n$.\r\nThis cannot happen since $a^{n}\\equiv 1 \\mod p$ be Lemma 1.\r\n\r\n[b]Corollary:[/b] There are $\\infty$ many primes $p \\equiv 1 \\mod n$.\r\nProof: easy, try yourself.",
"Solution_2": "[b]Corollary:[/b] There are $\\infty$ many primes $p \\equiv 1 \\mod n$.\r\nProof: easy, try yourself.[/quote]\r\n\r\nDoesn't it come directly from Dirichlet's theorem?",
"Solution_3": "Tried to prove the full Dirichlet theorem yourself\u00bf ;)",
"Solution_4": "[quote=\"ZetaX\"]Tried to prove the full Dirichlet theorem yourself\u00bf ;)[/quote]\r\n\r\n[b]I[/b], prove [b]Dirichlet's theorem[/b]? are you nutts? :P\r\n\r\nokey, here's the proof of the corollary:(compare with Euclid's)\r\n\r\nassume that there are only finitely many primes of the form $tn+1$.\r\nLet $P_{k}=kn+1$ be the largest prime of that form. \r\nconsider the number $P=P_{1}P_{2}...P_{k}+1=(n+1)(2n+1)...(kn+1)+1$. note that $P$ is a composite number of the form $tn+1$. and a number of this form must have a prime divisor of this form. but it is relatively prime to the rest of the prime numbers of the form. so we can conclude that a prime > $P_{k}$ divides $P$ and we are done.$\\blacksquare$\r\n\r\nPlease tell me the solution's correct!",
"Solution_5": "[quote=\"nayel\"][quote=\"ZetaX\"]Tried to prove the full Dirichlet theorem yourself\u00bf ;)[/quote]\n\nnote that $P$ is a composite number of the form $tn+1$. and a number of this form must have a prime divisor of this form. [/quote]\r\n\r\ncan someone explain this fact? how about you ZetaX?",
"Solution_6": "I don't know what you mean...\r\n- $P$ is not of the form $kn+1$, otherwise $P-1 = P_{1}\\cdots P_{k}$ would be a multiple of $n$... \r\n- If $P$ is of the form $kn+1$, then not necessarily every prime factor of $P$ is of that form! Take $n=4$, $9 = 3 \\cdot 3$\r\n\r\nZetaX did not write all that very interesting stuff for nothing... I think you shoud use those facts to prove che \"corollary\" :roll:\r\nbye!",
"Solution_7": "[quote=\"edriv\"]ZetaX did not write all that very interesting stuff for nothing... I think you shoud use those facts to prove che \"corollary\" :roll:\n[/quote]\r\n\r\nwell i'll try to study them soon but not rigth now. i have important exams. but don't worry i saved them in my file. :D",
"Solution_8": "[quote=\"edriv\"]ZetaX did not write all that very interesting stuff for nothing... I think you shoud use those facts to prove che \"corollary\" :roll:\nbye![/quote]\r\nYes, that's how it was intended. I know no easier way to prove Dirichlet's theorem for $1 \\mod m$, $m$ arbitrary. The case $-1 \\mod m$ is still possible, but harder (same for some other types of special cases).\r\n\r\nSorry by the way for not responding earlier, I was away the next day and didn't see this thread when I returned...",
"Solution_9": "[b]Lemma 1[/b] the set of prime divisors of the image set of a polynomial with integer coeffcients is infinite. (well, we suppose deg(P) > 0)\r\nBy contradiction: if $p_{1},p_{2},\\ldots,p_{k}$ are the elements of this finite set and $P(x) = a_{n}x^{n}+\\ldots+a_{1}x+a_{0}$, then $P(p_{1}p_{2}\\cdots p_{k}a_{0}) = a_{0}(a_{0}p_{1}\\cdots p_{k}(\\mbox{stuff})+1)$, where stuff can be very big and $p_{i}\\nmid a_{0}p_{1}\\cdots p_{k}(\\mbox{stuff})+1$ for i=1,2,...,k.\r\n\r\nNow just apply Lemma 1 to $\\Phi_{n}(x)$. Since the prime divisors of n are finite, and every prime p such that $p \\mid \\Phi_{n}(a), p \\nmid n$ is 1 modulo n, there are infinite primes of the form $kn+1$.\r\n\r\nGood night $\\zeta x$!",
"Solution_10": "Or else take a prime divisor of $ \\Phi_{n}(n)$ . Prime p divides $ \\Phi_{n}(n)$ divides $ n^{n}\\minus{}1$ implies $ (p,n)\\equal{}1$ and $ p\\equal{}1(modn)$ and there are infinitely many of them"
}
{
"Tag": [
"algebra",
"polynomial",
"algebra unsolved"
],
"Problem": "Are there any useful properties of polynomials that have positive values for all integers? Or all real numbers?",
"Solution_1": "What do you mean by \"useful\"? I could make up some weird properties that are true, like, if\r\n$ f(x)\\equal{}a_{n}x^{n}\\plus{}\\cdots\\plus{}a_{1}x\\plus{}a_{0}$\r\nis a polynomial, then $ a_{n}^{2}\\plus{}\\cdots\\plus{}a_{1}^{2}\\plus{}a_{0}^{2}\\geq 0$. Is this \"useful\"? Probably not, but it fulfills your requirement.",
"Solution_2": "This is a very vague question, but perhaps you have something like [url=http://en.wikipedia.org/wiki/Hilbert%27s_seventeenth_problem]Hilbert's 17th problem[/url] in mind. The solution to this problem (one of the first to be solved) says in particular that a polynomial which takes only non-negative values over the reals is a sum of squares of rational functions."
}
{
"Tag": [
"AMC",
"AIME",
"function",
"number theory"
],
"Problem": "AIME problem #1 from 1997, a sample problem on the AMC website:\r\n\r\nHow many of the integers between 1 and 1000, inclusive, can be expressed as the difference of the squares of two nonnegative integers?\r\n\r\nI have failed to solve this problem. This question #1 from an AIME seems considerably more difficult than the #1's from most AIME's. Any insight on the solution would be appreciated, thanks.\r\n\r\nFierytycoon",
"Solution_1": "The difference of squares = a:^2:-b:^2: = (a-b)(a+b). What can you tell about parity (even/odd) of these two brackets?",
"Solution_2": "All the odd numbers would be acheivable, because if x and y are consecutive, (x+y)(x-y) = 1 x (x+y) which would be odd because x+y = x+x+1 = 2x+1 which will be a odd number. Hope I was a help.",
"Solution_3": "Thanks for your responses. Hmm, well, I have gotten to a certain point:\r\n\r\nLet us say that 1 <= n <= 1000, and let n = a :^2: - b :^2: = (a+b)(a-b).\r\nNow let us take two cases.\r\n\r\n1) Let a = b + 1. Substituting, we have n = 2b + 1, and so for b = 0, 1, 2, ..., all odd numbers n can be satisfied.\r\n\r\n2) Let a = b + 2. Substituting, we have n = 4(b + 1), and so for b = 0, 1, 2, ..., all even numbers n that are divisible by 4 can be satisfied.\r\n\r\nThus, 3/4 of the cases can be satisfied, leading to an answer of 3/4 * 1000 = 750, which is the correct answer. However, I am still a bit vexed. I can show that 3/4 of the cases can be satisfied, but I have NOT shown that the other 1/4 of the cases CANNOT be satisfied. How do I know that if I let a = f(b), where f(b) is some function of b that I have not yet tried (such as f(b) = b + k), then more n could be satisfied?\r\n\r\nFierytycoon",
"Solution_4": "Two proofs:\r\n\r\nProof one: All perfect squares are congruent to 0 or 1 (mod 4), so you can't get to 2 (mod 4) by subtracting them.\r\n\r\nProof two: When you factor the difference as (a-b)(a+b), either both factors are even or both are odd. If they're both odd, you're not going to get 2 (mod 4). If they're both even you'll get a number divisible by 4, so it will be 0 (mod 4).",
"Solution_5": "ComplexZeta,\r\n\r\nThanks for your reply. However, since I am not too proficient at number theory (currently reading up on it right now), I do not understand how you came up with your statement:\r\n\r\n\"If they're both odd, you're not going to get 2 (mod 4).\"\r\n\r\nIs there a quick proof?\r\n\r\nFierytycoon",
"Solution_6": "If a and b are both odd, a+b and a-b are both even. So (a+b)(a-b) is a multiple of 4, and thus can't be 2 mod 4.",
"Solution_7": "TripleM,\r\n\r\nThanks for yoru response. However, I was confused about the case when the factors, not a and b, were both odd. as I understand, the \"they're\" in \"they're both odd\" refer to the factors, not a and b. I think you thought that it was a reference to a and b, and thus, took the case when the factors were even. :D \r\n\r\nFierytycoon",
"Solution_8": "Oh whoops, your right. Well if both brackets are odd, then their product is odd but 2 (mod 4) is always even.",
"Solution_9": "lol, whoops, sorta missed that...that was sorta simple. :lol: Anyways, thanks for your response TripleM.\r\n\r\nFierytycoon",
"Solution_10": "Ahhh.. i remember this. It wasn't that hard, just think about how to express all odds and evens as diff of squares. [i]Especially the evens.[/i]",
"Solution_11": "This problem is made simple using parity. All of the numbers from 1-1000 can be expressed as the difference of two squares if and only if (a+b)(a-b) both have the same parity. All numbers work that are odd and even, but there are special cases when the number is in the form, 4n+2. Because this factors as 2(2n+1) which have different parities. This happens 1/4 of the time. So 1000-1000/4= 1000-250=750 which is your answer hope this helped."
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "Let x, y, z, t be acute angles with sum \\pi. Show that\r\n1/cos(x) + 1/cos(y) + 1/cos(z) + 1/cos(t) <= \\sqrt (2)(tan(x)+tan(y)+tan(z) + tan(t))\r\n\r\n[i]unknown source[/i]\r\n\r\nNamdung",
"Solution_1": "I mus admit that this is a better variant of the inequality proposed by me in the Proposed problems section. But it is exactly the same problem. I won't post my solution now, I prefer to let all the others think a little bit about the problem.",
"Solution_2": "Yes, and solution based one Harazi indentity\r\nSum (1+tg(x)^2)/(tg(x)+t(y)) = Sum tg(x) :) \r\n\r\nNamdung"
}
{
"Tag": [
"function",
"integration",
"real analysis",
"real analysis unsolved"
],
"Problem": "Let $ f: [0,1]\\to\\mathbb{R}$ be a continuous function such that\r\n\\[ \\int_{0}^{1}f(x)dx\\equal{}\\int_{0}^{1}xf(x)dx.\\]\r\nProve that there exists $ c\\in (0,1)$ such that\r\n\\[ \\int_{0}^{c}xf(x)dx\\equal{}0.\\]",
"Solution_1": "Yet another typical Cezar Lupu problem.\r\n\r\n\r\n[color=blue]Lemma. Suppose that $ g : [0,1]\\to\\mathbb R$ is differentiable and that $ \\frac{g(1)\\minus{}g(0)}{1\\minus{}0}\\equal{} g^{\\prime}(0)$. Then, there is $ \\xi\\in (0,1)$ such that\n\\[ g^{\\prime}\\left(\\xi\\right)\\equal{}\\frac{g\\left(\\xi\\right)\\minus{}g(0)}{\\xi\\minus{}0}.\\]\n[/color]\r\n\r\n[i]Proof.[/i] Let $ h(x) \\equal{}\\frac{g(x)\\minus{}g(0)}{x\\minus{}0}$. If there is $ \\xi\\in (0,1)$ such that $ h^{\\prime}\\left(\\xi\\right) \\equal{} 0$, then we're done.\r\nElse, WLOG $ h$ is strictly increasing, i.e. $ \\frac{g(1)\\minus{}g(0)}{1\\minus{}0}>\\frac{g\\left(\\epsilon\\right)\\minus{}g(0)}{\\epsilon\\minus{}0}$.\r\nLetting $ \\epsilon\\to 0$, we see that $ \\frac{g(1)\\minus{}g(0)}{1\\minus{}0}> g^{\\prime}(0)$, contradiction.\r\n\r\n\r\nNow just apply this lemma to $ G$ such that $ G^{\\prime\\prime}\\equal{} f$."
}
{
"Tag": [
"geometry",
"3D geometry",
"sphere"
],
"Problem": "let $l,m,n$ be three lines ,wich we draw on earth and intersect each other in point $s$ prove that these three lines\r\n\r\nalso have another common point.",
"Solution_1": "What do you mean by a line? How do you define straight on a sphere?\r\nIsn't that like showing a \"line\" is a great circle of a sphere? If you define a line to be a great circle, then the result is trivial.",
"Solution_2": "But if they aren't great circles, isn't the result untrue? :huh:",
"Solution_3": "ure right guys ...\r\n\r\nand because of this i said:we draw them on $earth...$\r\n\r\ni want say that a line is a circle wich its radius is very very geat... [/list]"
}
{
"Tag": [
"geometry"
],
"Problem": "Let $ABCD$ be a cyclic quadrilateral. Let $X$ be the foot of the perpendicular from the point $A$ to the line $BC$, let $Y$ be the foot of the perpendicular from the point $B$ to the line $AC$, let $Z$ be the foot of the perpendicular from the point $A$ to the line $CD$, let $W$ be the foot of the perpendicular from the point $D$ to the line $AC$.\r\n\r\nProve that $XY\\parallel ZW$.\r\n\r\n Darij",
"Solution_1": "What is the QEDMO?",
"Solution_2": "The QED is a German mathematics association with about 150 members, most of them young (middle school, high school university students), many of them participants in regional,national or even international mathematical competitions.\r\n\r\nThe QEDMO is some sort of mathematical competition inside the QED. It has taken place twice so far, and works somewhat like this: When there's a meeting of the QED with a sufficient number of participants, the QEDMO Jury proposes a list of 14 problems. The participants are now split into two teams. Both teams now try to solve as many problems as possible (as QED meetings are approx. 3-4 days, problems are given to the teams on the first evening, solutions will be presented on the last day of the meeting), then present their solutions as follows:\r\n\r\nOne team starts and asks the other team to present their solution to a particular problem. The other team may accept or decline the challenge. If they decline, the first team has to present the solution. The team which presents their solution chooses one of their team members which will do that, the other team chooses a critic who may ask questions during the presentations and point out mistakes in the solution. If there are mistakes in the solutions that the presenting team's \"speaker\" cannot resolve, the critic is given the opportunity to correct the mistake or give his own proof. At the end, the jury awards 12 points distributed among the teams: For example, if the presenting team has solved the problem without a flaw, they will get 12 points. If the presenting team has made a minor mistake which they can't correct but the critic can, the critic's team receives 2 points, the presenting team receives 10; you get the idea. If the problem can't be solved by either the presenter or the critic, the jury will be awarded all remaining points. After points are distributed, the other team now asks for a solution to another problem.\r\n\r\nThere are some additional rules which I don't remember or can't explain correctly, but I think you get the general idea. If there are mistakes in my description (which is quite possible because I just attended 1 of the 2 QEDMOs so far), I'm sure the jury members will point them out.",
"Solution_3": "[quote=\"darij grinberg\"]Let $ABCD$ be a cyclic quadrilateral. Let $X$ be the foot of the perpendicular from the point $A$ to the line $BC$, let $Y$ be the foot of the perpendicular from the point $B$ to the line $AC$, let $Z$ be the foot of the perpendicular from the point $A$ to the line $CD$, let $W$ be the foot of the perpendicular from the point $D$ to the line $AC$.\n\nProve that $XY\\parallel ZW$.\n\n[/quote]\r\nIndeed very easy. :) \r\n\r\nSince $\\angle{AXB}+\\angle{AYB}=\\frac{\\pi}{2}+\\frac{\\pi}{2}=\\pi$, so $A, X, B, Y$ are cyclic. Hence $\\angle{ABX}=\\angle{AYX}$. \r\n\r\nSince $\\angle{AWD}=\\angle{AZD}=\\frac{\\pi}{2}$, so $A, W, Z, D$ are cyclic. Therefore $\\angle{AWZ}+\\angle{ADZ}=\\pi$.\r\n\r\nBecause $A, B, C, D$ lie on the same circle, so we have: $\\angle{ABX}=\\angle{ADZ}$, which means:$\\angle{ABX}+\\angle{AWD}=\\pi$, so $\\angle{AYX}+\\angle{AWD}=\\pi$, thus $\\angle{XYW}=\\angle{YWZ}$, which means $XY || WZ$."
}
{
"Tag": [],
"Problem": "Last year I was new to math, and applied to the only two summer math programs I knew of, however this year I know of a bunch and I was thinking about applying to 4 or 5, but that seems somewhat excessive. I mean the applications are fun so I will do them anyway, but I was wondering how many other people are applying to.",
"Solution_1": "Perhaps a poll would help show this. Me personally probably two or three.",
"Solution_2": "well i am only applying to one math one",
"Solution_3": "What happened to me was I found a couple I was really interested in, and decided to start applying to all of them. However, the response (and acceptance) from one of the first two camps I applied to came before the deadline for the application to the rest, so I only ended up applying to two.",
"Solution_4": "At my level and what I was looking for in a math camp, I considered several but ended up applying to only Awesomemath. If I didn't get in, then I would have considered other options. However, I am glad that it didn't come to that. Apply to as many as you want, but make sure that you know what each offers and for what you are looking."
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "Find the finished digit of this number \\[ 9999^{999^{99^9}}\\]",
"Solution_1": "You should not post such a problem in this forum.\r\nPlease make sure where you should post some problem?\r\n\r\nJapanese Communities Moderator\r\n\r\nkunny"
}
{
"Tag": [
"AMC",
"AIME",
"trigonometry",
"geometry",
"\\/closed"
],
"Problem": "I have lots of questions: \r\n1. Is it possible to take 2 classes in one day: AIME/ Intermediate Trig?\r\n2. I understand 70% of Intro Counting, but only 10% of Intermediate Counting. Should I take intro, or can I learn it through the aops book?\r\n3. How much will I be affected if I miss the wednesday geometry class, but am there for fridays? I really want to learn geometry. Can I catch up?\r\n4. How much classes it possible to take?\r\n5. Does this schedule look good?\r\n\r\nTuesday: Intermediate Trig or/and AIME\r\nWed: Geometry\r\nThursday: Intro to Counting\r\nFriday: Geometry\r\n\r\n6. Also: will more courses be offered in the spring of 2005?",
"Solution_1": "Some answers:\r\n\r\n1. Is it possible to take 2 classes in one day: AIME/ Intermediate Trig?\r\n\r\nIf the classes are scheduled at different hours, there is not reason why you couldn't. However if the classes are at the same hour, even if you now can actually log in into two instances of classes, I belive it is hard to catch up on two different subjects online at the same time. \r\n\r\nWhat you could do is to alternatively participate in each class, obtaining the transcript for the other one. \r\n\r\n2. I understand 70% of Intro Counting, but only 10% of Intermediate Counting. Should I take intro, or can I learn it through the aops book?\r\n\r\nYou should definitely take a class if you feel that you only handle 70% of the stuff the class is covering. In plus, almost all intermediate classes require matters discussed previously in the intro class. \r\n\r\n3. How much will I be affected if I miss the wednesday geometry class, but am there for fridays? I really want to learn geometry. Can I catch up?\r\n\r\nYou can use transcripts to see what happened in the class. Catching up differs from individual to individual, but usually it's not much of a problem. \r\n\r\n4. How much classes it possible to take?\r\n\r\nYou can take as many classes as you please. You can even register for classes just to take the transcripts, there is no need for you to actually be present in each class. \r\n\r\n5. Does this schedule look good?\r\n\r\nTuesday: Intermediate Trig or/and AIME\r\nWed: Geometry\r\nThursday: Intro to Counting\r\nFriday: Geometry\r\n\r\nRichard, Mathew or David can answer this one :) \r\n\r\n6. Also: will more courses be offered in the spring of 2005?\r\n\r\nYes, AoPS plans to offer more courses in the future. Just stay tooned :) \r\n\r\nHope this helps.",
"Solution_2": "When you say you understand 70% of the material, does that mean you can solve 70% of the problems? If so, that class will probably move too slowly for you. On the other hand, if you know 70% of the topics, it would still be wise for you to take the class. Some of the problems covered in class are quite difficult, and it will be useful for you to see how to apply the theorems to solving challenging problems.",
"Solution_3": "I'll second what ComplexZeta said. It is possible to be familiar with essentially ALL the topics in an AoPS course (from previous curricular math study in school) and still need lots practice in actually solving problems on those topics. My advice, when in doubt, is to register for the easier of two AoPS courses if you think your own level is somewhere in between the level of each course. \r\n\r\nGood luck in your course. I think the AoPS courses are an excellent, challenging look beyond what is usually taught in school math courses."
}
{
"Tag": [
"abstract algebra",
"linear algebra",
"matrix",
"geometry",
"projective geometry",
"superior algebra",
"superior algebra solved"
],
"Problem": "Hello all mathematicians out there!.\r\n\r\nLef u:L->F be a homomorphism of finitely generated free A-modules (A=commutative local ring with identity), fix a pair of bases for L and F respectively, and let M be the associated matrix of u. If n=rank(F) and N is the annihilator of the cokernel of u, show that the radical of N equals the radical of I, where I is the ideal of A generated by all nxn minors of M. \r\n(We assume rank(L)>rank(F)). \r\n\r\nRemark: I found this problem in a commutative algebra book by C.Peskine (An algebraic introduction to complex projective geometry), the author claims this is an important result, so i would appreciate any hint on how to prove the above statement.\r\n\r\n\r\nThanks!",
"Solution_1": "I'll prove that $I\\subset (Imu:F)=(0:F/Imu)=(0:coker u)=Ann(coker u)$. Actually I'll only prove the inclusion because the equalities are obvious :D \r\n\r\nTo prove the inclusion we only have to prove that the generators of $I$ are in $(Imu:F)$. \r\nIn Peskine, $M$ is defined as an $n\\times l$ matrix, $u(e_i)=\\sum_{j=1}^na_{ji}f_j\\ \\forall i=\\overline{1,l}$, where $rank(L)=l$.\r\n\r\nThe generators of $I$ are $n\\times n$ minors. Choose such a minor $x$. There exists a matrix $B$ such that $MB=xI_n$. Just put the comatrix of the matrix yielding that minor on the corresponding positions in $B$ and put $0$ in the rest of $B$.\r\nSo $x\\cdot f_i=\\sum_{j=\\overline{1,n},k=\\overline{1,l}}f_j\\cdot a_{jk}b_{ki}=\\sum_{k=\\overline{1,l}}u(e_k)b_{ki}\\in Imu\\ \\forall i=\\overline{1,n}\\Rightarrow x\\in Ann(coker u)\\Rightarrow I\\subset (Imu:F)$",
"Solution_2": "Now I prove $(Imu:F)\\subset\\sqrt I$.\r\n\r\n$x\\in (Imu:F)\\Leftrightarrow \\exists B=(b_{ij})_{i=\\overline{1,n}j=\\overline{1,l}}$ such that $x\\cdot f_i=\\sum_{j=1}^lu(e_j)b_{ji}$\r\nThis is easily seen, using the fact that $F$ is free, to be equivalent to $xI_n=AB$.\r\nThis means $x^n=\\det AB$. Using the Binet-Cauchy Formula, we get that $x^n\\in I$, meaning $x\\in\\sqrt I$\r\n\r\nWith this, the problem is solved."
}
{
"Tag": [],
"Problem": "Factor $ x^4 \\plus{}4$",
"Solution_1": "you can't :?:",
"Solution_2": "maybe you meant\r\n$ x^4 \\minus{} 4$\r\n\r\ncause I'm pretty sure you can factor that.\r\nAs for $ x^4 \\plus{} 4$..i cant find a way..",
"Solution_3": "I know there is an answer. Because I was worked handing out the questions in a math competition in my school and that problem showed up.. They showed the answer .In the answer they put something like $ \\minus{} 2i$ the $ i$ is an imaginary number.\r\n\r\nSo I think $ x^4 \\plus{} 4$\r\n$ \\equal{} x^4 \\minus{} ( \\minus{} 4)$\r\n$ \\equal{} x^4 \\minus{} ( \\minus{} 2i)^2$\r\n$ \\equal{} (x^2 \\plus{} ( \\minus{} 2i))(x^2 \\minus{} ( \\minus{} 2i)$ \r\n\r\nI am sure that the question is $ x^4 \\plus{} 4$ and that the answer has to do with imaginary numbers.",
"Solution_4": "If you wanted to factor this expression over integers you can add $ 4x^2$ and subtract $ 4x^2$ to obtain $ (x^4 \\plus{} 4x^2 \\plus{} 4) \\minus{} (4x^2)$ which equals $ (x^2 \\plus{} 2)^2 \\minus{} 4x^2$ which equals $ (x^2 \\plus{} 2x \\plus{} 2)(x^2 \\minus{} 2x \\plus{} 2)$.",
"Solution_5": "@phi-unit\r\nThat im preatty sure is also an answer.. there were two answers in the answer sheet they gave me at the competition and one had imaginary numbers and the other one had something like that.",
"Solution_6": "$ a^4\\plus{}4b^4 \\equal{}(a^2\\plus{}2b\\plus{}2b^2)(a^2\\minus{}2b\\plus{}2b^2)$ is also known as the Sophie Germain identity, I believe.",
"Solution_7": "Yes, and also, it turns out that in order for $ a^{4}\\plus{}n$ to be factorable, $ n$ must be of the form $ 4b^{4}$ for some integer $ b$. This was proved by Sophie Germain, and is why the identity is named for her (because the factorization, compared to the proof that $ n\\equal{}4b^{4}$, is quite obvious)."
}
{
"Tag": [
"USAMTS"
],
"Problem": "Hello,\r\n\r\nI am a USAMTS participant (USAMTS ID: 12048) and today is the deadline for the Round 1 Solutions. I attempted to submit my solutions from the home page, however, I recieve the following message: Sorry, the deadline has passed for submission. The due date was October 13, 2009.\r\n\r\nWhy is this? Am I not allowed to submit my solutions on the due date?\r\n\r\nAny feedback would be helpful.\r\n\r\nThanks,\r\n\r\n-- \r\n-SB",
"Solution_1": "Sorry dude... the deadline for uploading was at 3 pm pacific...\r\n\r\nSEE IF YOU CAN DRIVE TO THE POST OFFICE REALLY QUICK AND MAIL YOUR SOLUTIONS!!!",
"Solution_2": "To be clear, the deadline was 3 PM [b]Eastern Time[/b] on the due date. This information is on the first page of the problem document, and in several places on the USAMTS website (including the page where you download the problems)."
}
{
"Tag": [
"topology"
],
"Problem": "What is the Int. Math. Union logo exactly? I know it represents something in topology..\r\n\r\nsee http://mathunion.org/",
"Solution_1": "it's a picture of the (actually, a) borromean link..\r\ni think wikipedia has something to say about it :)\r\nthe first thing you can actually notice it that they unknot (unlink, if you prefer) as soon as you remove one component.. the second thing you could prove is that they're knotted (linked), but that's not as easy as the first ;)\r\n\r\nanyway, it seems a ropelength-minimizing representation: just take three \"real\" ropes with a constant section and try to knot them up like that using the least rope you can. what you seem to obtain (by computer approximations) should have a shape like that, just thickened as much as possible.\r\nby the way, the theory who tries to solve this kind of problems is called physical knot theory, in case you're interested.",
"Solution_2": "thanks for the fast answer :-) I see why this would be used as the 'MAIN' logo for an international mathematical union..kind of"
}
{
"Tag": [
"geometry"
],
"Problem": "For number 6, it says we can always remove 7 magazines but does it mean that we know the areas of each magazines or are they random and we have to explain how to remove in such a way.[/quote]",
"Solution_1": "you would know the areas of the magazines you remove. But the magazines could be any size, shape, area, arrangement, etc. as long as they entirely cover the table.\r\n\r\nIm pretty sure thats all the clarification I can give without breaking the rules.",
"Solution_2": "that is all I need. Thank you."
}
{
"Tag": [
"algebra proposed",
"algebra"
],
"Problem": "The real sequence is defined by $ x_1 \\equal{} 1, x_{n \\plus{} 1} \\equal{} \\frac {1}{x_1 \\plus{} x_2 \\plus{} ... \\plus{} x_n}$\r\nShow that $ x_1 \\plus{} x_2 \\plus{} ... \\plus{} x_n > 1989$ for sufficiently large $ n$.",
"Solution_1": "[quote=\"tdl\"]The real sequence is defined by $ x_1 \\equal{} 1, x_{n \\plus{} 1} \\equal{} \\frac {1}{x_1 \\plus{} x_2 \\plus{} ... \\plus{} x_n}$\nShow that $ x_1 \\plus{} x_2 \\plus{} ... \\plus{} x_n > 1989$ for sufficiently large $ n$.[/quote]\r\n[hide]\n$ \\frac{1}{x_n}\\equal{}x_1\\plus{}x_2\\plus{}\\ldots\\plus{}x_{n\\minus{}1}$\n\n$ x_{n\\plus{}1}\\equal{}\\frac{1}{\\frac{1}{x_n}\\plus{}x_n}\\equal{}\\frac{x_n}{x_n^2\\plus{}1}$ and the partial sums diverge...[/hide]",
"Solution_2": "All terms are positive, so if we can find one $ N$ such that $ \\sum_{k\\equal{}1}^N x_k >1989$, we'll be done. Suppose no such $ N$ exists. Then $ x_k\\geq \\frac{1}{1989}$ for all $ k$, and we get\r\n\\[ \\sum_{k\\equal{}1}^{1989^2\\plus{}1} x_k \\geq \\frac{1989^2\\plus{}1}{1989}>1989\\]\r\nanyway, a contradiction."
}
{
"Tag": [
"topology"
],
"Problem": "Let $ X$ be a topological space. Prove that: \r\n\r\n$ X$ is Hausdorff if and only if the diagonal $ \\{ (x,x) | x \\in X \\}$ is closed in $ X \\times X$.",
"Solution_1": "A set is closed if its complement is open. A set is open if we can fit one of the base open sets around any given point in it. A base for the product topology on $ X\\times X$ is the products of open sets.\r\n\r\nNow, what does it mean if for any $ (a,b)$ not on the diagonal, we can find $ U,V$ with $ (a,b)\\in U\\times V$ and $ U\\times V$ not intersecting the diagonal?"
}
{
"Tag": [
"number theory unsolved",
"number theory"
],
"Problem": "find funcition f N -> N for all n is integer n \\geq 0\r\n f(n)+f(n+1)=f(n+3)f(n+4)+2004 :D :?",
"Solution_1": "No solution ? It easy? :?"
}
{
"Tag": [
"calculus",
"integration",
"number theory",
"prime numbers"
],
"Problem": "What is the least odd integer greater than 4 that cannot be written as the sum of an integral power of 2 and a prime number?",
"Solution_1": "hmm dont really know how to do this.. is it infinity?",
"Solution_2": "maybe try finding odd numbers that are not prime when you subtract powers of two? :maybe:",
"Solution_3": "Yes.. and the answer is? I still cant get it, but I probably got careless and missed something\r\n\r\nwhats the answer?",
"Solution_4": "how do u solve this?",
"Solution_5": "Is it $25$ ?",
"Solution_6": "[quote=\"Astronum\"]Is it $25$ ?[/quote]\r\n$25=17+8=23+2$",
"Solution_7": "[quote=\"lotrgreengrapes7926\"][quote=\"Astronum\"]Is it $25$ ?[/quote]\n$25=17+8=23+2$[/quote]\r\n\r\nOps, my bad :blush: ... that's right. Thank you. :)",
"Solution_8": "I can't find an answer to this question, are you sure it is possible?",
"Solution_9": "[quote=\"mathgeniuse^ln(x)\"]I can't find an answer to this question, are you sure it is possible?[/quote]\r\nYes, it is possible. May be I will give some more time before posting the answer.",
"Solution_10": "Can you give us a hint first?",
"Solution_11": "[hide]149[/hide]",
"Solution_12": "[quote=\"perfect628\"][hide]149[/hide][/quote]\r\nYou should have given a hint. :roll: \r\nAny solution?",
"Solution_13": "My only solution was trial and error, by taking odds, subtracting the greatest power of 2, and seeing if it was prime. If not, I continued until I found a prime. 149 never yielded a prime.",
"Solution_14": "You could always try to write all the prime numbers and see what happens to the numbers you get. Maybe a pattern will occur. :?:",
"Solution_15": "[quote=\"perfect628\"][hide]149[/hide][/quote]\r\nUnfortunately, I dont have a method for this. I also did it by trial and error. \r\n\r\nperfect628's answer is incorrect. My only hint is that it is less than 149. If anyone knows a nice method for this, that will be nice.",
"Solution_16": "[quote=\"lotrgreengrapes7926\"]You should have given a hint. :roll: [/quote] Oops, sorry. I thought you were the one who posted the problem. :blush:",
"Solution_17": "[hide]I get 127 by Trial and Error.[/hide]",
"Solution_18": "[hide]95-16=79, so no, its not 95.[/hide]",
"Solution_19": "[hide]I meant 127, don't know where 95 came from. :?: \n\nI edited it in the previous post.[/hide]",
"Solution_20": "[quote=\"mathgeniuse^ln(x)\"]I meant 97, don't know where 95 came from. :?: \n\nI edited it in the previous post.[/quote]\r\n\r\n89+8=97",
"Solution_21": "I edited both of my previous posts. I don't know where 97 came from.",
"Solution_22": "[hide]127?I think that works.[/hide]",
"Solution_23": "Yeah, I think you actually finally found the answer.",
"Solution_24": "[quote=\"mathgeniuse^ln(x)\"][hide]I meant 127, don't know where 95 came from. :?: \n\nI edited it in the previous post.[/hide][/quote]\r\nCorrect! 127 it is.\r\n\r\nI am not sure if there is a sleek way to do this problem quickly without resorting to trial and error :?:",
"Solution_25": "Yeah, I just wrote all the primes and powers of 2 up to 150 and realized that 127 is impossible to achieve.",
"Solution_26": "primes are positivie right?",
"Solution_27": "Yes.\r\n\r\nIf primes weren't positive then it would be really hard to check the smallest integer...if any.",
"Solution_28": "I was just wondering that because on some test I remember seeing a question that asked to find all positive prime numbers.",
"Solution_29": "[quote=\"mathgeniuse^ln(x)\"]Yes.\n\nIf primes weren't positive then it would be really hard to check the smallest integer...if any.[/quote]\r\n\r\nI thought there were negative primes too? \r\n\r\nhttp://www.mazes.com/primes/negative.html\r\n\r\nOr did you mean for this particular problem?",
"Solution_30": "I know it is possible for negative primes, but if they ever ask that, they would usually say positive or negative primes. If they asked for both it would be a little more trickier, but maybe doable."
}
{
"Tag": [
"vector",
"function",
"limit",
"inequalities",
"real analysis",
"real analysis unsolved"
],
"Problem": "Let $1 \\leq p <\\infty$.\r\na).Prove that $(l^p)^*=l^q$ in a sense that there is a bijection between $(l^p)^*$ and $l^q$, provided $\\frac{1}{p}+\\frac{1}{q}=1$.\r\nb).Prove that for $10$ we can find such $K$ that for all $k>K$ we have that $|x^{(k)}_n|<\\frac{\\epsilon}{2^n}$, so $||x^{(k)}||_{l^1} = \\sum |x^{(k)}_n|<\\epsilon$, which means that $||x^{(k)}||_{l^1} \\to 0$, when $k \\to \\infty$."
}
{
"Tag": [
"vector",
"linear algebra"
],
"Problem": "If A is a symmetric nxn mx of rank<=r>=1 and X is a unit eigenvector of A, with eigenvalue a not= 0, let B=A - aXX^T. Show that B is symmetric and that N(A) is a proper subspace of N(B). Conclude that rank B= [1, infinity) satisfying\r\n\r\n f(x).f(2x)...f(nx) < 2008.n^2009 for all any positive real number x and positive integer n",
"Solution_1": "we suppose that $ \\exists a\\in\\mathbb{R}^\\plus{}: \\ f(a)>1$\r\n$ f$ continous gives $ \\exists b>a,\\forall x\\in[a,b]: \\ f(x)\\ge \\frac{f(a)\\plus{}1}{2}$ (we take $ c\\equal{}\\frac{f(a)\\plus{}1}{2}>1$)\r\n\r\nnow for $ e>0$ we take $ n\\in\\mathbb{N}$ such that $ ne\\le b<(n\\plus{}1)e$\r\nwe have $ |\\{k\\in[|1,n|]: \\ f(ke)>c\\}|\\ge |\\{k\\in[|1,n|]: \\ ke\\in[a,b]\\}|\\ge \\frac{b\\minus{}a}{2b}n$\r\nthen $ 2008>\\frac{\\prod_{k\\equal{}1}^{n}f(ke)}{n^{2009}}\\ge \\frac{A^n}{n^{2009}}$ where $ A\\equal{}c^{\\frac{b\\minus{}a}{2b}}>1$\r\nbut we have $ \\lim_{n\\to \\plus{}\\infty} \\frac{A^n}{n^{2009}}\\equal{}\\plus{}\\infty$ (impossible)\r\n\r\nthen $ \\forall a\\in\\mathbb{R}^\\plus{}: \\ f(a)\\equal{}1$"
}
{
"Tag": [],
"Problem": "Points $E$ and $F$ are chosen on the sides $AB$ and $BC$ respectively of a square $ABCD$ such that $BE = BF$. Let $BN$ be an altitude of the triangle $BCE$. Prove that the triangle $DNF$ is right-angled.",
"Solution_1": "See the attached diagram."
}
{
"Tag": [
"AMC",
"AMC 10",
"algebra",
"binomial theorem"
],
"Problem": "Well, i was going through some problems and i was having trouble figuring this one out:\r\n\r\nSimplify:\r\n\r\nsqrt(3+sqrt8)-sqrt(3-sqrt8)\r\n\r\nI think it has something to do with the binomial theorem, i don't know how to get it though.",
"Solution_1": "[hide=\"hint\"]\n\nTry squaring it\n\n[/hide]",
"Solution_2": "[quote=\"eli140\"]Well, i was going through some problems and i was having trouble figuring this one out:\n\nSimplify:\n\n$ \\sqrt{3\\plus{}\\sqrt8}\\minus{}\\sqrt{3\\minus{}\\sqrt8}$\n\nI think it has something to do with the binomial theorem, i don't know how to get it though.[/quote]\r\n\r\nWell, if we square it, we get $ 3 \\plus{} \\sqrt{8} \\minus{} 3 \\sqrt8 \\equal{} x^2 \\implies x \\equal{} 0$\r\n\r\nI think it is $ 0$, should belong in HSB.",
"Solution_3": "--------------------------------------------------------------------------------\r\n\r\neli140 wrote: \r\nWell, i was going through some problems and i was having trouble figuring this one out: \r\n\r\nSimplify: \r\n\r\n \r\n\r\nI think it has something to do with the binomial theorem, i don't know how to get it though. \r\n\r\n\r\nWell, if we square it, we get \r\n\r\nI think it is , should belong in HSB.\r\n\r\nSame thing from me. eli is right",
"Solution_4": "@user above me: The \"quote\" button near the bottom right most part of every post will quote the entire thing a user said and you can post your own words along with it. that will help retain the format of the post.",
"Solution_5": "Um...I don't really know what you said, but just let $ x\\equal{}\\sqrt{3\\plus{}\\sqrt8}\\minus{}\\sqrt{3\\minus{}\\sqrt8}\\implies x^2\\equal{}3\\plus{}\\sqrt8\\plus{}3\\minus{}\\sqrt8\\minus{}2\\sqrt{9\\minus{}8}\\equal{}4\\implies x\\equal{}\\boxed2$ (since the first radical is larger than the second, and both are positive).",
"Solution_6": "007 u know me and anyway sorry meant to say you were right. and thanks!",
"Solution_7": "[quote=\"math154\"]Um...I don't really know what you said, but just let $ x \\equal{} \\sqrt {3 \\plus{} \\sqrt8} \\minus{} \\sqrt {3 \\minus{} \\sqrt8}\\implies x^2 \\equal{} 3 \\plus{} \\sqrt8 \\plus{} 3 \\minus{} \\sqrt8 \\minus{} 2\\sqrt {9 \\minus{} 8} \\equal{} 4\\implies x \\equal{} \\boxed2$ (since the first radical is larger than the second, and both are positive).[/quote]\r\n\r\nARGH!!!! I keep on forgetting that i must multiply each thingie by all the other thingie's...phailure.....",
"Solution_8": "Obviously, it cannot be zero. If it were zero, it would imply that $ \\sqrt{3\\plus{}\\sqrt{8}}\\equal{}\\sqrt{3\\minus{}\\sqrt{8}} \\implies \\sqrt{8}\\equal{}0$, which is false.",
"Solution_9": "As I said earlier (why don't people read? xD), the answer is $ 2$.",
"Solution_10": "Umm, I get where the two 3s come from. How do you get the 2sqrt(9-8)?",
"Solution_11": "It's just $ x^2\\equal{}(3\\plus{}\\sqrt8)\\plus{}(3\\minus{}\\sqrt8)\\minus{}2\\sqrt{3\\plus{}\\sqrt8}\\cdot\\sqrt{3\\minus{}\\sqrt8}\\equal{}6\\minus{}2\\sqrt{9\\minus{}8}\\equal{}4$.",
"Solution_12": "[quote=\"eli140\"]Well, i was going through some problems and i was having trouble figuring this one out:\n\nSimplify:\n\nsqrt(3+sqrt8)-sqrt(3-sqrt8)\n\nI think it has something to do with the binomial theorem, i don't know how to get it though.[/quote]\r\n\r\n[hide]$ \\begin{array}{l}\\sqrt {3 \\plus{} \\sqrt 8 } \\minus{} \\sqrt {3 \\plus{} \\sqrt 8 } \\equal{} \\sqrt {3 \\plus{} 2\\sqrt 2 } \\minus{} \\sqrt {3 \\minus{} 2\\sqrt 2 } \\equal{} \\sqrt {\\left( {1 \\plus{} \\sqrt 2 } \\right)^2 } \\minus{} \\sqrt {\\left( {1 \\minus{} \\sqrt 2 } \\right)^2 } \\\\\\equal{} 1 \\plus{} \\sqrt 2 \\minus{} \\left( { \\minus{} 1 \\plus{} \\sqrt 2 } \\right) \\equal{} 2 \\\\\\end{array}$[/hide]",
"Solution_13": "maybe this might help\r\n$ \\sqrt{(a\\plus{}b)\\pm2\\sqrt{ab}}\\equal{}\\sqrt{a}\\pm\\sqrt{b}$",
"Solution_14": "Wait why wouldn't it be:\r\n\r\n$ \\begin{array}{l}\\sqrt {3 \\plus{} \\sqrt 8 } \\minus{} \\sqrt {3 \\minus{} \\sqrt 8 } \\equal{} \\sqrt {3 \\plus{} 2\\sqrt 2 } \\minus{} \\sqrt {3 \\minus{} 2\\sqrt 2 } \\equal{} \\sqrt {\\left( {1 \\plus{} \\sqrt 2 } \\right)^2 } \\minus{} \\sqrt {\\left( {1 \\minus{} \\sqrt 2 } \\right)^2 } \\\\\\equal{} 1 \\plus{} \\sqrt 2 \\minus{} \\left( {1 \\minus{} \\sqrt 2 } \\right) \\equal{} 2\\sqrt 2 \\\\\\end{array}$",
"Solution_15": "[quote=\"Fibonacci11235\"]Wait why wouldn't it be:\n\n$ \\begin{array}{l}\\sqrt {3 \\plus{} \\sqrt 8 } \\minus{} \\sqrt {3 \\minus{} \\sqrt 8 } \\equal{} \\sqrt {3 \\plus{} 2\\sqrt 2 } \\minus{} \\sqrt {3 \\minus{} 2\\sqrt 2 } \\equal{} \\sqrt {\\left( {1 \\plus{} \\sqrt 2 } \\right)^2 } \\minus{} \\sqrt {\\left( {1 \\minus{} \\sqrt 2 } \\right)^2 } \\\\\\equal{} 1 \\plus{} \\sqrt 2 \\minus{} \\left( {1 \\minus{} \\sqrt 2 } \\right) \\equal{} 2\\sqrt 2 \\\\\\end{array}$[/quote]\r\n\r\nBecause $ \\sqrt{(1\\minus{}\\sqrt2)^2}\\equal{}|1\\minus{}\\sqrt2|\\equal{}\\sqrt2\\minus{}1$. Always check that--but personally, I think my method is easier for this problem.",
"Solution_16": "Oh, thanks. :oops: \r\n\r\nI would have totally gotten that wrong."
}
{
"Tag": [],
"Problem": "This may be a silly question.\r\n\r\nYou pick 5 cards from 52, and at least two of them have to be aces. This can be done in how many ways?\r\n\r\nI know the answer: ${4\\choose2}{48\\choose3}+{4\\choose3}{48\\choose2}+{4\\choose4}{48\\choose1}$\r\n\r\nThis makes sense, and I understand it.\r\n\r\nHowever, I need someone to explain why ${4\\choose2}{50\\choose3}$ is incorrect.\r\n\r\nYou choose 2 aces from the 4, and then you choose three cards from the remaining 50 cards. Makes sense to me.\r\n\r\nAnyone? I'm tripping over a simple concept. :oops:",
"Solution_1": "[quote=\"mathkid\"]This may be a silly question.\n\nYou pick 5 cards from 52, and at least two of them have to be aces. This can be done in how many ways?\n\nI know the answer: ${4\\choose2}{48\\choose3}+{4\\choose3}{48\\choose2}+{4\\choose4}{48\\choose1}$\n\nThis makes sense, and I understand it.\n\nHowever, I need someone to explain why ${4\\choose2}{50\\choose3}$ is incorrect.\n\nYou choose 2 aces from the 4, and then you choose three cards from the remaining 50 cards. Makes sense to me.\n\nAnyone? I'm tripping over a simple concept. :oops:[/quote]\r\n$\\binom{4}{2}\\binom{50}{3}$ is incorrect because you aren't counting the possibility that you get 3 or 4 aces.",
"Solution_2": "Why not? Of those 50 cards, two will be aces, so choosing three could include two of those.\r\n\r\nAnd my incorrect answer is larger than the correct answer, if that means anything.",
"Solution_3": "[quote=\"mathkid\"]Why not? Of those 50 cards, two will be aces, so choosing three could include two of those.\n\nAnd my incorrect answer is larger than the correct answer, if that means anything.[/quote]\r\nIn that case... I think they're both correct. :rotfl:\r\nBut how? :rotfl:",
"Solution_4": "You only want to count a combination like As Ac Ah Ks (lolmaha) once, but you count it several times.\r\n\r\nFor example, you count it as (As Ac)(Ah Ks), (Ah As)(Ac Ks), (Ac Ah)(As Ks) in the quantity \\[\\binom{4}{2}\\binom{50}{3}\\]"
}
{
"Tag": [],
"Problem": "[size=150]\u5df2\u77e5$ \\Delta ABC$\u662f\u7b49\u8fb9\u4e09\u89d2\u5f62\uff0c$ O$\u662f$ \\Delta ABC$\u5185\u7684\u4efb\u610f\u4e00\u70b9\u3001$ OD\\perp BC$\uff0c$ OE\\perp AC$\uff0c$ OF\\perp AB$\uff0c\n$ D$\u3001$ E$\u3001$ F$\u5206\u522b\u662f\u5782\u8db3\uff0c\u8fde\u63a5$ OA$\u3001$ OB$\u3001$ OC$.\n\u6c42\u8bc1\uff1a$ S_{\\Delta OAF}\\plus{}S_{\\Delta OBD}\\plus{}S_{\\Delta OCE}\\equal{}\\dfrac{1}{2}S_{\\Delta ABC}$.[/size]",
"Solution_1": "[quote=\"wxd686\"][size=150]\u5df2\u77e5$ \\Delta ABC$\u662f\u7b49\u8fb9\u4e09\u89d2\u5f62\uff0c$ O$\u662f$ \\Delta ABC$\u5185\u7684\u4efb\u610f\u4e00\u70b9\u3001$ OD\\perp BC$\uff0c$ OE\\perp AC$\uff0c$ OF\\perp AB$\uff0c\n$ D$\u3001$ E$\u3001$ F$\u5206\u522b\u662f\u5782\u8db3\uff0c\u8fde\u63a5$ OA$\u3001$ OB$\u3001$ OC$.\n\u6c42\u8bc1\uff1a$ S_{\\Delta OAF} \\plus{} S_{\\Delta OBD} \\plus{} S_{\\Delta OCE} \\equal{} \\dfrac{1}{2}S_{\\Delta ABC}$.[/size][/quote]\r\n\r\n\u63d0\u793a\u56fe:\r\n(\u6709\u70b9\u8bef\u5dee)"
}
{
"Tag": [
"geometry",
"trigonometry",
"geometry proposed"
],
"Problem": "[color=darkred] Let $ABC$ be a triangle. The its incircle $C(I)$ touches the lines $AB$, $AC$ in the points $E$, $F$. Denote the middlepoint $M$ of the side $[BC]$. Prove that the $A$- simmedian and the lines $EF$, $MI$ are concurrently if and only if $\\boxed{\\cos B+\\cos C=1}$.[/color]",
"Solution_1": "[color=darkred][b]Generalization I.[/b] Let be $ABC$ an acute triangle. The its incircle $C(I)$ touches the lines $AB$, $AC$ in the points $E$, $F$. Denote the middlepoint $M$ of the side $[BC]$. Let $S\\in [BC]$ for which $\\frac{SB}{SC}=s$. Prove that the lines $AS$, $EF$, $MI$ are concurrently if and only if $\\boxed{\\frac{c-sb}{1-s}=\\frac{2(p-b)(p-c)}{a}}$. \n\n[b]Particular cases.[/b] \n\n$1.\\blacktriangleright$ For $S\\in BC$, $AS\\perp BC$, i.e. $s=\\frac{c\\cdot\\cos B}{b\\cdot\\cos C}$ obtain \n\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=141487 \n\n$2.\\blacktriangleright$ If the point $S$ is the foot of the $A$- simmedian, i.e. $\\frac{SB}{SC}=\\left(\\frac{c}{b}\\right)^{2}$, then \n\n$AS\\cap EF\\cap MI\\ne\\emptyset$ $\\Longleftrightarrow$ $\\cos B+\\cos C=1$.[/color]",
"Solution_2": "[color=darkred][b]Generalization II.[/b] Let be $ABC$ an acute triangle. The its incircle $C(I)$ touches the lines $AB$, $AC$ in the points $E$, $F$. Let $\\{M,S\\}\\subset (BC)$ be two points for which $\\frac{MB}{MC}=m$ and $\\frac{SB}{SC}=s$. Prove that the lines $AS$, $EF$, $MI$ are concurrently if and only if $\\boxed{(s-m)a^{2}+(m+1)(c-sb)a=(b-c)(s-1)(mb-c)}$. \n\n[b]Particular case.[/b] If $m=1$, i.e. the point $M$ is\n\nthe middlepoint of the side $[BC]$, then obtain the [u]generalization I[/u].[/color]"
}
{
"Tag": [
"calculus",
"integration",
"calculus computations"
],
"Problem": "In the $ xy$-plane, if $ C$ is the circle $ x^2\\plus{}y^2\\equal{}9$, oriented counterclockwise, then calculate $ \\int_C \\minus{}2ydx \\plus{} x^2dy$.\r\n\r\nI tried Green's Theorem:\r\n$ \\int\\int (2x \\plus{} 2) dA$ where that is through the circle of radius $ 3$ centered at the origin. This isn't all that pleasant without a nice polar sub (in fact I can't get the correct answer). Is there a better way?",
"Solution_1": "Green's theorem is completely fine and the resulting integral is not hard at all if you notice that the integral of $ x$ over the circle is $ 0$ due to antisymmetry. So I don't understand what your difficulty is :?. Can you show the details of what you did?",
"Solution_2": "$ 4\\int_0^3\\int_0^{\\sqrt{9\\minus{}y^2}} (2x \\plus{} 2) dxdy\\equal{} 4\\int_0^3 x^2\\plus{}2x\\Big|_0^{\\sqrt{9\\minus{}y^2}} dy\\equal{} 4\\int_0^3 (9\\minus{}y^2\\plus{}2\\sqrt{9\\minus{}y^2})dy\\equal{} 4\\left(9y|_0^3\\minus{}\\frac{1}{3}y^3|_0^3 \\plus{} 2\\int_0^3 \\sqrt{9\\minus{}y^2} dy\\right) \\equal{} 4(27\\minus{}9 \\plus{} 2(\\frac{1}{2}9\\pi))\\equal{}4(18\\plus{}9\\pi)\\neq 18\\pi$",
"Solution_3": "$ 2x\\plus{}2$ is not even with respect to the $ x$ variable, so the integral over the circle is not $ 4$ times the integral over the quarter-circle (it seems like you computed the latter integral correctly but, of course, since it was a wrong integral, you got a wrong value).",
"Solution_4": "Oh, wow that was stupid! I see now."
}
{
"Tag": [
"search",
"\\/closed"
],
"Problem": "If you view this person's profile, it says they have 146 posts. But click \"find all posts by pifreak\" and there are none. Now, how do 146 posts disappear just like that?\r\n\r\n[url]http://www.artofproblemsolving.com/Forum/profile.php?mode=viewprofile&u=515[/url]",
"Solution_1": "maybe all his/her posts come from those online classes?",
"Solution_2": "Did you PM him?\r\n\r\nI agree, either that or they were all in Fun and Games which is not likely since I keep a close eye there. In fact it's probably the earlier because I've never seen his post anywhere.\r\n\r\nP.S. Please excuse me for calling you he if you are a girl.",
"Solution_3": "Right, Syntax - posts in the classes wouldn't show up in a search (unless, perhaps, you are also in that class). pifreak's posts are all in online classes.",
"Solution_4": "pifreak is a she. She's in my class, so if you look at mine message a lot will be missing. Anyways, who's Vanessa Rusczyk? Her name is in the AoPS 2 books for the cover. And Claire and Stanley?",
"Solution_5": "Oh, wow. She must be very disciplined to not have ever posted in F&G...",
"Solution_6": "Vanessa = my wife; Claire = my mother; Stanley = my grandfather."
}
{
"Tag": [
"algebra",
"polynomial",
"Vieta",
"product of roots"
],
"Problem": "A polynomial with roots -4,3, and 24 is of the form $x^3+ax^2-bx+c$. Which of the following statements is true?\r\n\r\nA) $a$ is the sum of the roots\r\nB) $b = -36$\r\nC) $c$ is the product of roots\r\nD) $a$ is the product of the roots\r\nE) None of these",
"Solution_1": "Use Vieta's formulas for third-polynomial degree\r\n\r\n[hide]$e$[/hide]",
"Solution_2": "[hide=\"Answer\"]If the cubic was in the form $x^3-ax^2+bx-c$, then the first three would be true, but since the signs are reversed, the answer is $E$.[/hide]",
"Solution_3": "What's vieta's formulas for the third degree? (I can solve this by multiplying the expression out, but I doubt that's the best way.",
"Solution_4": "[quote=\"236factorial\"]What's vieta's formulas for the third degree? (I can solve this by multiplying the expression out, but I doubt that's the best way.[/quote]It's not hard to see why the formulas are true intuitively, so there's no need to memorize them. The coefficient for $ x^{n-k} $ in the expansion of $ (x-r_1)\\cdots(x-r_n) $ is just the sum of the products of all $ \\binom{n}k $ groups of $ k $ of the $ -r_i $.\r\n\r\nSo the coefficent of $ x^n $ is $ 1 $, the coefficient of $ x^{n-1} $ is $ -(r_1+\\ldots+r_n) $, etc.",
"Solution_5": "Also in simpler terms:\r\n\r\nFor any monic polynomial $ ax^3-bx^2+cx-d $ (monic just means that $ a=1 $), then $ b=r+s+t $, $ c=rs+rt+st $, and $ d=rst $. Note that the signs in the cubic are necessary. If we had, for example, $ ax^3+bx^2+cx+d $, then $ b=-(r+s+t) $ and $ d=-rst $. ($ r $, $ s $, and $ t $ are the roots of the equation, and $ ax^3-bx^2-cx-d $ is the equation $ (x-r)(x-s)(x-t) $ multiplied out.)",
"Solution_6": "[quote=\"Silverfalcon\"]A polynomial with roots -4,3, and 24 is of the form $ x^3+ax^2-bx+c $. Which of the following statements is true?\n\nA) $ a $ is the sum of the roots\nB) $ b = -36 $\nC) $ c $ is the product of roots\nD) $ a $ is the product of the roots\nE) None of these\n[/quote]\r\n\r\n[hide]$ 16a+4b+c=64\\\\ 9a-3b+c=-27\\\\ 576a-24b+c=-1384\\\\\\\\ \\text {Just solve with calculator...}\\\\ \\text {Problem is trivial after this.} $[/hide]\r\n\r\n:D :D",
"Solution_7": "[quote=\"JesusFreak197\"]Also in simpler terms:\n\nFor any monic polynomial $ ax^3-bx^2+cx-d $ (monic just means that $ a=1 $), then $ b=r+s+t $, $ c=rs+rt+st $, and $ d=rst $. Note that the signs in the cubic are necessary. If we had, for example, $ ax^3+bx^2+cx+d $, then $ b=-(r+s+t) $ and $ d=-rst $. ($ r $, $ s $, and $ t $ are the roots of the equation, and $ ax^3-bx^2-cx-d $ is the equation $ (x-r)(x-s)(x-t) $ multiplied out.)[/quote]One more thing to note, if the leading coefficient $ a $ is not $ 1 $, then you can just divide the equation by $ a $ and use what JesusFreak said. This works because you can divide an equation by anything $ \\neq 0 $ and have the same roots.",
"Solution_8": "Right, I should have mentioned that. The general formula would actually be: $ \\frac{b}{a} =r+s+t $, $ \\frac{c}{a} =rs+rt+st $, and $ \\frac{d}{a} =rst $.",
"Solution_9": "[quote=\"236factorial\"]What's vieta's formulas for the third degree? (I can solve this by multiplying the expression out, but I doubt that's the best way.[/quote]\r\n\r\nVieta's forlumas for the third degree states that: \r\n\r\nIf you have polynomial: $ ax^3+bx^2+cx+d=0 $ with 3 real roots say $ x1, x2, x3 $\r\n\r\nThen: $ x1 + x2 + x3 =\\frac{-b}{a} $\r\n $ x1*x2+x2*x3+x3*x1=\\frac{c}{a} $\r\n $ x1*x2*x3 =\\frac{-d}{a} $"
}
{
"Tag": [
"function",
"integration",
"calculus",
"inequalities",
"real analysis",
"calculus computations"
],
"Problem": "Suppose $ f$ and $ g$ are Riemann Integrable Functions on $ [a,b]$ and $ f(x)\\equal{}g(x)$ if $ x$ not equal to $ c$.\r\nShow that: $ \\int ^a_b f\\equal{} \\int ^a_b g$",
"Solution_1": "A rough sketch, using the Darboux theory (upper and lower sums). Find a partition so that the upper and lower sums for the integral of $ f$ are within $ \\epsilon$ of each other. (Maybe $ \\frac{\\epsilon}2$ - that can be worked out later.) Take the common refinement. Refine that further to make the subinterval containing $ c$ extremely narrow. How narrow? We can figure that out once we know the size of $ |f(c)\\minus{}g(c)|$ and the $ \\epsilon$ we want to match. The upper and lower sums for $ f$ and $ g$ are identical except for the one term for the subinterval containing $ c,$ and we can control how big that is.\r\n\r\nThere's a straightforward inductive extension to the case in which $ f$ and $ g$ differ at finitely many points.",
"Solution_2": "We can show $ \\int^{a}_{b}f = \\int^{a}_{b}g$ also directly from Riemman sums.\r\n\r\n[b]Definition:[/b]\r\n$ f$ is Riemann integrable on $ [a,b]$ if and only if for each $ \\epsilon > 0$ there exists $ \\delta > 0$ such that for all partitions $ D$ of $ [a,b]$, where $ \\max{\\delta_k} < \\delta$ $ (\\delta_k$ is the length of $ k$-th interval of $ D \\right)$, and all choices of $ \\xi_k$, we have $ |\\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\int^{b}_{a}f| < \\epsilon$.\r\n\r\n[b]For the problem:[/b]\r\n\r\nWe choose $ \\epsilon > 0$. \r\n\r\nNow we find $ \\delta_f$, for which $ |\\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\int^{b}_{a}f| < \\frac {\\epsilon}{2}$,\r\nwhenever $ \\max{\\delta_k} < \\delta_f$.\r\n\r\nAnaloguously $ \\delta_g$, for which $ |\\sum_{k = 1}^{n}(g(\\xi^{\\prime}_k)\\delta^{\\prime}_k) - \\int^{b}_{a}g| < \\frac {\\epsilon}{2}$,\r\nwhenever $ \\max{\\delta^{\\prime}_k} < \\delta_g$.\r\n\r\n\r\nSince the upper inequalities must be true for all partitions and choices of points, they must be true also for a partition, where $ \\delta_k < \\min{\\delta_f, \\delta_g}$ and where $ \\xi$-s are the same for $ g$ and $ f$ and $ c$ is not among them. Therefore \r\n\r\n$ |\\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\sum_{k = 1}^{n}(g(\\xi_k)\\delta_k) - \\int^{b}_{a}f + \\int^{b}_{a}g | \\leq |\\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\int^{b}_{a}f| + |\\sum_{k = 1}^{n}(g(\\xi_k)\\delta_k) - \\int^{b}_{a}g| < \\epsilon$\r\n\r\nNow: $ \\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\sum_{k = 1}^{n}(g(\\xi_k)\\delta_k) = \\sum_{k = 1}^{n}(f(\\xi_k) - g(\\xi_k))\\delta_k = 0$\r\n\r\nSo $ |\\sum_{k = 1}^{n}(f(\\xi_k)\\delta_k) - \\sum_{k = 1}^{n}(g(\\xi_k)\\delta_k) - \\int^{b}_{a}f + \\int^{b}_{a}g | = |\\int^{b}_{a}f + \\int^{b}_{a}g | < \\epsilon$.\r\n\r\nSince $ \\epsilon$ was arbitrary, we have $ \\int^{b}_{a}f = \\int^{b}_{a}g$.",
"Solution_3": "[quote=\"Kent Merryfield\"]There's a straightforward inductive extension to the case in which $ f$ and $ g$ differ at finitely many points.[/quote]\r\n\r\nUsing my approach, I think, that the result is valid even if $ f$ and $ g$ differ at infinitely many points, just that they do not form any closed interval $ [c,d]\\subset [a,b].$",
"Solution_4": "[quote=\"tmrfea\"]Using my approach, I think, that the result is valid even if $ f$ and $ g$ differ at infinitely many points, just that they do not form any closed interval $ [c,d]\\subset [a,b].$[/quote]\r\nIt's more subtle than that. For example, the result isn't valid if $ f$ and $ g$ differ by a constant on $ \\mathbb{Q}$ - in that case, at least one of $ f$ and $ g$ aren't Riemann integrable. The most general result I know of for the Riemann integral is that $ f$ and $ g$ differ on a compact set of [url=http://mathworld.wolfram.com/MeasureZero.html]measure zero[/url] (and note that this is well-defined even if you haven't defined the Lebesgue measure). Extending this result to all sets of measure zero is one of the motivations for the theory of the Lebesgue integral."
}
{
"Tag": [
"inequalities proposed",
"inequalities"
],
"Problem": "$ a,b,c>0\\ ,\\ a\\plus{}b\\plus{}c\\equal{}1\\Rightarrow 1\\plus{}\\sum_{cyc}\\frac{(a\\plus{}b)(b\\plus{}c)}{c\\plus{}a}\\ge \\sum_{cyc}\\frac{1\\plus{}9ab}{3(1\\minus{}c)}$",
"Solution_1": "[quote=\"alex2008\"]$ a,b,c > 0\\ ,\\ a \\plus{} b \\plus{} c \\equal{} 1\\Rightarrow 1 \\plus{} \\sum_{cyc}\\frac {(a \\plus{} b)(b \\plus{} c)}{c \\plus{} a}\\ge \\sum_{cyc}\\frac {1 \\plus{} 9ab}{3(1 \\minus{} c)}(*)$[/quote]\r\nIT's very easy and not interesting!\r\n$ (*)<\\equal{}> \\sum_{cyc}\\frac{a^2}{b\\plus{}c}\\plus{}a\\plus{}b\\plus{}c \\geq \\sum_{cyc}\\frac{3ab}{a\\plus{}b}$"
}
{
"Tag": [
"calculus",
"integration",
"function",
"logarithms",
"inequalities",
"ratio",
"abstract algebra"
],
"Problem": "Let $ a \\in [0,1)$ and $ \\phi_n(u) \\equal{} \\left( \\frac {u \\minus{} a}{u \\plus{} a} \\right)^n, I_n(a) \\equal{} \\int_0^1 |\\phi_n(u)| \\frac {du}{\\sqrt {1 \\minus{} u^2}}$.\r\n\r\nI'm interested in the asymptotoics(behavious depending on $ n$) of $ \\displaystyle \\inf_{a \\in [0,1)} I_n(a)$.\r\n\r\nDo you have any ideas how to find this asymptotics.",
"Solution_1": "Here is a \"back of envelope\" computation. Split the integral into $ \\int_0^a$ and $ \\int_a^1$. Note that the first one is an increasing function of $ a$ and the second is a decreasing one. So, basically we are interested in the meeting point. Let's make an assumption that they meet with $ na\\gg 1$ and use the standard Laplace asymptotic method for both of them (this is exactly the assumption needed to justify Laplace for the second integral). The first integral in this range is about $ \\frac a{2n}$. The second one is a bit trickier but it finally evaluates to something about $ \\frac C{\\sqrt {an}}\\left(\\frac {1\\minus{}a}{1\\plus{}a}\\right)^n\\approx C\\frac {e^{\\minus{}2an}}{\\sqrt {an}}$, provided that $ a\\ll n^{\\minus{}1/2}$, where $ C$ is some constant whose exact value is of no interest to us unless we want higher order terms in our asymptotic formula. Now, denoting $ t\\equal{}an$, we see that our objective is to minimize $ \\frac{t}{2n^2}\\plus{}\\frac{Ce^{\\minus{}2t}}{\\sqrt t}$. This is a piece of cake because the first term is slow and the second one is fast, so the answer for this minimization problem is $ \\frac {\\log n}{2n^2}$, which is the answer for the original problem as well. It remains to note that the optimal $ a$ is about $ \\frac{\\log n}n$, which is within the assumed range needed to justify our formulae. The end.\r\n\r\nP.S. This argument is extremely sketchy, of course, but if you manage to understand what is done here yourself, next time you'll be able to do such problem without any outside help and in next to no time ;). If you see that it is too hard to recover the details from what I wrote, just let me know and I post the full argument when I have more time :).",
"Solution_2": "Thanks a lot for you answer.\r\nI've got the asymptotics you said for the first integral. But i don't see for the moment how to deal with the second integral, when we also have that at the point $ u\\equal{}1$ where $ \\phi_n(u)$ attains its maximum in $ [a,1]$, the value of $ \\frac{1}{\\sqrt{1\\minus{}u^2}}$ tends to $ \\infty.$",
"Solution_3": "Ah, but Laplace asymptotic formula works for weights with weak power singularities at the point of maximum as well! The idea is the same: if you need the asymptotics for the integral $ \\int_0^\\text{whatever}x^\\nu f(x)^n$ with $ \\nu>\\minus{}1$ and $ f$ attains its only maximum $ M$ at $ 0$ with $ f'(0)\\equal{}\\minus{}L<0$, you can replace $ f(x)$ by $ Me^{\\minus{}(L/M)x}$, expand the integral to $ \\plus{}\\infty$ and get the answer involving the Gamma-function.",
"Solution_4": "Ok, now it seems that i see how to get the asymptotics of the second integral as well.\r\nBut i have a few probably stupid question about the next steps you did :\r\n\r\n[quote=\"fedja\"] The second one is a bit trickier but it finally evaluates to something about $ \\frac C{\\sqrt {an}}\\left(\\frac {1 \\minus{} a}{1 \\plus{} a}\\right)^n\\approx C\\frac {e^{ \\minus{} 2an}}{\\sqrt {an}}$, provided that $ a\\ll n^{ \\minus{} 1/2}$, where $ C$ is some constant whose exact value is of no interest to us unless we want higher order terms in our asymptotic formula. [/quote]\n\nWhere is this $ 2a$ in $ C\\frac {e^{ \\minus{} 2an}}{\\sqrt {an}}$ coming from ? Shouldn't it just be $ C\\frac {e^{ \\minus{}n \\ln{\\frac{1\\minus{}a}{1\\plus{}a} } }}{\\sqrt {an}}$ be ?\n\n[quote=\"fedja\"]\nNow, denoting $ t \\equal{} an$, we see that our objective is to minimize $ \\frac {t}{2n^2} \\plus{} \\frac {Ce^{ \\minus{} 2t}}{\\sqrt t}$. This is a piece of cake because the first term is slow and the second one is fast, so the answer for this minimization problem is $ \\frac {\\log n}{2n^2}$, which is the answer for the original problem as well.[/quote]\r\n\r\nCould you please explain how you got this value $ \\frac {\\log n}{2n^2}$.",
"Solution_5": "[quote=\"eugene\"]\nWhere is this $ 2a$ in $ C\\frac {e^{ - 2an}}{\\sqrt {an}}$ coming from ? Shouldn't it just be $ C\\frac {e^{ - n \\ln{\\frac {1 - a}{1 + a} } }}{\\sqrt {an}}$ ?\n[/quote]\nYou mean $ e^{ n \\ln{\\frac {1 - a}{1 + a} } }$ in the numerator? (minus is an obvious misprint, I hope). Yes, that's correct but $ \\ln{\\frac {1 - a}{1 + a}=-2a+O(a^2)}$ and if $ a$ is not too large, $ O(a^2)$ can be neglected :)\n\n[quote=\"eugene\"][quote=\"fedja\"]\nNow, denoting $ t = an$, we see that our objective is to minimize $ \\frac {t}{2n^2} + \\frac {Ce^{ - 2t}}{\\sqrt t}$.\n[/quote]\nCould you please explain how you got this value $ \\frac {\\log n}{2n^2}$.[/quote]\r\nSure. Take a look at the second term. It goes down exponentially and starts at some constant when $ t=1$. You want it to get comparable to the first term at the point of minimum. The first term grows quite slowly and starts at $ n^{-2}$. So, you want the second term to be also about (this \"about\" is really very lax) $ n^{-2}$. The main factor is, of course, $ e^{-2t}$, which tells you that $ t$ must be about $ \\log n$. Now, once you got a clear idea of what you expect your optimal $ t$ to be, the formal proof is easy. Plugging $ t=\\log n$, you get $ \\frac{\\log n}{2n^2}+\\text{something much less}$. If $ t>(1-\\delta)\\log n$, you get at least $ (1-\\delta)\\frac{\\log n}{2n^2}$ from the first term. If $ t<(1-\\delta)\\log n$, then the second term is at least $ Cn^{-2+2\\delta}/\\sqrt{\\log n}$, which is much greater. So the minimum is, indeed, almost $ \\frac{\\log n}{2n^2}$. Generally speaking, if you have a turtle crawling up the mountain and a hare rushing down, the minimum of the sum of their altitudes (assuming that the hare cannot go below the sea level) is almost the same as the altitude of the turtle at the point of their meeting (actually you have to wait a short time after they meet to let the hare descend to a negligible altitude but the turtle won't be able to go far in that time anyway :).",
"Solution_6": "Thanks a lot, Fedja, for nice explanation.\r\n\r\n[quote=\"fedja\"]\nNow, once you got a clear idea of what you expect your optimal $ t$ to be, the formal proof is easy. Plugging $ t \\equal{} \\log n$, you get $ \\frac {\\log n}{2n^2} \\plus{} \\text{something much less}$. If $ t > (1 \\minus{} \\delta)\\log n$, you get at least $ (1 \\minus{} \\delta)\\frac {\\log n}{2n^2}$ from the first term. If $ t < (1 \\minus{} \\delta)\\log n$, then the second term is at least $ Cn^{ \\minus{} 2 \\plus{} 2\\delta}/\\sqrt {\\log n}$, which is much greater. So the minimum is, indeed, almost $ \\frac {\\log n}{2n^2}$. [/quote]\r\n\r\nYou probably meant here if $ t > (1 \\plus{} \\delta)\\log n$.",
"Solution_7": "Nope, I meant exactly what I wrote (it is $ e^{\\minus{}2t}$, remember?).",
"Solution_8": "I meant $ (1 \\plus{} \\delta)$ in the first inequality: you took $ t \\equal{} \\log n$ and then want to consider values $ t > (1 \\plus{} \\delta)\\log n$ and $ t < (1 \\minus{} \\delta)\\log n$ and show that for both of these cases we get a bigger value.\r\nBut, anyway, it doesn't matter much.\r\n\r\nBtw, would it be possible to solve the following generalization: \r\n\r\nLet $ a_i \\in [0,1)$ and $ \\phi_{n_i}(u) \\equal{} \\left( \\frac {u \\minus{} a_i}{u \\plus{} a_i} \\right)^n, i \\equal{} 1,2,\\ldots ,k$ , $ a \\equal{} (a_1,a_2, \\ldots, a_k)$.\r\n$ I_n(a) \\equal{} \\int_0^1 |\\phi_{n_1}(u)| |\\phi_{n_2}(u)| \\ldots |\\phi_{n_k}(u)| \\frac {du}{\\sqrt {1 \\minus{} u^2}}$.\r\n\r\nI'm interested in the asymptotoics(behavious depending on $ n$) of $ \\displaystyle \\inf_{a_i \\in [0,1)} I_n(a)$.\r\n\r\nDo you have any ideas how to find this asymptotics.",
"Solution_9": "Uh-oh! The part near $ 1$ gives little trouble: under very mild assumptions, it is equivalent to $ C\\frac{e^{\\minus{}2n\\sum a_j}}{\\sqrt {n\\sum a_j}}$, but the one near $ 0$ is a big headache. I'll see what I can do but I do not promise anything fast: multivariable optimization is always much harder than one variable one, you know... :) All I can say right away is that I expect a power law (with, perhaps, an occasional $ \\log$ factor) with power growing linearly in $ k$. Do you need the exact asymptotics or just the order of magnitude?\r\n\r\nBy the way, do I understand right that $ n$ is the same in all $ \\varphi$? (Your notation $ \\varphi_{n_i}$ is a bit confusing because it makes one think that there are different $ n_i$ but the right hand side in the formula you wrote has no such things) :?",
"Solution_10": "OK, it took me some time to figure out what is going on here and by now I believe that the answer is \r\n\\[ \\frac{k\\cdot (k\\minus{}1)!^2\\cdot (\\log n)^{2k\\minus{}1}}{2\\cdot 4^{k\\minus{}1}\\cdot n^{2k}},\\]\r\nwhich more or less agrees with my intuition. The argument that led me to it is still quite shaky. I'll post it when I have more time unless you have a solid reason to believe that the answer cannot possibly be this one. By the way, where did such fancy minimization problem come from?",
"Solution_11": "The problem comes from asymptotics of some another expression.\r\nI was trying to find the asymptotic behaviour of it, and on the way run into the one posted here.\r\n\r\nI knew before that with $ k$ numbers $ a_i$ the answer should asymptotically be $ \\frac{(\\log n)^{2k\\minus{}1}}{n^{2k}}$ for other reasons for the initial expression, so i would say that i have solid reason that the answer is the one you mentioned :). (i didn't thought about the constant though).\r\n\r\nI would be very grateful if you could post you ideas how you came to this answer, when you have time.",
"Solution_12": "[color=red]I edited a bit because trying to correct one mistake I made, I forgot to correct another one, so the initial answer turned out to be correct after all but it was only because I originally made an even number of mistakes. Now everything should be fine :)[/color]\r\n\r\nHere is a brief ketch of what I did (I hope that it makes sense but there are a couple of things I didn't try to justify). Assume that $ a_1\\ge a_2\\ge\\dots\\ge a_k$. First of all, let's consider the case when $ k \\equal{} 2$. If $ a_1$ is comparable to $ a_2$, it can easily be seen that you do not gain much compared with the case $ a_1 \\equal{} a_2$, i.e., the order of magnitude is still about $ n^{ \\minus{} 2}$, so to get less, one needs to consider $ a_1\\gg a_2$. Then the contribution of the part near $ 1$ is about $ \\frac {e^{ \\minus{} 2a_1n}}{\\sqrt {a_1n}}$ (with very lax \"about\", but, since this is a fast term, we can affors to be very sloppy in its estimate). \r\n\r\nThe term near $ 0$ splits into $ \\int_0^{a_2} \\plus{} \\int_{a_2}^{a_1}$. The first integral is about $ \\frac {a_2}{2n}$ (same as in the one-variable version) and the whole game is to estimate the second integral. Now, the function $ \\frac {x \\minus{} a_2}{x \\plus{} a_2}\\frac {a_1 \\minus{} x}{a_1 \\plus{} x}$ attains its maximum for $ a_2\\ll x\\ll a_1$, which (with certain reservations) allows to use exponential approximation $ \\exp\\left\\{ \\minus{} 2\\frac {a_2}x \\minus{} 2\\frac {x}{a_1}\\right\\}$ and conclude that the maximum is about $ e^{ \\minus{} 4\\sqrt {\\frac {a_2}{a_1}}}$ and it is attained somewhere near $ x \\equal{} \\sqrt {a_1a_2}$. Now, using Laplace asymptotic formula (again, with some reservations: the ratio $ a_2/a_1$ should be neither too small nor too large for this step), we see that this second integral is about $ a_1\\sqrt [4]{\\frac {a_2}{a_1}}n^{\\minus{}1/2}e^{ \\minus{} 4n\\sqrt {\\frac {a_2}{a_1}}}$. \r\n\r\nIf we believe in the approximate expression for this term, the sum is not hard to optimize (the same old hare and turtle story). We see that one needs to make $ a_1$ of order $ \\log n/n$ and $ a_2/a_1$ of order $ (\\log n/n)^2$ (I'll not try to figure out the constants for this particular case; we'll do it for the general case below). Now comes a big leap of faith: we shall assume that in general we have pretty much the same situation: the minimum is attained if $ a_1\\gg a_2\\gg\\dots\\gg a_k$ and, moreover, the ratios $ a_{j}/a_{j \\minus{} 1}$ are about $ (\\log n/n)^2$. How do we know that? There is a rational way to come to this conclusion but it is neither short, not completely rigorous, so we'll better see what follows from here and try to justify everything a posteriori. \r\n\r\nThen $ a_j \\equal{} C_j(\\log n/n)^{2j \\minus{} 1}$ and the final answer should be $ \\frac {C_k(\\log n)^{2k \\minus{} 1}}{2n^{2k}}$. This is our \"turtle\". All \"hares\" (integrals from $ a_{j}$ to $ a_{j \\minus{} 1}$; with such fast decay of $ a_j$ only 2 factors will really matter in each of them and we shall get the same approximation formula for each of them as in the $ k \\equal{} 2$ case) should be below this level. This allows to find the relations for the constants: $ C_1\\approx k$, $ 4\\sqrt {\\frac {C_j}{C_{j \\minus{} 1}}}\\approx 2k \\minus{} (2j \\minus{} 3) \\minus{} 1$ leading to the answer\r\n\\[ C_k \\equal{} k\\cdot 4^{ \\minus{} (k \\minus{} 1)}\\cdot[(k\\minus{}1)!]^2\\,.\r\n\\]\r\nNow, how to justify everything. The first thing to do is to check the validity of the Laplace approximation for the second integral in the $ k \\equal{} 2$ case when $ a_2/a_1\\asymp (\\log n/n)^2$. To be honest, I haven't done it but it should be a routine thing and even if it breaks (it should break really badly to change the answer because it is a hare estimate; for instance, if it turns out that you have an extra $ \\log n$ factor in the right approximation formula, it won't move the final answer by an inch; only a missing power of $ n$ has enough strength to change the constant in the final asymptotics), it is clear what correction should be made in the argument. After that, you need to show that if $ a_{j}/a_{j \\minus{} 1}\\ll (\\log n/n)^2$, then the integral from $ a_j$ to $ a_{j \\minus{} 1}$ is much greater than, say, $ a_{j \\minus{} 1}/n$, so we shall effectively reduce the number of variables in this case. Finally, once you know that each $ a_j/a_{j \\minus{} 1} > c(\\log n/n)^2$, you can say that to beat the assumed answer, you'll need to also have $ a_{j}/a_{j \\minus{} 1} < C(\\log n/n)^2$, which brings you right into the range where all the above estimates were done. And in that range we made an optimal choice.\r\n\r\nUph-ph-ph. That was a long story! Hope it helps a bit :) Feel free to ask as many questions as you need to understand everything.",
"Solution_13": "Thank you very much for you answer.\r\n\r\n[quote=\"fedja\"]\n\nThen $ a_j \\equal{} C_j(\\log n/n)^{2j \\minus{} 1}$ and the final answer should be $ \\frac {C_k(\\log n)^{2k \\minus{} 1}}{2n^{2k}}$. This is our \"turtle\". All \"hares\" (integrals from $ a_{j}$ to $ a_{j \\minus{} 1}$; with such fast decay of $ a_j$ only 2 factors will really matter in each of them and we shall get the same approximation formula for each of them as in the $ k \\equal{} 2$ case) should be below this level. This allows to find the relations for the constants: $ C_1\\approx k$, $ 4\\sqrt {\\frac {C_j}{C_{j \\minus{} 1}}}\\approx 2k \\minus{} (2j \\minus{} 3) \\minus{} 1$ leading to the answer\n\\[ C_k \\equal{} k\\cdot 4^{ \\minus{} (k \\minus{} 1)}\\cdot[(k \\minus{} 1)!]^2\\,.\n\\]\n[/quote]\r\n\r\nI'll write my understanding how you find the constants $ C_j$ just to be sure i understood it right.\r\nIn the whole expression $ \\int_{0}^{a_n} \\plus{} \\int_{a_n}^{a_{n \\minus{} 1}} \\plus{} \\ldots \\int_{a_1}^{1}$, the last integral has asymptotics \r\n$ \\frac {e^{ \\minus{} 2a_1n}}{\\sqrt {a_1n}}$, and if $ a_1 \\equal{} C_1 \\frac {\\log n}{n}$, then this is approximately $ \\frac {1}{n^{2C_1} \\sqrt {C_1 \\log n}}$, from the other side the main asymptotic",
"Solution_14": "Seems like your connection got broken when you were posting :? I can see that you, most likely, understood everything correctly but, to confirm it with certainty, I would prefer to see the entire text you intended to post :)",
"Solution_15": "Sorry, indeed i didn't finish my post. I've made a correction above by now.\r\n\r\n* Strange, but it disappeared once again. I'll try to post it separately here:\r\n\r\nI'll write my understanding how you find the constants $ C_j$ just to be sure i understood it right.\r\nIn the whole expression $ \\int_{0}^{a_n} \\plus{} \\int_{a_n}^{a_{n \\minus{} 1}} \\plus{} \\ldots \\int_{a_1}^{1}$,\r\nthe last integral has asymptotics $ \\frac {e^{ \\minus{} 2a_1n}}{\\sqrt {a_1n}}$, and if $ a_1 \\equal{} C_1 \\frac {\\log n}{n}$,\r\nthen this is approximately $ \\frac {1}{n^{2C_1} \\sqrt {C_1 \\log n}}$, from the other side the \r\nmain asymptotic term $ \\frac {a_k}{2k} \\asymp \\frac {1}{n^{2k}}$ , taking into account that\r\n$ a_k \\asymp C_k \\left( \\frac {log n}{n} \\right)^{2k \\minus{} 1}$.\r\n\r\nThen requiring that the asymptotics of $ \\int_{a_{j}}^{a_{j \\minus{} 1}}$, which is \r\n$ a_{j \\minus{} 1}\\sqrt [4]{\\frac {a_j}{a_{j \\minus{} 1}}}n^{ \\minus{} 1/2}e^{ \\minus{} 4n\\sqrt {\\frac {a_j}{a_{j \\minus{} 1}}}}$ \r\nmust be $ \\frac {1}{n^{2k}}$ we find the ratios $ \\frac {C_j}{C_{j \\minus{} 1}}$.",
"Solution_16": "Yes, that's it. :lol:",
"Solution_17": "Sorry for reviving an old thread for probably stupid reason.\r\n\r\nSuppose i initially wanted to find the asymptotics of $ \\displaystyle \\inf_{a \\in [0,1)} I_n(a)$, where\r\n$ I_n(a) = \\int_0^1 \\frac{|\\phi_n(u)|}{1+\\phi_n(u)^2} \\frac {du}{\\sqrt {1 - u^2}}$ and\r\n$ \\phi_n(u)$ is still the same function $ \\phi_n(u)= \\left( \\frac {u - a}{u + a} \\right)^n$\r\n\r\nWell, $ \\phi_n(u)$ definitely tends to $ 0$ at any point $ 0\\log n$ then $ |I_n(x,a)| \\asymp \\frac{a}{2n} \\minus{}x^2C\\frac{\\exp(\\minus{}2an)}{\\sqrt{an}}$ and the supremum $ \\sup_{x \\in [\\minus{}1,1]} | I_n(x,a) |$ is attained at $ x\\equal{}0$ hence the infumum in this case is greater than $ \\frac{\\log n}{2n^2}$. \r\nIf $ a < \\log n$ then $ |I_n(x,a)| \\asymp x^2C\\frac{\\exp(\\minus{}2an)}{\\sqrt{an}} \\minus{}\\frac{a}{2n}$ the supremum $ \\sup_{x \\in [\\minus{}1,1]} | I_n(x,a) |$ is attained at $ x\\equal{}1$ hence the infumum in this case is again greater than $ \\frac{\\log n}{2n^2}$.",
"Solution_23": "I'm terribly sorry for making you wait so long :blush:. Here is the (sketch of an) argument I had in mind.\r\n\r\nFirst of all, the factor $ \\frac{x^2}{x^2\\plus{}(1\\minus{}x^2)u^2}$, $ x\\in[\\minus{}1,1]$ can be written as $ \\frac 1{1\\plus{}Mu^2}$, $ M\\equal{}x^{\\minus{}2}\\minus{}1\\in[0,\\plus{}\\infty]$. \r\n\r\nSecondly, the contribution of the intervals $ [a_i/\\sqrt n,a_i\\sqrt n]$ can be estimated by $ e^{\\minus{}c\\sqrt n}$ (the factor, corresponding to $ a_n$ is already small enough).\r\n\r\nLet now $ J_j\\equal{}[a_j\\sqrt n, a_{j\\plus{}1}/\\sqrt n]$ with the convention that the first integral starts at $ 0$, the last one ends at $ 1$, and that $ J_j\\equal{}\\varnothing$ if the left end exceeds the right one. The total integral is then $ \\sum_j \\int_{J_j}\\plus{}O(e^{\\minus{}c\\sqrt n})$.\r\n\r\nThe next observation is that, for each $ j$, we can choose $ M$ such that the factor $ \\frac 1{1\\plus{}Mu^2}$ is close to $ 1$ on $ J_k$ with $ k\\le j$ and to $ 0$ on $ J_k$ with $ k>j$. Since the multiplicative gap between the intervals $ J_k$ is of order $ n$, this is possible.\r\n\r\nThus, for each $ j$, we can choose $ M$ such that the whole integral is\r\n\\[ \\Bigl|\\sum_0^j I_j\\Bigr| \\pm \\Bigl[\\delta\\sum_k|I_k|\\plus{}e^{\\minus{}c\\sqrt n}\\Bigr]\\]\r\nwith arbitrarily small $ \\delta>0$ where $ I_j$ is the integral over $ J_j$ without the factor $ \\frac 1{1\\plus{}Mu^2}$.\r\n\r\nNow the key observation is that, if \r\n\\[ \\Bigl|\\sum_0^j I_j\\Bigr|\\le A\\plus{}\\delta\\sum_k|I_k|,\\]\r\nthen \r\n$ |I_0|\\le C(\\delta) A$, $ |I_j|\\le 2C(\\delta)A$ with $ C(\\delta)\\to 1$ as $ \\delta\\to 0\\plus{}$. Indeed, we have $ \\Bigl|I_j\\Bigr|\\le 2A\\plus{}2\\delta\\sum_k|I_k|$ for all $ j\\ge 1$. Adding these inequalities with the inequality for $ I_0$, we obtain $ \\sum_j|I_j\\Bigr|\\le 2AK\\plus{}2K\\delta\\sum_k|I_k|$ where $ K$ is the (fixed) number of intervals, which allows us to estimate the sum of $ |I_j|$ by $ \\frac{2K}{1\\minus{}2K\\delta}A$. The rest should be clear.\r\n\r\nThe final step is to notice that, if $ A$ is less than the answer to the original problem, then the inequalities $ |I_j|\\le 2A$ for $ j\\ge 1$ imply the same estimates for the ratios $ a_{j}/a_{j\\plus{}1}$ as before (they were at the level of power of $ n$, not of the constant in front of it), and, thereby, the estimate $ |I_0|>A$, which gives the desired contradiction. \r\n\r\nAs usual, feel free to ask questions and, please, accept my apologies for the delay.",
"Solution_24": "Thanks a lot for your answer, Fedja. No need for apologies :) \r\n\r\nI think i understand the idea and the structure except the last paragraph with estimates:\r\n\r\n[quote=\"fedja\"]\nThe final step is to notice that, if $ A$ is less than the answer to the original problem, then the inequalities $ |I_j|\\le 2A$ for $ j\\ge 1$ imply the same estimates for the ratios $ a_{j}/a_{j \\plus{} 1}$ as before (they were at the level of power of $ n$, not of the constant in front of it), and, thereby, the estimate $ |I_0| > A$, which gives the desired contradiction. \n[/quote]\r\n\r\nAs far as i understand the $ |I_j| \\asymp a_j\\sqrt [4]{\\frac {a_j}{a_{j \\plus{} 1}}\\cdot n}\\exp\\left( \\minus{} 4n\\sqrt {\\frac {a_j}{a_{j \\plus{} 1}}\\cdot n}\\right)$ (*) Now if $ A \\equal{} A_0 \\equal{} C \\frac {\\log^{2k \\minus{} 1} n}{n^{2k}}$, the constant obtained in the previous case, then requiring that $ |I_j|\\leq 2A$ for all $ j \\geq 1$ basically gives an estimate for $ a_j/a_{j \\plus{} 1}$ of the following type:\r\n$ a_j/a_{j \\plus{} 1} \\geq C_j \\frac {\\log^2n}{n^3}$ (unfortunately with $ n^3$ in the denominator because of the multiplier $ n$ in the square root in (*)) for appropriate constants $ C_j$ similarly to the previous case.\r\nMultiplying such inequalities we get an estimate for $ a_0$, somewhat similar to \r\n$ a_0 \\geq C_0 \\frac {\\log^{2k}}{n^{3k}}$ (given that we also know the estimate from below for $ a_k$). But $ |I_0| \\asymp \\frac {a_0}{n\\sqrt {n}}$ which as far as i understand doesn't give immediately that $ |I_0| > A$ ($ A$ is of order $ \\frac {1}{n^{2k}}$) ($ k$ is the number of numbers $ a_1,a_2,\\ldots,a_k$).\r\n\r\nYou probably meant some another way to get better estimates or i am just wrong somewhere.",
"Solution_25": "Where did you get that extra $ n$ under the square root in the exponent from? :? The approximation for the integral from $ a_j$ to $ a_{j\\plus{}1}$ is [b]exactly[/b] the same as before (up to $ \\pm$, of course). All that I said was that the integrals from $ a_j$ to $ a_j\\sqrt n$ and from $ a_{j\\plus{}1}/\\sqrt n$ to $ a_{j\\plus{}1}$ can be safely ignored, not that we need to replace $ a_j$ by $ a_{j}\\sqrt n$ in the previous estimate for one hump or anything like that. So, there is no $ n^3$, just the same good old $ n^2$.",
"Solution_26": "Thanks for the answer, Fedja. Now i see what you mean.\r\nBelow i write my understanding of your proof. Sorry that it is rather long. If you happen to have time,please tell me whether i got your point right. Thanks :)\r\n\r\n\r\nIn the beginning fix the numbers $ a_1,a_2,\\ldots,a_q$. Just to be consistent with the previous case i let's consider the opposite ordering of $ a_i$'s, i.e. assume $ a_q < \\ldots < a_2 < a_1$. The integrals\r\n$ I_j \\equal{} \\int_{a_j}^{a_j\\sqrt n} \\plus{} \\int_{a_j\\sqrt n}^{a_{j \\plus{} 1}/\\sqrt n} \\plus{} \\int_{a_{j \\plus{} 1}/ \\sqrt n}^{a_{j \\plus{} 1}} \\equal{} O(e^{ \\minus{} c\\sqrt n}) \\plus{} \\int_{J_j}$ with $ J_j \\equal{} [a_j \\sqrt n,a_{j \\plus{} 1}/ \\sqrt n]$. Therefore the whole integral $ I \\equal{} \\sum_j \\int_{J_j} \\plus{} O(e^{ \\minus{} c\\sqrt n})$.\r\n\r\nLet's fix $ \\epsilon > 0$ and find such $ M$ that $ 1 \\minus{} \\frac {1}{1 \\plus{} Mu^2} < \\epsilon$ on $ J_k$ with $ k \\leq j$ and $ \\frac {1}{1 \\plus{} Mu^2} < \\epsilon$ on $ J_k$ with $ k > j$ (*). Indeed due to\r\nthe inequality $ \\frac {1 \\minus{} \\epsilon}{\\epsilon a^2_{j \\plus{} 1}n} < \\frac {\\epsilon n}{(1 \\minus{} \\epsilon) a^2_{j \\plus{} 1}}$ (which is true say for $ \\epsilon \\equal{} 1/n$) one has that there exists $ M$ with\r\n$ \\frac {1 \\minus{} \\epsilon}{\\epsilon a^2_{j \\plus{} 1}n} < M < \\frac {\\epsilon n}{(1 \\minus{} \\epsilon) a^2_{j \\plus{} 1}}$ and for such $ M$ one indeed has (*).\r\n\r\nThus, for each $ j$, we can choose $ M$ such that\r\n\\[ |I| \\equal{} \\Bigl|\\sum_0^j I_i\\Bigr| \\pm \\Bigl[\\delta\\sum_{k \\equal{} 1}^K |I_k| \\plus{} e^{ \\minus{} c\\sqrt n}\\Bigr]\r\n\\]\r\nLet's prove that there exists $ j$ such that\r\n\\[ \\Bigl|\\sum_0^j I_i\\Bigr| \\geq A_0 \\plus{} \\delta\\sum_{k \\equal{} 1}^K |I_k| \\quad \\quad \\quad (1)\r\n\\]\r\nwhere $ A_0$ is the asymptotic with constant obtained in the previous case. \r\n\r\nIndeed suppose that on the contrary for all $ j$:\r\n\\[ \\Bigl|\\sum_0^j I_i\\Bigr| < A_0 \\plus{} \\delta\\sum_{k \\equal{} 1}^K |I_k|\r\n\\]\r\nThen for $ j \\equal{} 0$, $ |I_0|\\leq C(\\delta)A_0$. For $ j \\geq 1$\r\n\\[ \\Bigr|I_j \\Bigr| \\minus{} \\Bigr| \\sum_{i \\equal{} 0}^{j \\minus{} 1} I_i \\Bigr| \\leq A_0 \\plus{} \\delta\\sum_{k \\equal{} 1}^K |I_k|.\r\n\\]\r\nHence the estimate $ \\Bigl|I_j\\Bigr|\\le 2A_0 \\plus{} 2\\delta\\sum_k|I_k|$. Summing such inequalities for all $ j$ obtain $ \\sum_{k \\equal{} 1}^K |I_k| \\leq \\frac {2K}{1 \\minus{} 2K\\delta}A_0$ and plugging into the estimate for $ |I_j|$ we get that $ |I_j|\\le 2C(\\delta)A_0$.\r\n\r\nNow since we know that $ |I_j| \\asymp a_{j \\minus{} 1}\\sqrt [4]{\\frac {a_{j \\minus{} 1}}{a_j} } \\exp \\left( \\minus{} 4n\\sqrt {\\frac {a_{j \\minus{} 1}}{a_j}} \\right)$ (we can forget about the summand $ O(e^{ \\minus{} c\\sqrt n})$ which is $ o(A_0)$ we can estimate $ a_{j \\minus{} 1}/a_j$. First we note that there exist constants $ B_j$ such that $ a_{j \\minus{} 1}/a_j \\geq B_j \\frac {\\log^2 n}{n^2}$ therefore $ a_j \\geq C_j \\left(\\frac {\\log n}{n} \\right)^{2j}$ for some constants $ C_j$.\r\n\r\nIf $ a_{j \\minus{} 1}/a_j \\equal{} (2q \\minus{} 2j \\plus{} 2 \\minus{} \\tau)^2/16 \\frac {\\log^2 n}{n^2}$ for some positive $ \\tau$ then we obtain the contradiction. Therefore for all $ j$ : $ a_{j \\minus{} 1}/a_j \\geq (2q \\minus{} 2j \\plus{} 2)^2/16 \\frac {\\log^2 n}{n^2}$ and $ a_0 \\geq D_0$ where $ D_0 \\equal{} \\frac {q\\cdot (q \\minus{} 1)!^2\\cdot (\\log n)^{2q \\minus{} 1}}{2\\cdot 4^{q \\minus{} 1}\\cdot n^{2q}}$ is the asymptotic in $ A_0$ obtained in the previous case and $ |I_0|\\geq A_0$. \r\n\r\nThis contradiction proves (1) which implies that for $ \\sup_M |I| \\geq A_0$ for any set $ a \\equal{} (a_1,a_2,\\ldots,a_q)$ and hence $ \\inf_a \\sup_M |I| \\geq A_0$.",
"Solution_27": "I glanced at it briefly and nothing stroke my eye as \"obviously wrong\" except that the ordering of $ a_j$ should be direct here: $ a_1<\\dots0,\\sum_j p_j\\equal{}1$, then equal $ p_j$ give you the optimal choice for the asymptotic constant in the end (you should be able to easily do that yourself by now, just go over the proof and introduce $ p_j$ everywhere; you'll end up getting their product squared in the denominator in the final asymptotic expression).",
"Solution_30": "Thank you, Fedja, for your reply. Yes, indeed I see the way you meant via recalculating all the expressions introducing $ p_j$ and having their product squared in $ C_k \\equal{} \\frac{k [(k\\minus{}1)!]^2 }{2^{2k\\minus{}2} (p_1p_2 \\ldots p_k)^2}$. If we want to minimize this expression we must have all $ p_j$'s equal. \r\nAs you corrrectly assumed i was indeed hoping for some a bit different reason, as you said \"slick argument\". So, if you happen to have one some time in future when you have time, i would be glad if you posted it.\r\nThanks again :)"
}
{
"Tag": [
"advanced fields",
"advanced fields unsolved"
],
"Problem": "Let $ \\langle\\cdot,\\cdot\\rangle$ be the Lorentz metric on $ \\mathbb R^{n\\plus{}1}$ defined by $ \\langle x,y\\rangle\\equal{}\\minus{}x_0y_0\\plus{}x_1y_1\\plus{}\\ldots\\plus{}x_ny_n$, where $ x\\equal{}(x_0,\\ldots,x_n)$ and $ y\\equal{}(y_0,\\ldots,y_n)$.\r\nLet $ H^n\\equal{}\\{x\\in\\mathbb R^{n\\plus{}1}\\mid\\langle x,y\\rangle\\equal{}\\minus{}1,x_0>0\\}$.\r\nProve that $ \\langle\\cdot,\\cdot\\rangle$ induces a complete Riemannian metric on $ H^n$.\r\n\r\nTo show that $ \\langle\\cdot,\\cdot\\rangle$ induces a Riemannian metric on $ H^n$, do I need to show that, given a fixed $ x\\in H^n$, for $ v\\in T_xH^n$, $ \\langle v,v\\rangle>0$?\r\n\r\nIf so, how do I show that $ v_0^2d-2x_0-(n+1)\\frac{u}{q} > c$. Then : Let $ z\\in(c,d-2x_0-(n+1)\\frac{u}{q})$\r\n\r\n$ g(x_0+(x_0+(n+1)\\frac{u}{q})+z)=g(x_0)+g(x_0+(n+1)\\frac{u}{q})+g(z)$\r\n$ g((x_0+\\frac{u}{q})+(x_0+n\\frac{u}{q})+z)=g(x_0+\\frac{u}{q})+g(x_0+n\\frac{u}{q})+g(z)$\r\n\r\nAnd so $ g(x_0+(n+1)\\frac{u}{q})=g(x_0+n\\frac{u}{q})+(g(x_0+\\frac{u}{q})-g(x_0))$\r\n\r\nAnd so $ g(x_0+k\\frac{u}{q})=g(x_0)+k(g(x_0+\\frac{u}{q})-g(x_0))$ $ \\forall k\\in\\mathbb N_0$ such that $ x_0+k\\frac{u}{q} < d-c-x_0$\r\n\r\nSo :\r\n$ k=p$ $ \\implies$ $ g(x_0+p\\frac{u}{q})=g(x_0)+p(g(x_0+\\frac{u}{q})-g(x_0))$\r\n$ k=q$ $ \\implies$ $ g(x_0+q\\frac{u}{q})=g(x_0)+q(g(x_0+\\frac{u}{q})-g(x_0))$\r\n\r\n$ \\implies$ $ g(x_0+\\frac{p}{q}u)=g(x_0)+\\frac{p}{q}(g(x_0+u)-g(x_0))$\r\n\r\n$ \\implies$ $ g(x_0+r\\cdot u)=g(x_0)+r(g(x_0+u)-g(x_0))$ $ \\forall r\\in\\mathbb Q^+$ such that $ x_0+r\\cdot u3c$, we have $ c<\\frac{d}{3} 3c$\r\nIn this case, the two intervals $ (c,d-2c)$ and $ (3c,d)$ have a common part and so $ \\beta=0$ and $ g(x)=\\alpha x$ $ \\forall x\\in (c,d)$\r\nAnd these conditions are obviously enough.\r\n\r\n2) $ d-2c < 3c$\r\nIn this case, the two intervals $ (c,d-2c)$ and $ (3c,d)$ have no common part and so :\r\n$ g(x)=\\alpha x + \\beta \\forall x\\in (c,d-2c)$\r\n$ g(x)=\\alpha x + 3\\beta \\forall x\\in (3c,d)$\r\n$ g(x)$ may have any value for $ x\\in[d-2c,3c]$ (in respect with continuity)\r\nAnd these conditions are enough since $ x+y+z\\in(c,d)$ $ \\implies$ $ x,y,z\\in(c,d-2c)$\r\n\r\n3) $ d-2c=3c$\r\nWe are in the same situation than in point 2 but the continuity at $ d-2c=3c$ implies $ \\beta=0$ and so we are in the case 1).\r\n\r\nFinal solution :\r\n==============\r\n1) If $ 1 < a < a^3 < a^5\\leq b$ : $ f(x)=\\alpha\\ln(x)$ ($ \\alpha\\in\\mathbb R$)\r\n\r\n2) If $ 1 < a < a^3 < b < a^5$ : Let $ \\alpha,\\beta\\in\\mathbb R$\r\n\r\n$ \\forall x\\in(a,\\frac{b}{a^2})$ : $ f(x)=\\alpha\\ln(x)+\\beta$\r\n\r\n$ \\forall x\\in(a^3,b)$ : $ f(x)=\\alpha\\ln(x)+3\\beta$\r\n\r\n$ \\forall x\\in[\\frac{b}{a^2},a^3]$ : $ f(x)$ may have any value in respect with continuity"
}
{
"Tag": [
"USAMTS",
"probability"
],
"Problem": "Solutions to the Round 4 problems are now posted on the [url=http://www.usamts.org]USAMTS website[/url].",
"Solution_1": "Typo in the solution for #1:\r\n\r\n$ 11^4 < 16072 < 12^2$\r\n\r\nIt should say\r\n\r\n$ 11^4 < 16072 < 12^4$",
"Solution_2": "Thanks. The typo has been corrected.",
"Solution_3": "Another typo:\r\n\r\n\"If the particle follows the blue arrow ending at $ (3, 4)$, then the probability of winning from there is $ \\frac{1}{2}$, by symmetry.\"\r\n\r\nSeems it should be $ (4, 3)$.",
"Solution_4": "When will we know our Round 4 scores/feedback and prize choices?"
}
{
"Tag": [
"geometry",
"geometric transformation",
"homothety",
"parallelogram",
"AwesomeMath",
"summer program",
"ratio"
],
"Problem": "help me out!!!",
"Solution_1": "I think the midpoints suggest a homothety, or whatever.\r\nBut this is definitely not HSB!",
"Solution_2": "how are we supposed to do the problem if the only description you give is the ppicture. its not even high resolution",
"Solution_3": "ya its a real bad diagram too I cannot tell which sides are supposed to be equal in length :o",
"Solution_4": "Well the diagram is pretty clear if you zoom in, but maybe this should be moved to the Pre-Olympiad forum.",
"Solution_5": "so its three different sized equilateral triangles joined at their vertices to form a scalene triangle in the center..\r\n\r\nhm",
"Solution_6": "[quote=\"rnwang2\"]I think the midpoints suggest a homothety, or whatever.\nBut this is definitely not HSB![/quote]\r\nAha... at least you remembered that... now if you find the right centre of homothety the problem is solved almost immediately.",
"Solution_7": "[quote=\"probability1.01\"]Aha... at least you remembered that... now if you find the right centre of homothety the problem is solved almost immediately.[/quote]\r\nSo you're supposed to draw in the parallelograms, right? I don't see the center...:help:\r\n\r\nNB: I have 666 posts.",
"Solution_8": "hmmm...this problem is very similar to the number 10 on the awesomemath test B. the only difference is that t was the centers of the triangles. ill move it",
"Solution_9": "please transfer it to the its appropriate forum, if needed.",
"Solution_10": "Let AR be a median of the original triangle. Consider a homothety (with ratio 2:1) about A taking P to D, Q to E, and R to F. Prove DEF is equilateral with congruent triangles.",
"Solution_11": "can some1 please explain what is homothety? I EARLIER HEARD IT FROM MOONMATH.... :blush:"
}
{
"Tag": [
"AMC",
"AIME"
],
"Problem": "This is Problem 8 on the 1993 AIME:\r\nLet $ S\\,$ be a set with six elements. In how many different ways can one select two not necessarily distinct subsets of $ S\\,$ so that the union of the two subsets is $ S\\,$? The order of selection does not matter; for example, the pair of subsets $ \\{a, c\\}\\,$, $ \\{b, c, d, e, f\\}\\,$ represents the same selection as the pair $ \\{b, c, d, e, f\\}\\,$, $ \\{a, c\\}\\,$.\r\n\r\nThe solution in the wiki was:\r\nCall the two subsets $ m$ and $ n$. For each of the elements in $ S$, we can assign it to either $ m$, $ n$, or both. This gives us $ 3^6$ possible methods of selection. However, because the order of the subsets does not matter, each possible selection is double counted, except the case where both $ m$ and $ n$ contain all $ 6$ elements of $ s$. So our final answer is then $ \\frac {3^6 \\minus{} 1}{2} \\plus{} 1 \\equal{} \\boxed{365}$\r\n\r\nThen:\r\nI tried to solve it, and I found that we first partition the 6 elements into two sets, and there should be $ 2^6/2\\equal{}32$ ways to do so. Then, for each element, we can choose if we want to duplicate it by also putting it in the other subset, and there are $ 2^6\\minus{}1\\equal{}63$ ways to do so. My final answer was then $ 32 \\cdot 63\\equal{}2016$. Obviously, this is the wrong answer, but where did I go wrong?",
"Solution_1": "The duplication step overcounts. Putting an element in set $ A$ and choosing to duplicate it is the same as putting an element in set $ B$ and choosing to duplicate it, which is why the official solution does the duplication along with the assignment.",
"Solution_2": "OK. Thanks. Can you explain how much it overcounted by? It overcounted by a large amount.",
"Solution_3": "You're basically doing $ 4^6$ instead of the $ 3^6$ in the solution you gave."
}
{
"Tag": [
"inequalities",
"inequalities solved"
],
"Problem": "[img]http://img291.imageshack.us/img291/1562/untitledly7.jpg[/img]",
"Solution_1": "[quote=\"eagerkill\"][img]http://img291.imageshack.us/img291/1562/untitledly7.jpg[/img][/quote]\r\n\r\nWhy can you set c = 0 ?\r\n\r\nAnd please don't use Imageshack or similar image storage sites. As far as I know, images uploaded there will be deleted eventually. You can attach images here on MathLinks by using the \"Add an Attachment\" feature when posting.\r\n\r\n Darij",
"Solution_2": "Because after homogenization, I thinks I can forget the constraint a^2+b^2+c^2=1 and establish another constraint. And new constraint I establish is c=0. So, is it wrong ??",
"Solution_3": "[quote=\"eagerkill\"]Because after homogenization, I thinks I can forget the constraint a^2+b^2+c^2=1 and establish another constraint. And new constraint I establish is c=0. So, is it wrong ??[/quote]\r\n\r\nSorry, I didn't see your reply until now.\r\nYou cannot establish every constraint you would like to. For instance, I hope it is clear to you that you cannot establish the constraint $a^{2}+b^{2}+c^{2}=0$ ;) . Before establishing a constraint you have to make sure that it is justified, i. e. that proving the inequality under the constraint will easily yield a proof for the general case. For instance, you can add a constraint $a+b+c=1$ to an inequality for three numbers $a$, $b$, $c$ if you are sure that proving the inequality for the case $a+b+c=1$ will yield a proof of this inequality for all $a$, $b$, $c$. Basically, establishing the constraint $a+b+c=1$ is the same as saying \"WLOG we assume that $a+b+c=1$\", and every time you say \"WLOG\" you have to make sure that there is really no loss of generality. See http://www.mathlinks.ro/Forum/viewtopic.php?t=20553 for more explanation.\r\n\r\nActually, dehomogenization is a pretty useless tool except for some special cases (mostly those constraints that allow a trigonometric substitution).\r\n\r\n Darij"
}
{
"Tag": [
"abstract algebra",
"advanced fields",
"advanced fields unsolved"
],
"Problem": "If $ f: S^n \\to S^n$ is continuous and $ \\Sigma f$ is the suspension of $ f$, how to show that $ f$ and $ \\Sigma f$ have the same (Brouwer) degree?\r\n\r\n\r\n(using some Mayer-Vietoris technique)",
"Solution_1": "naturality",
"Solution_2": "Sorry, you'll have to explain that to me...",
"Solution_3": "The isomorphism $ H^n(S^n) \\equal{} H^{n\\plus{}1}(\\Sigma S^n)$ is natural, so if $ f$ is multiplication by $ n$, so is $ \\Sigma f$.",
"Solution_4": "Thanks. Why is this an isomorphism? How does one construct it?",
"Solution_5": "It is the boundary map of Mayer-Vietoris. Its kernel and cokernel are the homology groups of hemispheres, which are contractible, hence $ 0$ for $ i > 0$.",
"Solution_6": "Great! Thanks!!!"
}
{
"Tag": [
"geometry",
"circumcircle",
"trapezoid"
],
"Problem": "Let the medians of the triangle $ABC$ intersect at point $M$. A line $d$ through $M$ intersects the circumcircle $ABC$ at $X$ and $Y$ so that $A$ and $C$ lie on the same side of $d$. Prove that $BX \\cdot BY = AX \\cdot AY+CX \\cdot CY$.",
"Solution_1": "Let $N$ be the midpoint of $AC$ and $A',B',C',N'$ the orthogonal projections of $A,B,C,N$ on the line $d.$ Segment $NN'$ becomes the median of the trapezoid $ACC'A'$ $\\Longrightarrow$ $NN' \\equal{} \\frac {_1}{^2}(AA' \\plus{} CC').$ But from $\\triangle MBB' \\sim \\triangle MNN',$ we get the proportion $\\frac {_{BB'}}{^{NN'}} \\equal{} \\frac {_{BM}}{^{NM}} \\equal{} 2.$ Hence, it follows that $BB' \\equal{} AA' \\plus{} CC' \\ (\\star).$ \n\nOn the other hand, if $R$ denotes the circumradius of $\\triangle ABC,$ we have the relations\n\n$BX \\cdot BY \\equal{} 2R \\cdot BB' \\ , \\ AX \\cdot AY \\equal{} 2R \\cdot AA' \\ , \\ CX \\cdot CY \\equal{} 2R \\cdot CC'.$ \n\nCombining these expressions with $(\\star)$ yields\n\n$\\frac {BX \\cdot BY}{2R} \\equal{} \\frac {AX \\cdot AY}{2R} \\plus{} \\frac {CX \\cdot CY}{2R} \\Longrightarrow BX \\cdot BY \\equal{} AX \\cdot AY \\plus{} CX \\cdot CY.$"
}
{
"Tag": [
"LaTeX"
],
"Problem": "\u600e\u4e48\u4e0b\u8f7d\u3001\u5b89\u88c5\u3001\u4f7f\u7528LATEX\uff1f",
"Solution_1": "\u5355\u51fb\u6700\u4e0a\u9762\u4e00\u6392\u7684\u7b2c\u4e09\u4e86: Latex Help\r\n\u5982\u679c\u4f60\u53ea\u60f3\u5728\u8fd9\u4e2a\u8bba\u575b\u5185\u4f7f\u7528\u7684\u8bdd, \u5c31\u6ca1\u5fc5\u8981\u4e0b\u8f7d. \u53ea\u9700\u5728code\u4e24\u8fb9\u52a0\u4e0adollar sign\u5c31\u53ef\u4ee5\u4e86.\r\n\u4f8b\u5982: a+b \\geq 2 \\sqrt{ab} \u52a0\u4e0a\u4e4b\u540e,\u5c31\u53d8\u6210\u4e86:\r\n$ a+b \\geq 2 \\sqrt{ab} $",
"Solution_2": "\u600e\u6837\u5199CODE?\u8bf7\u8be6\u7ec6\u544a\u8bc9\u6211.\u4e07\u5206\u611f\u8c22! :blush:",
"Solution_3": "[quote=\"honglei\"]\u600e\u6837\u5199CODE?\u8bf7\u8be6\u7ec6\u544a\u8bc9\u6211.\u4e07\u5206\u611f\u8c22! :blush:[/quote]\r\n\u8fd8\u662f\u5728\u90a3\u91cc, \u6709\u4e00\u4e2a\u8054\u63a5\u597d\u8c61\u53eb: symbols and signs."
}
{
"Tag": [
"LaTeX",
"binomial coefficients"
],
"Problem": "Problem 23 on the 1986 AHSME:\r\n\r\nLet $ N\\equal{}69^5\\plus{}5 \\cdot 59^4\\plus{}10 \\cdot 69^3\\plus{}10 \\cdot 69^2\\plus{}5 \\cdot 69\\plus{}1$. How many positive integers are factors of $ N$?\r\n\r\n(A) 3\r\n(B) 5\r\n(C) 69\r\n(D) 125\r\n(E) 216",
"Solution_1": "Let x=69\r\n\r\nTherefore, the expression becomes:\r\n\r\nx^5 + 5x^4 + 10x^3 + 10x^2 + 5x + 1\r\n\r\nnotice the binomial coefficients, we factor and get\r\n\r\n(x+1)^5\r\n\r\nreplacing x with 69\r\n\r\n70^5 = 2^5 * 5^5 * 7^5\r\n\r\nThus, the number of factors is \r\n\r\n(5+1)*(5+1)*(5+1) = [b]216[/b]\r\n\r\nSorry I don't know any LaTeX :blush:",
"Solution_2": "[quote=\"Eulers_Apprentice\"]$ x^{5} \\plus{} 5x^{4} \\plus{} 10x^{3} \\plus{} 10x^{2} \\plus{} 5x \\plus{} 1 \\equal{} (x \\plus{} 1)^{5} \\Longleftrightarrow$\n$ N \\equal{} 70^{5} \\equal{} 2^{5}\\cdot 5^{5} \\cdot7^{5}$\n\n$ \\therefore \\; d(N) \\equal{} (5 \\plus{} 1)^{3} \\equal{} \\boxed{216}\\ \\ \\ \\ \\blacksquare$\n[/quote]\r\n[hide=\"Code\"]\n[code]$x^{5} + 5x^{4} + 10x^{3} + 10x^{2} + 5x + 1 = (x + 1)^{5} \\Longleftrightarrow$\n$N = 70^{5} = 2^{5}\\cdot 5^{5} \\cdot7^{5}$\n\n$\\therefore \\; d(N) = (5 + 1)^{3} = \\boxed{216}\\ \\ \\ \\ \\blacksquare$[/code][/hide][url=http://www.artofproblemsolving.com/Wiki/index.php/LaTeX:About]LaTeX Wiki[/url]",
"Solution_3": "The problem in the original post says, :P \r\n[quote=\"ThetaPi\"]Problem 23 on the 1986 AHSME:\n\nLet $ N \\equal{} 69^5 \\plus{} 5 \\cdot \\color{blue} 59 \\color{black} ^4 \\plus{} 10 \\cdot 69^3 \\plus{} 10 \\cdot 69^2 \\plus{} 5 \\cdot 69 \\plus{} 1$. How many positive integers are factors of $ N$?\n[/quote]\r\n\r\nProbably just a typographical error."
}
{
"Tag": [
"calculus",
"integration",
"geometry unsolved",
"geometry"
],
"Problem": "What is the average thickness of a slab sawed at random from a round log ?",
"Solution_1": "Orl, are you sure this is not \"Advanced Fields\" or \"Calculus\"?\r\n\r\n darij",
"Solution_2": "[quote=\"darij grinberg\"]Orl, are you sure this is not \"Advanced Fields\" or \"Calculus\"?\n[/quote]\r\n\r\nYes, the solution involves in $\\int$ or even $\\int \\int$ depending on which solution you prefer. But yet the problems sounds so elementary. :lol:"
}
{
"Tag": [
"MATHCOUNTS",
"AMC",
"USA(J)MO",
"USAMO",
"AIME",
"FTW",
"number theory"
],
"Problem": "I'm gonna make a tournament, with difficulty ranging from hard Classroom Math to National Mathcounts. The tests will not be in any particular form(a.k.a you might have a short answer problem right in-between two multiple-choice problems. You may not use calculators at all, and I have a system of sign-ups:\r\n\r\nAnyone who has not completed middle-school can compete, with a few exceptions:\r\n\r\nIf you have gotten in the top 3rd at National Mathcounts, or have made USAMO at least once, you may not compete, so other middle-schoolers have a chance :wink:\r\nThe first test should be up in a few days.",
"Solution_1": "I join. Hi. Put USAMO problems in it.",
"Solution_2": "[quote=\"pythag011\"]I join. Hi. Put USAMO problems in it.[/quote]\r\n\r\nYou made usamo...\r\ncan i join?",
"Solution_3": "[quote=\"Urc\"]can i join?[/quote]\r\n\r\nSure! If you want to join, just say so!",
"Solution_4": "[quote=\"1=2\"]\nIf you have gotten in the top 3rd at National Mathcounts, or have made USAMO at least once, you may not compete, so other middle-schoolers have a chance :wink:[/quote]\r\n\r\nI have to comment on this. 75th is BARELY in the top 3rd :rotfl:\r\nWhy didn't you say top 57?\r\n\r\nNow I feel honored. :D\r\n\r\nBut I finished middle school... :(",
"Solution_5": "I wanna join! I made AIME, but not USAMO. Also, I've never done MC.\r\n\r\nBut my FTW rating is 1881 from memming, so don't use FTW problems :D",
"Solution_6": "You got it.\r\n\r\nPost here if you want to be in!",
"Solution_7": "And from the lack of popularity I have decided to withdraw from middleschool tournaments altogether. Bye! :bye:",
"Solution_8": "Can I join?",
"Solution_9": "[quote=\"1=2\"]And from the lack of popularity I have decided to withdraw from middleschool tournaments altogether. Bye! :bye:[/quote]\r\n\r\nWickedestjr, have you seen the last post of 1=2?",
"Solution_10": "oh wow... :rotfl:",
"Solution_11": "...Since the rules do not allow me to participate, I am taking over the tournament, if that's ok with you, 1=2?",
"Solution_12": "[quote=\"pythag011\"]...Since the rules do not allow me to participate, I am taking over the tournament, if that's ok with you, 1=2?[/quote]\r\n\r\nhm dont you have ur own tourney?",
"Solution_13": "naw, just let this die. You have your own USAMO tournament.",
"Solution_14": "yeah, but it died...maybe I made the problems a little hard...",
"Solution_15": "Due to the death of this tourney, I am hosting it now, and this time I will make easy equestions. But they still might be sort of hard.",
"Solution_16": "[quote=\"pythag011\"]Due to the death of this tourney, I am hosting it now, and this time I will make easy equestions. But they still might be sort of hard.[/quote]\r\n\r\nHmm...easy but hard? Like:\r\n\r\n1. (insert easy problem here, like 1+1)\r\n2. (insert hard problem here, like IMO)\r\n3. (insert easy)\r\n4. (hard)\r\n\r\nlike that?",
"Solution_17": "Like this is the first round.\r\n\r\n1. There is a Venn diagram. All of the regions have the same area. If the radius of each of teh circles is 1, and the centers from an equilateral triangle, find the side length.\r\n\r\n2.How many ways are tehre to choose 5 people from 10 people standing in a line if none of them can be standing next to each otehr.\r\n\r\n3. Find all primes p such that $ p^2\\plus{}2$ is NOT a multiple of 3.",
"Solution_18": "1. how many circles are in the venn diagram???",
"Solution_19": "the cetners form an equilateral triangle. Quote.",
"Solution_20": "Wondering: would we need proofs? or just answers?",
"Solution_21": "NOOOOOOOO Proofs. If you want proofs, go turn in solutions for my other tourney.",
"Solution_22": "Was that the real first round, or just an example of a possible first round?",
"Solution_23": "The real first round.",
"Solution_24": "Okay, well I am going to have to quit. Sorry. I really want to compete but I don't think that I will have enough time, because I am already hosting 5 math tournaments, and 3 other games.",
"Solution_25": "THE BONUS ROUND (for USAMO qualifiers)\r\n\r\nif you want to make USAMO this year, or already have, solve the following quesitons.\r\n\r\nThe first problem. PM me if you solve it with SOLUTIONS to get the second one.\r\n\r\nThis will be good practice for writing proofs.\r\n\r\nThe fisrt problem is pretty easy.\r\n\r\nDefine the fibonacci sequence as $ F_1 \\equal{} 1,$ $ F_2 \\equal{} 1,$, and $ F_{n\\plus{}1} \\equal{} F_n \\plus{} F_{n\\minus{}1}.$\r\n\r\nIf m and n are positive integers, prove taht gcd($ F_m, F_n$) = $ F_{gcd(m,n)}$"
}
{
"Tag": [
"algebra",
"polynomial",
"analytic geometry",
"inequalities",
"floor function",
"combinatorics unsolved",
"combinatorics"
],
"Problem": "Let $ A$ be set of all $ n$-tuples $ (a_1, a_2, \\ldots, a_n)$, where $ 0 \\leqslant a_i \\leqslant k$. \r\n\r\nWe say that element $ a \\equal{} (a_1, a_2, \\ldots, a_n)$ dominate element $ b \\equal{} (b_1, b_2, \\ldots, b_n)$ iff $ a_i >\\equal{} b_i$ for every $ i \\equal{} 1, 2, \\ldots n$. \r\n\r\nDeterminate maximal subset $ F$ of $ A$, such that there are no two elements $ a, b \\in F$ with $ a$ dominating $ b$.",
"Solution_1": "For $ k \\equal{} 1$ we have original Sperner's lemma, and domination is simple subset relation.\r\nI think that final result is family with all $ n$-tuples with equal sum, or equivalently maximal coefficient in polynomial \r\n\\[ (1 \\plus{} x \\plus{} x^2 \\plus{} \\ldots \\plus{} x^k)^n\\]",
"Solution_2": "Thank you for this beautiful problem. Did you ask to determine a maximal subset $ F$ or its maximal length? I can provide with the maximum length and that equals to $ (k + 1)^{n - 1}$.\r\n\r\nProof: The set if all n-tuples $ T(n)$ with each co ordinate $ 0\\leq a_i\\leq k,1\\leq i\\leq n$ forms a poset under the relation $ a < b\\quad iff \\quad a_i\\leq b_i\\quad \\forall i$. \r\n\r\nFor $ n > 1$,let $ \\{(i,e_{n - 1})|e_{n - 1}\\in T(n - 1)\\}$ denote all n-tuples whose first co ordinate is $ i$ and $ j$th co ordinate is $ (j - 1)$th coordinate of $ e_{n - 1}%Error. \"QUAD\" is a bad command.\n\\forall 2\\leq j\\leq n$\r\n\r\nLet $ C_{e_{n - 1}} = \\{(i,e_{n - 1})|\\forall 1\\leq i\\leq n\\}$.\r\nObserve that,\r\n$ \\{C_e|\\forall e\\in T(n - 1)\\}$ forms $ (k + 1)^{n - 1}$ disjoint minimum chains.By Dilworth's theorem, we have a antichain of desired length.(qed)\r\n\r\n[quote=\"alexilic\"]We say that element $ a = (a_1, a_2, \\ldots, a_n)$ dominate element $ b = (b_1, b_2, \\ldots, b_n)$ iff $ a_i > = b_i$ for every $ i = 1, 2, \\ldots n$. [/quote]\r\nWe can get to another variant of the problem by changing the inequality in the line to ...$ a_i>b_i$...",
"Solution_3": "You are wrong! \r\nFor $ k \\equal{} 1$, maximal number of elements is $ \\binom{2n}{n}$ which is greater than $ 2^{n \\minus{} 1}$, as you claim. \r\n\r\nKeep thinking...",
"Solution_4": "yes, I am wrong but not in the sense you said. I disagree on the fact that if $ k\\equal{}1$, the maximal antichain has length $ 2n\\choose n$. $ T(n,k\\equal{}1)$ contains $ 2^n$ elements. Let me list all of them for n=3,\r\n\r\n{(000),(001),(010),(011),(100),(101),(110),(111)}\r\n\r\n$ {6\\choose 3}\\equal{}20>|T(n,k\\equal{}1)|$. Here you are wrong.\r\n\r\nIn my previous post I suggested to break $ T(n)$ into chains like this,\r\n$ C_1\\equal{}\\{(000),(100)\\}$\r\n$ C_2\\equal{}\\{(001),(101)\\}$\r\n$ C_3\\equal{}\\{(010),(110)\\}$\r\n$ C_4\\equal{}\\{(011),(111)\\}$\r\n\r\nIf this was a minimum number of chains as I claimed, we would have got $ 2^2$ length antichain. But the longest antichains present here are $ \\{(100),(010),(001)\\}$ and $ \\{(011),(101),(110)\\}$. Here , I went wrong.\r\n\r\nHere is a minimum configuration of chains\r\n$ C_1\\equal{}\\{(000),(001),(101)\\}$\r\n$ C_2\\equal{}\\{(100),(110)\\}$\r\n$ C_3\\equal{}\\{(010),(011),(111)\\}$\r\n\r\nI am thinking about the possible closed form ie., the minimum number of chains for a general n and k. Let me know about new developments.",
"Solution_5": "I mistyped bound... The correct answer for $ k \\equal{} 1$ is $ \\binom{n}{\\lfloor \\frac{n}{2} \\rfloor}$. \r\n\r\nFor $ n \\equal{} 3$ we get $ 3$ sets and one such example is $ \\{001, 010, 100\\}$.",
"Solution_6": "We shall pick out the pairs in the following fashion: (0,k) (1,k-1),..\r\nFor a specific pair (m,n) we construct a n-tuple using exactly $ \\lfloor \\frac {n}{2} \\rfloor$ m's. This can be done in $ n\\choose {\\lfloor \\frac {n}{2} \\rfloor}$ ways ensuring maximality in this process. Any two tuples here are not comparable. Build such $ n$ tuples for every pair and we observe that any two n-tuples across two pairs are not comparable. We thus obtain a antichain of length ${ \\lfloor \\frac {k+1}{2} \\rfloor {n\\choose {\\lfloor \\frac {n}{2} \\rfloor} }}$. Observe that we cannot add any other n-tuple to this antichain. How do we prove or disprove that this is the longest antichain ?",
"Solution_7": "Imitating proof of a Sperner lemma we can obtain following inequality $ \\sum_{(a_1,a_2,...,a_n)\\in F}{\\frac {{k \\choose a_1}{k\\choose a_2}...{k \\choose a_n}}{{nk \\choose a_1 + a_2 + ... + a_n}}}\\leq 1$, where $ \\mathcal l{F}$ denotes set with required properties."
}
{
"Tag": [
"geometry",
"LaTeX"
],
"Problem": "Group $ \\beta$\r\n[b]Theoretical Round. Problems to solve.[/b] \r\n\r\n[b]1. Galaxy.[/b] The bolometric (total) stellar magnitudeof a galaxy visible from a distance of $ L_1\\equal{}3\\text{Mpc}$ is $ m_1\\equal{}6^m.88$. Find the bolometric stellar magnitude $ m_2$ of this galaxy visible from a distance of $ L_2\\equal{}3\\text{Gpc}$\r\n\r\n[b]2. Space Sail.[/b] A space probe for research of the asteriod belt is sent to objects of researches by using a solar sail. The probe-system of mass $ m\\approx 1\\text{ton}$ was moving around the Sun in a circular orbit of radius $ 1\\text{A.U.}$ Then such a sail (that was earlier a part of the probe-system) was opened and after half a revolution around the Sun the probe has reached Ceres. Estimate the area $ S$ and thickness $ d$ of this sail. Consider the sail is of mirror type. The typical distances of the asteroid belt bodies from the Sun are $ 2.8\\text{A.U.}$. The constant of total solar irradiation is $ A\\approx 1.37 \\text{kW/m}^2$\r\n\r\n[b]3. Alcohol in Universe.[/b] Astronomers based at Britain's Jodrell Bank Observatory say they have spotted a cloud of alcohol in deep space that measure 288 billion miles (463 billion kilometers) across, a finding that could shed light on how giant stars are formed from primordial gas. Some nations even decided to send expeditions to this area to taste the cloud but... enthusiasm disappeared after learning that the cloud is so remote that it will take us many million years to reach it on the modern techniques. The density of molecules in the cloud is large in comparison with interstellar gas density but very small from our custom point of view, only about $ 10\\text{atoms/mm}^3$. Estimate the temperature such a cloud should have to be stable and not get dispersed in future (before the international expedition will reach the cloud). Suppose the alcohol is ethyl ($ C_2H_5OH$) \r\n\r\n[b]5. Visit.[/b] Extraterrestrial animals (animaloids) decided to invade the Earth. The navigation devices of the spacecrafts of vituloids (lat. vitulus=seal) needs that sometimes after the landing the Polar star ($ \\alpha\\equal{}2^h32^m$, $ \\delta\\equal{}89^o16'$ at the epoch of starting the colonization) nust be exactly in Zenith and the devices of crocodiloids needs that sometime the Polar star must be exactly on the Horizon. The devices are placed on the top of the spacecrafts and these can land on the Earth only in vertical positions. Also every device has to be invisible directly for the device of any other spacecraft. Estimate, how many spacecraft of every type of animaloids can land on the Earth. Consider the Earth to be spherical, the spacecrafts' bottom to be on its surface (may stand either on ground or on the water) and height of every spacecraft is $ h\\equal{}10m$ \r\n\r\nEnjoy! :lol: \r\n\r\nP.S.: I have ommited $ 4$ as it involves a photograph. :)",
"Solution_1": "Anyone trying?\r\nNo. $ 2$ and $ 3$ are my favorites. I think there is a very elegant method(though not tried yet) for $ 2$, using the fact that luminosity follows inverse square law.",
"Solution_2": "hey i have done most of these and yes that elegant method was tremendous but the thing is i dont know latex !@#$ 98$",
"Solution_3": "No. 2 is a very nice problem. I couldn't solve it at all. Shreyas Patankar, can you please post your solution?"
}
{
"Tag": [
"geometry solved",
"geometry"
],
"Problem": "Let $ABC$ be an acute triangle, let $H$ be the foot of its altitude from $A$, and let $D$, $E$, $Q$ be the feet of the perpendiculars from an arbitrary point $P$ inside the triangle onto the lines $AB$, $AC$, $AH$, respectively. Prove that $\\left|AB\\cdot AD-AC\\cdot AE\\right|=BC\\cdot PQ$.",
"Solution_1": "Let $M=AP\\cap BC$, and let $B',C'$ be the feet pf the perpendiculars from $B,C$ to $AP$ respectively.\r\n\r\nWe have $AB\\cdot AD=AP\\cdot AB',\\ AC\\cdot AE=AP\\cdot AC'$, and we get $|AB\\cdot AD-AC\\cdot AE|=AP\\cdot B'C'$. What we must show now is $AP\\cdot B'C'=BC\\cdot PQ$, which is equivalent to $\\frac{AP}{PQ}=\\frac{BC}{B'C'}$. \r\n\r\nSince $\\frac{BM}{MB'}=\\frac{CM}{MC'}$, we get $\\frac{BC}{B'C'}=\\frac{BM}{MB'}$, so we have to prove $\\frac{AP}{PQ}=\\frac{BM}{MB'}$. This follows from the fact that $APQ,BMB'$ are similar right triangles (simple angle chase).",
"Solution_2": "Another solution:\r\n\r\n[b]Lemma 1.[/b] If $k_1$ and $k_2$ are two circles with centers $O_1$ and $O_2$, and if P and Q are two points such that $PQ\\perp O_1O_2$, then the difference between the powers of the point P with respect to the circles $k_1$ and $k_2$ equals the difference between the powers of the point Q with respect to the circles $k_1$ and $k_2$.\r\n\r\n[i]Proof.[/i] Since $PQ\\perp O_1O_2$, we have $\\angle PTO_1 = 90^{\\circ}$, where T is the point of intersection of the lines PQ and $O_1O_2$. Thus, by the Pythagoras theorem, $O_1P^2=O_1T^2+TP^2$. Similarly, $O_2P^2=O_2T^2+TP^2$.\r\n\r\nIf $r_1$ and $r_2$ are the radii of the circles $k_1$ and $k_2$, then, since the centers of these two circles are $O_1$ and $O_2$, the power of the point P with respect to the circle $k_1$ equals $O_1P^2-r_1\\ ^2$, and the power of the point P with respect to the circle $k_2$ equals $O_2P^2-r_2\\ ^2$. Therefore, the difference between these two powers is\r\n\r\n$\\left(O_1P^2-r_1\\ ^2\\right)-\\left(O_2P^2-r_2\\ ^2\\right)=O_1P^2-O_2P^2-r_1\\ ^2+r_2\\ ^2$\r\n$=\\left(O_1T^2+TP^2\\right)-\\left(O_2T^2+TP^2\\right)-r_1\\ ^2+r_2\\ ^2=O_1T^2-O_2T^2-r_1\\ ^2+r_2\\ ^2$.\r\n\r\nSo this is the difference between the powers of the point P with respect to the circles $k_1$ and $k_2$. But similarly, we get the same expression for the difference between the powers of the point Q with respect to the circles $k_1$ and $k_2$. Thus, the two powers are equal, and Lemma 1 is proven.\r\n\r\nNow, to the solution of the problem:\r\n\r\nWe will work with directed segments.\r\n\r\nIf X is the orthogonal projection of the point P on the line BC, we have < PXH = 90\u00b0; since we also have < XHQ = 90\u00b0 and < PQH = 90\u00b0, it follows that the quadrilateral PQHX has three right angles, so that it is a rectangle. Thus, PQ = XH.\r\n\r\nSince < PXB = 90\u00b0 and < PDB = 90\u00b0, the points X and D lie on the circle with diameter PB. Denote this circle by y; the center of this circle is the midpoint Y of the segment PB. Similarly, the points X and E lie on the circle z with diameter PC, and the center of this circle is the midpoint Z of the segment PC. Now, since the points Y and Z are the midpoints of the sides PB and PC of triangle BPC, respectively, we have YZ || BC. On the other hand, since the line AH is an altitude in triangle ABC, we have $AH\\perp BC$. Thus, $AH\\perp YZ$. Since the points Y and Z are the centers of the circles y and z, we can thus apply Lemma 1 and conclude that the difference between the powers of the point A with respect to the circles y and z equals the difference between the powers of the point H with respect to the circles y and z. Now, the power of the point A with respect to the circle y is $AB\\cdot AD$, the power of the point A with respect to the circle z is $AC\\cdot AE$, the power of the point H with respect to the circle y is $HB\\cdot HX$, and the power of the point H with respect to the circle z is $HC\\cdot HX$. Thus, we have\r\n\r\n$AB\\cdot AD-AC\\cdot AE = HB\\cdot HX-HC\\cdot HX$.\r\n\r\nIn other words,\r\n\r\n$AB\\cdot AD-AC\\cdot AE = \\left(HB-HC\\right)\\cdot HX = CB\\cdot HX = BC\\cdot XH$.\r\n\r\nSince PQ = XH, we thus get\r\n\r\n$AB\\cdot AD-AC\\cdot AE = BC\\cdot PQ$.\r\n\r\nFor non-directed segments, this rewrites as\r\n\r\n$\\left|AB\\cdot AD-AC\\cdot AE\\right| = BC\\cdot PQ$.\r\n\r\n$\\blacksquare$\r\n\r\n Darij"
}
{
"Tag": [
"number theory open",
"number theory"
],
"Problem": "I have been trying to prove that $ 12\\cdot{x^3}\\minus{}3\\equal{}n^2$ has no solutions in naturals numbers with $ x > 1$ \r\n\r\nI first thought of finding a mod in which it makes no sense, but I just couldn't find one (after some attempts).\r\nThanks in advance.",
"Solution_1": "If we are allowed to use Fermat's Last Theorem, I can solve this.\r\n\r\nClearly, $ 3 \\mid n$ and $ n > 0$ since $ x > 1$. Let $ n \\equal{} 3m$ for some integer $ m > 0$. Then \\[ 12x^3 \\minus{} 3 \\equal{} 9m^2 \\Rightarrow 4x^3 \\minus{} 1 \\equal{} 3m^2. \\] Since the left-hand side is an odd number, $ m$ is odd. Let $ m \\equal{} 2k\\plus{}1$ for some integer $ k \\geq 0$. By expanding, we get that \\[ 4x^3 \\equal{} 3(2k\\plus{}1)^2 \\equal{} 12k^2 \\plus{} 12k \\plus{} 4. \\] Therefore, by dividing by four on both sides, \\[ x^3 \\equal{} 3k^2 \\plus{} 3k \\plus{}1 \\Rightarrow x^3 \\plus{} k^3 \\equal{} (k\\plus{}1)^3. \\] If $ k \\equal{} 0$, then $ x \\equal{} 1$, which is impossible since $ x > 1$. Otherwise, $ k > 0$ and by Fermat's Last Theorem, there are no positive integer solutions to this equation.\r\n\r\n(PS. I am sure that the equation $ x^3 \\plus{} k^3 \\equal{} (k\\plus{}1)^3$ not having positive integer solutions can be proved using elementary means and using FLT is probably overkill.)"
}
{
"Tag": [
"calculus",
"integration"
],
"Problem": "Let p>2 be a prime, and a, b, c, d be integers not divisible by p, such that:\r\n{ra/p} + {rb/p} + {rc/p} + {rd/p} = 2\r\nfor any integer r not divisible by p.\r\n\r\nProve that at least two of the numbers a+b, a+c, a+d, b+c, b+d, c+d are divisible by p.\r\n\r\n{} is the fractional part function.",
"Solution_1": "An initial thought:\r\nWe probably have to have p | (a + b) and p | (c + d), or some permutation thereof. If that is the case, we definitely have a working solution. So we have to prove that in any other arrangement, there will be some r such that the sum doesn't work. Also, we can take a, b, c, d and r in [1, p - 1]. Hmm...\r\n\r\nOkay, further thoughts: letting r = 1, we get \r\n{a/p} + {b/p} + {c/p} + {d/p} = 2 which gives us\r\na + b + c + d = 2p.\r\nNow, we know that p/2 is not integral, so each of a, b, c and d is strictly greater or less than p/2. Furthermore, we know that at least one of the four must be greater and at least one of the four must be less. WLOG assume a > p/2, b < p/2. Then letting r = 2, we get\r\n{2a/p} + {2b/p} + {2c/p} + {2d/p} = 2\r\n(2a - p) + 2b + (2c + 2d - np) = 2p where n = 0, 1, or 2: 0 if c, d < p/2, 2 if c, d > p/2, 1 if they are split (one above, one below). \r\nThis gives us 2(a + b + c + d) = (3 + n)p from which we get n = 1. So assume WLOG a, c > p/2 > b, d.\r\nNow we just have to show the sums are what we want.\r\n\r\n\r\nI wonder if I used a similar argument for r = p - 1, would that finish it off? Hmmm ... will continue thinking on it.",
"Solution_2": "bump\r\n\r\nI think you can also assume that say a = 1 <= b < p/2 <=c <=d, and try to prove that d = p-1\r\n\r\n*note, you can assume a = 1, because if a != 1, there will be some k s.t. ak=1 mod p, and then a' = 1, b' = bk, c' = ck, d'=dk and relabel",
"Solution_3": "rebump"
}
{
"Tag": [
"inequalities",
"geometry",
"\\/closed"
],
"Problem": "I wanted to post a problem to advanced forum, but I don't know to which category (I was thinking about inequalities or geometry). Maybe there will be more such problems. What should I do with them? Maybe there's forum named other problems somewhere but I haven't found it.",
"Solution_1": "You can put it in either Geometry or Inequalities. If it is mostly unappropriate moderators will move it from one to another. :)"
}
{
"Tag": [],
"Problem": "prove that there are no positive integer solutions $ x,y,z$ such that $ 2xz\\equal{}y^{2}$ and $ x \\plus{} z\\equal{}1994$\r\n\r\n\r\n\r\n\r\n\r\n[b]PLEASE HIDE YOUR SOLUTIONS[/b]",
"Solution_1": "well here is my solution .i am not sure whether i am right\r\n\r\n[hide=\"Unsure solution\"] the given equation we have $ 2xz$ to be a perfect square.$ 2$ is not a perfect square ,so we have $ xz = 2^{2n + 1}$ where $ n %Error. \"belongsto\" is a bad command.\n\\ N$\nsince a power of 2 will be divisible by numbers which are multiples of 2 alone,we have $ x = 2^{k}$ and $ z = 2^{l}$\nlets assume without loss of generality that $ l > k$\nwe have $ 2^{k} + 2^{l} = 1994$\n$ \\implies$\n$ 2^{k}(1 + 2^{l - k}) = 1994$\n$ 2^{k - 1}(1 + 2^{l - k}) = 997$\nthis is not possible because the [b]L.H.S[/b] is even and the [b]R.H.S[/b] is odd.\nso........... there are no solutions.\n\nAm i correct in my solution[/hide]",
"Solution_2": "well i think it can be solved [hide=\"using\"]some infinite descent or something like that case.but i dont know wat it is :( .[/hide]",
"Solution_3": "this is wat i have \r\n[hide]we have $ 1994 \\equal{} x \\plus{} z$\nnote that $ 1994 \\equal{} 2 \\cdot 997$(997 is a prime)\nsquare both sides to get-\n$ 4 \\cdot (997)^2 \\equal{} x^2 \\plus{} z^2 \\plus{} 2xz$\nor,\n$ 4 \\cdot (997)^2 \\equal{} x^2 \\plus{} z^2 \\plus{} y^2$\nwe have too cases,\nsince $ y^2$ is a perfect square and even .\n$ y^2\\equiv 0\\mod 4$\nsince $ y^2 \\equal{} 2xz$,\nwe have two cases,\n [u]Case 1[/u] - $ x,z \\equal{} 4k$\n this means $ (x \\plus{} z)^2 \\equal{} (8a)^2 \\equal{} 64\\cdot a^2$\n also $ 64\\cdot a^2 \\equal{} 4 \\cdot (997)^2$\n $ 997 \\equal{} 4 \\cdot 249 \\plus{} 1$\n therefore, $ 997 \\equal{} 4b \\plus{} 1$\n or, $ 16 \\cdot a^2 \\equal{} 4b \\plus{} 1$\n which is not possible since the LHS is a multiple of 4 but the RHS isnt\n [u]Case 2[/u]- $ x,z\\equiv 2\\mod 4$\n therefore either $ x,z\\equiv 2\\mod 4$\n or $ z,x\\equiv 1\\ mod 4$ respectively\n but this makes $ x^2 \\plus{} y^2 \\plus{} z^2\\equiv 1\\mod 4$\n which is a contradiction\nhence no solutions in integers.[/hide]",
"Solution_4": "ok i guess my solution was long!\r\nanyway can anyone tell what infinte descent is?",
"Solution_5": "geniusbliss,\r\ntwo cases need not be taken\r\n$ y^2$ is always congruent to 0 mod4 as it is even\r\n\r\nok you have edited it. :lol:",
"Solution_6": "srinath.r: your solution is ocmpletely wrong.\r\n\r\ngeniusbliss:good solution but you have made a mistake though still the solution is correct.\r\n$ x,z$ are even {whihc can be proved by divisibility}.\r\nIn Case 1: $ x\\equal{}4k, z\\equal{} 4k'$ is how you should have written because it is not necessary that $ x$ and $ z$ are same.\r\nStill it comes to $ 16(k\\plus{}k')^2 \\equal{} 4*997$ whihc is not possible clearly.\r\nOtherwise, a nice one.\r\n\r\nWell, infinite descent cannot solve this one as for the second time u cannot say the same even divisibilty when you get $ 997$.",
"Solution_7": "well ok but i dint mean to write it as $ x,z \\equal{} 4k$ i was thinking one thing but typed another :)"
}
{
"Tag": [
"algebra",
"polynomial",
"geometry",
"angle bisector"
],
"Problem": "1. let f(x) = x^6+x^5+x^4+x^3+x^2+x + 1 determine remnants for f(x^7)/f(x)\r\n 2. the sum of a four-digit number ands its four digits is 2005. what is this four-digit number?\r\n 3. In triangle ABC, AB=10 and AC=18. Mis the midpoint of of BC, and the line through M parrallel to the bisector of CAB cuts AC at D. find the length of AD.\r\n 4. Let x,y and z be positive numbers such that x+y+xy=8 y+z+yz=15 z+x+zx=35 find the value of x+y+z+xy\r\n 5. Let A be the larger root of (2004x)^2 -(2003)(2005)x -1=0 and B be the smaller root of x^2 + 2003x - 2004 =0. Determine the value of A-B \r\n 6.Determine all integers n such that n^4 - 4n^3 +15n^2-30n + 27 is a prime number\r\n 7.Let a,b and c be real numbers such that a+bc = b+ca = c+ab =501. If M is the maximum value of a+b+c and m is the minimum value of a+b+c . Determine the value of M + 2m\r\n 8.the positive integers a,b and care such that a+b+c =20=ab + bc -ca-b^2 Determine all positive value of abc.",
"Solution_1": "[hide=\"A Start for #2\"]\\begin{align*}(1000a+100b+10c+d)+(a+b+c+d)&=2005\\\\ & \\Downarrow\\\\ 1001a+101b+11c+2d&=2005\\end{align*}\n$ a=0$ results in a 3-digit number; $ a=2$ is impossible unless $ d=1.5$, but $ d$ must be an integer. $ \\therefore a=1$.\n\n\\begin{align*}1001+101b+11c+2d&=2005\\\\ & \\Downarrow\\\\\n101b+11c+2d&=1004\\end{align*}[/hide]",
"Solution_2": "[quote=\"watcharapon\"]1. let f(x) = x^6+x^5+x^4+x^3+x^2+x + 1 determine remnants for f(x^7)/f(x)\n 2. the sum of a four-digit number ands its four digits is 2005. what is this four-digit number?\n 3. In triangle ABC, AB=10 and AC=18. Mis the midpoint of of BC, and the line through M parrallel to the bisector of CAB cuts AC at D. find the length of AD.\n 4. Let x,y and z be positive numbers such that x+y+xy=8 y+z+yz=15 z+x+zx=35 find the value of x+y+z+xy\n 5. Let A be the larger root of (2004x)^2 -(2003)(2005)x -1=0 and B be the smaller root of x^2 + 2003x - 2004 =0. Determine the value of A-B \n 6.Determine all integers n such that n^4 - 4n^3 +15n^2-30n + 27 is a prime number\n 7.Let a,b and c be real numbers such that a+bc = b+ca = c+ab =501. If M is the maximum value of a+b+c and m is the minimum value of a+b+c . Determine the value of M + 2m\n 8.the positive integers a,b and care such that a+b+c =20=ab + bc -ca-b^2 Determine all positive value of abc.[/quote]\r\n\r\nBrief answers. \r\n\r\n[hide=\"1\"]\n$ f(x) = (x^6 + x^5 + x^4 + x^3 + x^2 + x + 1)Q(x) + R(x)$, where $ R(x)$ is at most a fifth degree polynomial.\n\nLet $ \\omega = e^{2i\\pi/7}$, then $ f(\\omega) = f(\\omega^2) = \\cdots = f(\\omega^6) = 0$ and $ f(\\omega^7) = f((\\omega^2)^7) = \\cdots = f((\\omega^6)^7) = 7$, so $ R(\\omega) = R(\\omega^2) = \\cdots = R(\\omega^6) = 7$. But thats six values, and $ R(x)$ is fifth degree polynomial, so $ \\boxed{R(x) = 7}$. \n[/hide]\n\n[hide=\"2\"]\nLet $ s(n)$ be sum of digits of $ n$. Since $ s(n) \\le 1 + 9 + 9 + 9 = 28$, nothing $ \\le 1976$ works; trivial casework shows $ n \\ge 2000$ doesn't work either. So $ n + s(n) = 2005 \\Longrightarrow \\overline{19ab} + (1 + 9 + a + b) = 2005 \\Longrightarrow 11a + 2b = 95$. Since $ a, b$ are digits, by checking parities we see that $ \\boxed{1979}$ is the unique answer. \n[/hide]\n\n[hide=\"3\"]\nLet $ N$ be foot of angle bisector. By [[angle bisector theorem]] we have $ BN = 5x, CN = 9x$ for some $ x$, and thus $ MC = \\frac {BC}{2} = \\frac {BN + CN}{2} = 7x$. Since $ \\triangle CDM \\sim \\triangle CAN$ by corresponding angles etc,\n\\[ \\frac {CD}{18} = \\frac {7x}{7x + 2x} \\Longrightarrow CD = 14\n\\]\nAnd so AD = $ 18 - CD = \\boxed{4}$.\n[/hide]\n\n[hide=\"4\"]\nRewrite $ x + y + xy + 1 = 9$ etc, and factoring,\n\\begin{align} (x + 1)(y + 1) & = 9 \\\\\n(y + 1)(z + 1) & = 16 \\\\\n(x + 1)(z + 1) & = 36 \\end{align}\nTheir product is\n\\begin{align}(x + 1)^2(y + 1)^2(z + 1)^2 = 9 \\cdot 16 \\cdot 36 \\Longrightarrow (x + 1)(y + 1)(z + 1) = 72\\tag{4}\\end{align}\nDivide each of $ (1),(2),(3)$ from $ (4)$ to get\n\\[ z + 1 = 8,\\ x + 1 = 4.5,\\ y + 1 = 2\n\\]\n, and so $ x + y + z + xy = 3.5 + 1 + 7 + 3.5 = \\boxed{15}$.\n[/hide]\n\n[hide=\"5\"]\n$ (2004^2x + 1)(x - 1) = 0$ and $ (x + 2004)(x - 1) = 0$, so $ A - B = 1 - ( - 2004) = \\boxed{2005}$.\n[/hide]\n\n[hide=\"6\"]\nRewrite as $ [(n - 2)n(n + 2)]n^2 + 3(5n^2 - 10n + 9)$; modulo 3, we see that at least one of $ n - 2, n, n + 2$ must be divisible by $ 3$ so the entire expression is vidisible by $ 3$. So the only time the answer can be prime is if\n\\[ n^4 - 4n^3 + 15n^2 - 30n + 27 = 3\n\\]\nwhich doesn't seem to occur. \n[/hide]\r\n\r\n7\r\n\r\n8",
"Solution_3": "5.\r\nfrom the RRT we know that A=1 and B=-2004 so $ A \\minus{} B \\equal{} 2005$\r\n7\r\n.$ a \\plus{} bc \\equal{} b \\plus{} ca < \\equal{} > a \\minus{} ca \\plus{} bc \\minus{} b \\equal{} 0 < \\equal{} > a \\minus{} b \\minus{} c(a \\minus{} b) \\equal{} 0 < \\equal{} > (a \\minus{} b)(1 \\minus{} c) \\equal{} 0$\r\nlike this we get $ (1 \\minus{} a)(b \\minus{} c) \\equal{} 0$ and $ (1 \\minus{} b)(a \\minus{} c) \\equal{} 0$\r\nbecause if $ a \\equal{} b \\equal{} c \\equal{} 1 \\equal{} > a \\plus{} bc \\equal{} 2$ and not $ 501$ we just know that $ a \\equal{} b \\equal{} c$\r\nand \r\n$ a^2 \\plus{} a \\minus{} 501 \\equal{} 0$\r\n$ D \\equal{} 2008$\r\n$ a1 \\equal{} (1 \\plus{} \\sqrt{2008})/2$ and $ a2 \\equal{} (1 \\minus{} \\sqrt{2008})/2$\r\nso \r\n$ M \\equal{} 3*a1 \\equal{} (3 \\plus{} 3*(\\sqrt{2008}))/2$\r\n and\r\n $ 2m \\equal{} 6*a2 \\equal{} 3 \\minus{} 3*(\\sqrt{2008})$\r\nso\r\n [u]$ M \\plus{} 2m \\equal{} 15/2 \\minus{} 3/2(\\sqrt{2008})$[/u]"
}
{
"Tag": [
"inequalities",
"number theory unsolved",
"number theory"
],
"Problem": "prove that the equation dose not have any answer in Fibonacci numbers with a1=a2=1\r\n\r\n\r\n Fn = p^a and p=4t+3 t>0",
"Solution_1": "$ n \\equal{} 10, t \\equal{} 13, a \\equal{} 1$.\r\n\r\nUPD Even easier: $ n\\equal{}1, t\\equal{}21991023834438738, a\\equal{}0$.",
"Solution_2": "i searched for the condition(s) that $ n$ must have if such equation holds and i got this:\r\n$ p$ is prime ,no?\r\nif so,let's assume $ n \\equal{} q_1^{\\alpha_1}...q_k^{\\alpha_k}$ and let $ k > 1$.\r\nif $ n \\equal{} 2^s.\\beta$ and $ \\beta > 1$ then obviously we must have $ F_{\\beta} \\equal{} p^b$ where $ b > 0$.but we know that for all $ m > 0$ we have $ F_{2m \\minus{} 1}^2 \\plus{} F_{2m \\plus{} 1}^2 \\plus{} 1 \\equal{} 3F_{2m \\minus{} 1}F_{2m \\plus{} 1}$(see PEN L5).and if we put $ m \\equal{} \\frac {\\beta \\minus{} 1}{2}$ we get that $ F_{\\beta}$ doesn't have any prime divisors equivalent to $ \\minus{} 1$ modulo $ 4$.\r\nso we get that $ n \\equal{} 2^s$.but in this case since $ s$ must be greater than $ 1$ we get that $ F_4 | F_n \\implies 3 |F_n \\implies p \\equal{} 3$ and it is a contradiction with $ p \\equal{} 4t \\plus{} 3,t > 0$.u can see that if $ s>2$ then it doesn't work at all!\r\ni tried another way to solve this nice problem(without using what was mentioned in PEN L5) and i got only this: :( \r\nwe'll try to prove that $ n$ is a prime(i assume that $ n$ is odd if not then take the odd part of $ n$ and replace it by $ n$).\r\nnote that this odd part which is $ > 1$ exists (if not then $ n \\equal{} 2^s \\implies p \\equal{} 3$).\r\nwe know that $ F_l | F_{lt}$.for all $ l,t \\in N$.\r\nit's obvious that in this case we must have \r\n\r\n$ F_{q_1} \\equal{} p^{t_1}$ and also $ F_{q_k} \\equal{} p^{t_k}$ \r\n\r\nbut since we had assumed $ k > 1$ and also \r\n\r\n$ \\gcd(F_{q_1},F_{q_k}) \\equal{} F_1 \\equal{} 1$.\r\n\r\nso we have a clear contradiction by assuming that $ k > 1$.\r\n\r\nso we must have $ n \\equal{} q^b$ for some odd prime $ q$ and integer $ b$.\r\n\r\nobviously we have $ b < a \\plus{} 1$.\r\n\r\nto prove it just see that $ F_q < F_{q^2} < ... < F_{q^{b}}$ and all $ F_{q^{i}},0 < i < b \\plus{} 1$ divide $ p^a$ which implies that \r\n\r\n$ F_q \\equal{} p^{a_1} < F_{q^2} \\equal{} p^{a_2} < ... < F_{q^{b}} \\equal{} p^{a_b}$ \r\n\r\nand \r\n\r\n$ 0 < a_1 < a_2 < ... < a_{b} \\equal{} a$.\r\n\r\nso it's obvious that $ b < a \\plus{} 1$.\r\n\r\nlet's assume $ b > 1$ and $ q \\neq p$.\r\n\r\nwe have proven in PEN L4 this property : \r\n\r\n$ F_t^3 | F_{tl} \\minus{} F_{t \\plus{} 1}^l \\plus{} F_{t \\minus{} 1}^l$.now we put $ t \\equal{} l \\equal{} q$ which implies that \r\n\r\n$ F_q^3 | F_{q^2} \\minus{} F_{q \\plus{} 1}^q \\plus{} F_{q \\minus{} 1}^q$ but we have $ F_q \\equal{} p^{a_1}$ and\r\n\r\n $ p^{a_1 \\plus{} 1} | F_q^3 | F_{q^2} \\minus{} F_{q \\plus{} 1}^q \\plus{} F_{q \\minus{} 1}^q$ and also since \r\n\r\n$ F_{q^2} \\equal{} p^{a_2},a_2 > a_1$ \r\nwe get that \r\n$ p^{a_1 \\plus{} 1} | F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q \\equal{} (F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}).(\\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}}) \\equal{} (p^{a_1}).(\\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{p^{a_1}}) \\implies p | \\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{p^{a_1}}$\r\nwhich is a clear contradiction since\r\n $ \\gcd(F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1},\\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}}) \\equal{} (F_q,q) \\equal{} 1 \\implies p \\nmid \\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{p^{a_1}}$\r\nit means that we have $ b \\equal{} 1$ or $ q \\equal{} p$.\r\nthe first case seems difficult.(for me :D )it is $ F_q \\equal{} p^a$.\r\nfor the second one \r\nwe get that $ F_q \\equal{} q^{a_1},F_{q^2} \\equal{} q^{a_2},...,F_{q^{b}} \\equal{} q^{a_b} \\equal{} q^a$. \r\nby using the same theorem mentioned above this time for $ p \\equal{} q$ we get that if $ b > 1$ then \r\n$ F_{q}^{3}| F_{q^{2}} \\minus{} F_{q \\plus{} 1}^{q} \\plus{} F_{q \\minus{} 1}^{q} \\implies q^{a_1 \\plus{} 2} | F_{q \\plus{} 1}^{q} \\minus{} F_{q \\minus{} 1}^{q} \\equal{} \\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}}.F_q \\implies q^2 | \\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}}$\r\nnote that i have used the fact that $ q^{a_1 \\plus{} 2} | F_{q^2}$.to prove it just use inequality and prove that $ q.F_q < F_{q^2}$.\r\nbut u can easily see that (by the well-known theorem of lifting the exponent)\r\n$ q \\parallel{} \\frac {F_{q \\plus{} 1}^q \\minus{} F_{q \\minus{} 1}^q}{F_{q \\plus{} 1} \\minus{} F_{q \\minus{} 1}}$.so we have a contradiction with $ b > 1$.\r\nit means we have $ b \\equal{} 1$.\r\nand we have now $ F_q \\equal{} q^a$.\r\nso i have proven that if there exists $ n$ s.t. $ F_n \\equal{} p^a$ and $ p$ and $ n$ are odd then $ n \\equal{} q$ where $ q$ is prime.\r\nhope i didn't make any mistake! :D",
"Solution_3": "In fact, if we look for solution with $ a>1$ and $ p$ arbitrary, there are only $ 4$ such solutions: $ 1,1,8,144$.",
"Solution_4": "[quote=\"zhoraster\"]In fact, if we look for solution with $ a > 1$ and $ p$ arbitrary, there are only $ 4$ such solutions: $ 1,1,8,144$.[/quote]\r\ni didn't try on it but i'm curious to know how u got this ?",
"Solution_5": "It's not me who proved this, but [url=http://arxiv.org/abs/math/0403046]Bugeaud et al[/url]"
}
{
"Tag": [],
"Problem": "I need some help figuring out this complicated formula.\r\n\r\nAn event with 5 possibilities.\r\nThe chance of each event is based on X. X can range from 0 to 1. We are given the following information.\r\n\r\nThe chance of event 1 occurring is:\r\nx = 1/5, chance = 0%\r\nx = 2/5 chance = 1%\r\nx = 3/5 chance = 1%\r\nx = 4/5 chance = 11%\r\nx = 1 chance = 25%\r\n\r\nThe chance of event 2 occurring is:\r\nx = 1/5, chance = 2%\r\nx = 2/5 chance = 6%\r\nx = 3/5 chance = 18%\r\nx = 4/5 chance = 34%\r\nx = 1 chance = 55%\r\n\r\nThe chance of event 3 occurring is:\r\nx = 1/5, chance = 11%\r\nx = 2/5 chance = 24%\r\nx = 3/5 chance = 48%\r\nx = 4/5 chance = 40%\r\nx = 1 chance = 16%\r\n\r\nThe chance of event 4 occurring is:\r\nx = 1/5, chance = 33%\r\nx = 2/5 chance = 49%\r\nx = 3/5 chance = 26%\r\nx = 4/5 chance = 14%\r\nx = 1 chance = 3%\r\n\r\nThe chance of event 5 occurring is:\r\nx = 1/5, chance = 54%\r\nx = 2/5 chance = 20%\r\nx = 3/5 chance = 7%\r\nx = 4/5 chance = 1%\r\nx = 1 chance = 1%\r\n\r\nEach event is mutually exclusive (only 1 event can occur at a time). The total chance over the 5 events must equal 100% for any one X value.\r\n\r\nHow do I figure out the chance of any event when x is any number?",
"Solution_1": "i dont get it. could u explain why 2/5 chance=10% for example?",
"Solution_2": "No, he's saying that first you pick an X value:\r\n\r\nSay I pick 3/5.\r\n\r\nThen the chance of event one happening is 1%.\r\nThe chance of event two is 18%\r\nThe chance of event three is 48%\r\nThe chance of event four is 26%\r\nThe chance of event five is 7%\r\n\r\nThat only applies when X = 3/5.",
"Solution_3": "can no one help me with this?"
}
{
"Tag": [
"logarithms",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Prove that if $ a,x,y$ are $ p$-adic integers different from $ 0$ and $ p | x, pa | xy$, then \\[ \\frac 1y \\frac{(1\\plus{}x)^y\\minus{}1}{x} \\equiv \\frac{\\log (1\\plus{}x)}{x} \\;\\;\\;\\; ( \\textrm{mod} \\; a\\ ) \\\\\\\\ .\\] \r\n\r\n[i]L. Redei[/i]",
"Solution_1": "Using the power series expansions for the $p$-adic [url=https://en.wikipedia.org/wiki/P-adic_exponential_function]exponential and logarithm[/url], we have for $\\log(1+x)=z$ that\n\\[\n\\frac{(1+x)^y-1}{y} = \\frac{\\exp\\big(y\\log(1+x)\\big)-1}{y}=z+\\frac{yz^2}{2!}+\\frac{y^2z^3}{3!}+\\dots\n\\]\nTherefore, the problem reduces to\n\\[\nS=\\frac{yz^2}{2!}+\\frac{y^2z^3}{3!}+\\dots\\equiv 0\\pmod{ax},\n\\]\nwhere it suffices to show $pS\\equiv 0\\pmod{x^2y}$ because $pa|xy$. However, this follows from the fact that $x^2|p\\cdot \\frac{z^k}{k!}$, where\n\\[\n\\frac{z}{x}=\\frac{\\log(1+x)}{x}=1-\\frac{x}{2}+\\frac{x^2}{3}\\mp\\dots\n\\]\nis a $p$-adic integer (due to $v_p(x^{k-1})\\ge k-1\\ge v_p(k)$), so our divisibility in turn reduces to $p\\cdot \\frac{z^{k-2}}{k!}$ being a $p$-adic integer, which is clear, because by [url=http://artofproblemsolving.com/wiki/index.php?title=Legendre%27s_Formula]Legendre's formula[/url], $v_p(k!)\\le \\frac{k-1}{p-1}\\le (k-2)v_p(z)+1$. $\\blacksquare$"
}
{
"Tag": [
"trigonometry"
],
"Problem": "I'm not sure if this belongs in this forum, but here goes:\r\nUsing only algebra, simplify the below:\r\n1.$\\sqrt[3]{i}$\r\n\r\n2. $\\frac{1}{\\sqrt{i}+i}$",
"Solution_1": "Yes, probably this forum... by using only algebra do you mean complex trig stuff allowed?\r\n[hide=\"1\"]$i=e^{i\\frac{\\pi}{2}}, i^{\\frac13}=e^{i\\frac{\\pi}{6}},e^{i\\frac{5\\pi}{6}},e^{i\\frac{9\\pi}{6}}$... that's pretty messy stuff...[/hide]\n[hide=\"2\"]rationalize twice: first multiply thru by $\\sqrt{i}=i$, then by $i-1$. Now you just calculate the top...[/hide]",
"Solution_2": "If \"using only algebra\" means that De Moivre is shunned, then we should proceed thus: $x^{3}=i\\implies x^{6}=-1$, hence we should solve $x^{6}+1=0$ and then take only the three solutions that satisfy $x^{3}=i$ (the remaining three will give $x^{3}=-i$, of course).\r\n\r\nSimilar for the second problem: find the two values of $\\sqrt{i}$ by solving $x^{4}+1=0$, and then substitute them and simplify.\r\n\r\nHint for both equations: Factorize"
}
{
"Tag": [],
"Problem": "ok. Well, basically this topic is to right about funny things that happen....and to have fun!! JOKES, FUNNY STUFF, RIDDLES, and other. \r\nOk. now that's all I have to say..:P\r\nhm.Do you want to have a riddle for the day? I will do it, but if I don't do it one day, someone else do it!\r\nWE CAN!!! Then the next person below this post, posts a riddle, joke, but the riddle a day, I do....ok anyway LETS START!\r\n[b]\nRiddle 1 (Sunday riddle):What do you call a fish without an eye?[/b]\r\n[hide=\"Answer, figure riddle out before looking!\"]FSH!Get it, the i...?[/hide]",
"Solution_1": "haha? that was just sooo funny k00lperson\r\nbut \r\n\r\n\r\n[hide=\"What did one Smuckers bottle say to the other?\"]frankly scarlett, i dont give a jam[/hide]",
"Solution_2": "what is this?\r\n<3~\r\nmce mce mce \r\n\r\n[hide=\"hint\"]not <:3~[/hide]",
"Solution_3": "..idk?\r\n\r\n[b]Monday Riddle: Why was the baby ant confused?[/b]\r\n[hide=\"Answer, solve riddle before looking!\"]All his uncles were ants![/hide]",
"Solution_4": "@k00lperson... 3 BLIND MICE!\r\n\r\nkinda funny to give an IDK after your fish one.",
"Solution_5": "[hide=\"i found this amazing joke online\"]Fresh out of business school, the young man answered a want ad for an accountant. Now he was being interviewed by a very nervous man who ran a small business that he had started himself. \"I need someone with an accounting degree,\" the man said. \"But mainly, I'm looking for someone to do my worrying for me.\" \n\n\"Excuse me?\" the accountant said. \n\n\"I worry about a lot of things,\" the man said. \"But I don't want to have to worry about money.\nYour job will be to take all the money worries off my back.\" \n\n\"I see,\" the accountant said. \"And how much does the job pay?\" \n\n\"I'll start you at eighty thousand.\" \n\n\"Eighty thousand dollars!\" the accountant exclaimed. \"How can such a small business afford a sum like that?\" \n\n\"That,\" the owner said, \"is your first worry.\"[/hide]",
"Solution_6": "[b]Tuesday Riddle:Why was six afraid of seven? (easy one!)[/b]\r\n[hide=\"Answer\"]Because seven ate nine![/hide]",
"Solution_7": "[hide=\" clicky-->:rotfl: \"]An inmate on death row was scheduled to be put to death by firing squad the following morning. One of the prison guards asked the inmate if he wanted something special for his last meal. The inmate declined the offer. Later, the prison guard asked the inmate if there was something special he wanted to do on his final day. Again, the inmate declined the offer. The following morning, as he inmate was being put before the firing squad, the guard asked him if he wanted a cigarette and a blindfold. \"No,\" the inmate said, \"just get it over with.\" \n\n\"Well, is there anything that I can do for you before you go?\" asked the guard. The inmate thought for a moment, then replied, \"Actually, music is my life. One thing I would really like would be to sing my favorite song, from beginning to end, without any interruptions.\" The guard nodded and agreed. \"Go ahead,\" said the guard. The inmate started, \"One billion bottles of beer on the wall...\"\n\n\n[/hide]",
"Solution_8": "lol, but after he sang down to 1 ..did he die??lol",
"Solution_9": "k00lperson, the point of the song was to delay his death. it takes a life time to count to 1 billion. so, he probably died halfway through it from old age[hide=\".\"]what did the webkinz say at Christmas( P.S. you have to know about webkinz to answer this question.)[/hide] \nthere are two hide tags. one is miniminiminiminiminimini. who cares? [hide=\"Cheater!\"] Answer:\nI hope Santakinz brought me a lot of loot!!!! :blush: [/hide]",
"Solution_10": "[hide=\"another one\"]When Albert Einstein was making the rounds of the speaker's circuit, he usually found himself eagerly longing to get back to his laboratory work. One night as they were driving to yet another rubber-chicken dinner, Einstein mentioned to his chauffeur (a man who somewhat resembled Einstein in looks & manner) that he was tired of speechmaking. \n\n\"I have an idea, boss,\" his chauffeur said. \"I've heard you give this speech so many times. I'll bet I could give it for you.\" Einstein laughed loudly and said, \"Why not? Let's do it!\" When they arrive at the dinner, Einstein donned the chauffeur's cap and jacket and sat in the back of the room. The chauffeur gave a beautiful rendition of Einstein's speech and even answered a few questions expertly. \n\nThen a supremely pompous professor asked an extremely esoteric question about anti-matter formation, digressing here and there to let everyone in the audience know that he was nobody's fool. Without missing a beat, the chauffeur fixed the professor with a steely stare and said, \"Sir, the answer to that question is so simple that I will let my chauffeur, who is sitting in the back, answer it for me.\"[/hide]",
"Solution_11": "[b]Wednesday Riddle:What can fill a room but takes up no space?[/b]\r\n[hide=\"Answer!\"]Light! :P :wink: [/hide]",
"Solution_12": "your wrong k00l. think about shadows. a good answer would be matter.\r\n\r\nwhat do you call this?\r\n\r\nfce\r\n\r\n[hide=\"look behind you!\"]when a blind fish meets a blind mouse.[/hide]",
"Solution_13": "[b]Thursday Riddle:What walks all day on its head?[/b]\r\n[hide=\"Answer\"]A nail in a horse shoe![/hide]",
"Solution_14": "A champion jumps over a castle and lands on a priest. The priest is then dragged of the landscape. What is happening here?\r\n\r\n[hide=\"Answer\"]\nA chess game!! champion-knight, castle-rook, priest-bishop.[/hide]",
"Solution_15": "Happy New Year! \r\n[b]Friday Riddle:What did the boy squirrel say to the girl squirrel on Valentine's day?[/b]\r\n[hide=\"Answer..\"]I'm nuts about you! :P[/hide]"
}
{
"Tag": [
"geometry",
"geometry proposed"
],
"Problem": "Three given circles have the same radius and pass through a common point $ P$. Their other points of pairwise intersections are $ A$, $ B$, $ C$. We define triangle $ A'B'C'$, each of whose sides is tangent to two of the three circles. The three circles are contained in $ \\triangle A'B'C'$. Prove that the area of $ \\triangle A'B'C'$ is at least nine times the area of $ \\triangle ABC$",
"Solution_1": "Hint: Apply Jensen's inequality.",
"Solution_2": "?? Anyone?"
}
{
"Tag": [],
"Problem": "Maria boarded a plane from the airport on the 10th of august . The airport clock showed that it was Saturday shortly after 6 pm\r\nwhen Maria returned the hand of the day clock had moved 135 degree from its position when Maria left the airport.\r\nHow many hours have passed since Maria boarded the plane on Saturday.\r\nans 63.",
"Solution_1": "Well since the day clock moved $ 135$ degrees, it has moved $ \\frac{135}{360}$ of the total week. Thus the number of days that have moved on is $ \\frac{135}{360} \\cdot 7 \\equal{} 2.625$.\r\n\r\nThis is two whole days and $ \\frac{5}{8}$ths of a day, so this is $ 2 \\cdot 24 \\plus{} \\frac{5}{8} \\cdot 24 \\equal{} 48\\plus{}15\\equal{}63$ hours. (This is assuming that less than a week passed since she borded the plane.)\r\n\r\nHope this helped.\r\n\r\nPS - Notice that the time Maria boarded the plane ins't important at all - all we need to know about is the day clock."
}
{
"Tag": [
"analytic geometry"
],
"Problem": "I have 3 points A, B, and C located in a 3D plan.\r\nThe x, y, and z coordinates of point A and B are known and also the distance between points A, B, and C is known as well (81 feet).\r\nWhat I want to do is to determine the x, y, and z coordinates of point C.",
"Solution_1": "So you mean $d\\left(A,B\\right)=d\\left(A,C\\right)=d\\left(B,C\\right)=81$, right? In that case, you have $A$ and $B$ fixed, but there are an infinite amount of possibilities for $C$ which satisfy this. The possibilities for $C$ actually turn out to form a circle.",
"Solution_2": "A circle centered at the midpoint of $AB$ and having a radius of $\\frac{81\\sqrt3}{2}$ as a matter of fact..."
}
{
"Tag": [
"combinatorics unsolved",
"combinatorics"
],
"Problem": "The Dwarfs in the Land-under-the-Mountain have just adopted a completely new decimal currency system based on the Pippin, with gold coins to the value of $ 1$ Pippin, $ 10$ Pippins, $ 100$ Pippins and $ 1000$ Pippins.\r\n\r\nIn how many ways is it possible for a Dwarf to pay, in exact coinage, a bill of $ 1997$ Pippins?",
"Solution_1": "Isn't this more of a informatics problem, more particular dynamic programming?",
"Solution_2": "Hi Andrew;\r\n\r\nThis is a standard partition problem, easy to set up but hard to compute.\r\n\r\n$ \\frac {1}{(1 \\minus{} x)(1 \\minus{} x^{10})(1 \\minus{} x^{100})(1 \\minus{} x^{1000})}$\r\n\r\nand your looking for the coefficient of x^1997, or\r\n\r\n$ (1 \\plus{} x \\plus{} x^2 \\plus{} x^3 \\plus{} ... \\plus{} )(1 \\plus{} x^{10} \\plus{} x^{20} \\plus{} ... \\plus{} )(1 \\plus{} x^{100} \\plus{} x^{200} \\plus{} x^{300} \\plus{} ... \\plus{} )(1 \\plus{} x^{1000})$\r\n\r\nand your looking for the coefficient of x^1997. I expanded it with a computer to get that coefficient and it is 2650. So there 2650 ways to make 1997 in pippen currency."
}
{
"Tag": [
"abstract algebra",
"group theory",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Let $ N$ and $ H$ two groups, $ \\varphi$ and $ \\psi$ two morphisms $ H \\to \\mathrm{Aut}(N)$.\r\nIf there exist $ u \\in \\mathrm{Aut}(N)$ such that $ \\psi(h) \\equal{} u \\circ \\varphi(h) \\circ u^{\\minus{}1}$, or if there exist $ v \\in \\mathrm{Aut}(H)$ such that $ \\varphi \\equal{} \\psi \\circ v$, then $ N \\rtimes_{\\varphi} H \\simeq N \\rtimes_{\\psi} H$.\r\nSo these are two suffisiant conditions for semidirect products to be isomorphic.\r\n\r\nI am looking for necessary conditions. In particular, I am interested to find conditions so that if $ N$ and $ H$ are two [b]finite[/b] groups there is a unique (up to isomorphism) semidirect product $ N \\rtimes H$ that is not direct.",
"Solution_1": "Nobody interested?",
"Solution_2": "This question is interesting but I think it is too difficult (or too general).\r\nEven for $ N\\equal{}\\mathbb{Z}/p^2\\mathbb{Z}\\times \\mathbb{Z}/p\\mathbb{Z}$ and $ H\\equal{}\\mathbb{Z}/p\\mathbb{Z}$, with $ p>2$ prime, I think there are 3 non isomorphic, non abelian groups of the form $ N\\rtimes H$.\r\nShowing that two semidirect products are non isomorphic is often a difficult task.",
"Solution_3": "[quote=\"JC_math\"]This question is interesting but I think it is too difficult (or too general).\nEven for $ N \\equal{} \\mathbb{Z}/p^2\\mathbb{Z}\\times \\mathbb{Z}/p\\mathbb{Z}$ and $ H \\equal{} \\mathbb{Z}/p\\mathbb{Z}$, with $ p > 2$ prime, I think there are 3 non isomorphic, non abelian groups of the form $ N\\rtimes H$.\nShowing that two semidirect products are non isomorphic is often a difficult task.[/quote]\r\n\r\nFor what it is worth:\r\n\r\nFor p=2, there are three non-abelian groups of order 2^4 with abelian normal subgroups of type 4 \u00d7 2 that are complemented. For p\u22653, there are five non-abelian groups of order p^4 with abelian normal subgroups of type p^2 \u00d7 p that are complemented.\r\n\r\nIn other words, your claim is true for p=2 too, and for p>2 it is even more true than you claimed. The proof is just consulting the list of groups of order p^4.",
"Solution_4": "Indeed JackSchmidt.\r\nDo you think the initial question from Portland is \"solvable\"?",
"Solution_5": "[quote=\"JC_math\"]Do you think the initial question from Portland is \"solvable\"?[/quote]\r\n\r\nI don't normally look at solving random questions like this. First I look to see if there are any interesting special cases or motivation. I don't see any interesting special cases or motivation for this question.\r\n\r\nSemidirect products can be isomorphic for random reasons. Even listing the ones where some isomorphism takes N to N does not seem too interesting. When H is simple, and N is elementary abelian, this might be a question about the representation theory of H, but even that question seems contrived (some sort of postage stamp problem where the denominations available are the dimensions of the non-principal irreducible representations).\r\n\r\nAnother aspect of the problem that strikes me as contrived is trying to ignore direct products as if they are not semidirect. For instance, what about H=A5, N=A5? There are up to conjugacy exactly two actions of H on N, the trivial action giving the direct product, and the conjugation action, in which H definitely does not centralize N. I would not think that one would call the semidirect product formed from the second action the direct product, but on the other hand there is an isomorphism (centralizing N) between it and H \u00d7 N. How many semidirect products that are not direct does the pair (H,N) have?",
"Solution_6": "Thank you JackSchmidt and JC_Math.\r\n\r\nIn relation to \r\n[quote=\"JackSchmidt\"]First I look to see if there are any interesting special cases or motivation. I don't see any interesting special cases or motivation for this question...[/quote]\r\nThe motivation was simple, when you find one or several \"sufficient conditions\", it is fairly natrual to ask wheter or not we know some \"necessary conditions\", don't you think?\r\nI thought about the question before posting, and because it seemed fairly difficult (at least to me) in the generic case, I reduced it to finite groups, in hope that it would be simpler. The example given by JC_Math convinced me that even for finite (even abelian) groups this is probably too difficult a question.",
"Solution_7": "JackSchmidt,\r\n\r\nLike you I normally don't answer random questions.\r\nI asked myself a question not far from the one at the beginning of that thread a while back, which is why I knew about the characterization of groups of order $ p^4$. \r\nIndeed, \"this might be a question about the representation theory of H\", or about wreath product of cyclic groups (to be used in representation theory:)), which was my case."
}
{
"Tag": [
"LaTeX",
"inequalities",
"inequalities proposed"
],
"Problem": "Just look at this one :\r\n\r\n If for the positive real x,y,z is true xy+yz+xz=2 then provethe following enequality:\r\n\r\n [b] symsum[{xy}^2]+symsum[{xyz}^2]>= symsum[(x^3)(y^2)z]+symsum[(x^2)yz][/b]",
"Solution_1": "Can you put it into Latex? ;)",
"Solution_2": "If $x,y,z >0$ and $xy + yz + zx = 2,$ show that \\[ \\sum x^2y^2 + \\sum x^2y^2z^2 \\geq \\sum x^3y^2z + \\sum x^2yz. \\]",
"Solution_3": "[quote=\"Arne\"]If $x,y,z >0$ and $xy + yz + zx = 2,$ show that \\[ \\sum x^2y^2 + \\sum x^2y^2z^2 \\geq \\sum x^3y^2z + \\sum x^2yz. \\][/quote]\r\nThe inequality is equivalent to $\\dfrac{1}{2}(\\sum x^3y^3+\\sum_{sym}x^3y^2z)+3x^2y^2z^2\\ge \\sum_{sym}x^3y^2z+\\dfrac{1}{2}(\\sum_{sym}x^3y^2z+3x^2y^2z^2)\\Leftrightarrow\\sum x^3y^3+3x^2y^2z^2\\ge 2\\sum_{sym}x^3y^2z$ that is incorrect for $x=y=z$ but correct the version $\\sum x^3y^3+3x^2y^2z^2\\ge \\sum_{sym}x^3y^2z$ which is just Shur for substitution $xy=a,yz=b,zx=c$",
"Solution_4": "[quote=\"amanin\"][quote=\"Arne\"]If $x,y,z >0$ and $xy + yz + zx = 2,$ show that \\[ \\sum x^2y^2 + \\sum x^2y^2z^2 \\geq \\sum x^3y^2z + \\sum x^2yz. \\][/quote]\nThe inequality is equivalent to $\\frac{1}{2}(\\sum x^3y^3+\\sum_{sym}x^3y^2z)+3x^2y^2z^2\\ge \\sum_{sym}x^3y^2z+\\frac{1}{2}(\\sum_{sym}x^3y^2z+3x^2y^2z^2)\\Leftrightarrow\\sum x^3y^3+3x^2y^2z^2\\ge 2\\sum_{sym}x^3y^2z$ that is incorrect for $x=y=z$ but correct the version $\\sum x^3y^3+3x^2y^2z^2\\ge \\sum_{sym}x^3y^2z$ which is just Shur for substitution $xy=a,yz=b,zx=c$[/quote]\r\nwhy the inequality is incorrect for x=y=z?",
"Solution_5": "After the substitution a=xy, b= xz, c=yz we get using the equality a+b+c=2 and\r\n\r\n[b]$a^3+b^3+c^3+3abc=2(a^2+b^2+c^2-ab-ac-bc)+6abc$[/b]\r\n\r\na shur's inequaliti and the inequality is proved"
}
{
"Tag": [],
"Problem": "I can type the word munchkins in 72 minutes. How many minutes will it take for me to type the word floccinoccinihilapilaphacation?\r\n :starwars:",
"Solution_1": "Assuming constant rate of typing per letter, it takes $ 8$ minutes to type one letter. There are $ 30$ letters, so $ 8 \\times 30\\equal{}\\boxed{240 \\text{ minutes}}$.",
"Solution_2": "Since you can type 8 min per one letter. there are thirty letters in the word floccinoccinihilapilaphacation so $ 8\\times 30\\equal{}\\boxed{240\\text{ minutes}}$. I did not look at the answer.",
"Solution_3": "dude, don't post a solution unless it's different from the person above or u think the guy above (AIME15 :P ) is wrong!\r\n\r\nEDIT: do u really think AIME15 is wrong?",
"Solution_4": "according to my calculation I am right and thanks for the tip :furious: :no: :cursing:",
"Solution_5": "munchkins has 9 letters \r\n72 divided by nine is 8\r\nFlo-blabla-somtin-somtin-tion has got 30 letters\r\n8x30=240 min or 4 hours\r\n\r\n\r\nps: the problem has 130 character (with spaces) so it took you 17 hours and 20 minutes to write this. :sleep2:",
"Solution_6": "I think we already know the answer and there is no need for anymore solutions."
}
{
"Tag": [
"probability",
"combinatorics unsolved",
"combinatorics"
],
"Problem": "Hi every body \r\ncould any one help me to solve this problem\r\ni did not know where to write the topic\r\n[img]http://img409.imageshack.us/img409/4158/22rw1.jpg[/img]\r\nREGARDS\r\nPlease help me",
"Solution_1": "For the second question: \r\nThe probability of rolling a $6$ is $\\frac{1}{6}$ and the probability of rolling anything but a $6$ is $\\frac{5}{6}$. So for the first part of that question there are $\\binom{4}{1}= 4$ ways of getting exactly one $6$. So we have $4\\cdot\\left( \\frac{1}{6}\\right) \\cdot\\left(\\frac{5}{6}\\right)^{3}$. That will give you the probability of exactly one - see if you can use that to figure out the prob of at least one.",
"Solution_2": "For the first question\r\nP(1 works | system is functioning)\r\n=P(1 works and system is functioning) / P(system is functioning)\r\n=1/2 * (1 - (1/2)^(n-1)) / (1- (1/2)^(n-1))"
}
{
"Tag": [
"inequalities",
"vector",
"linear algebra",
"matrix",
"linear algebra solved"
],
"Problem": "Let $P,Q,R \\in M_n(K)$. Prove that $rang(PQ)+rang(QR) \\leq rang(Q)+rang(PQR)$",
"Solution_1": "what is 'rang'?\r\nRange or Rank?",
"Solution_2": "[b]rang[/b] it means rank",
"Solution_3": "It is Frobenius inequality:\r\nFirstly let's prove useful inequality,considering our matrices as mapping on vector space over $K$.Let $U \\subset V$ and $X$:$V \\to W $. Then $dim(KerX|_{U})\\leq dimKerX=dimV-dimImX$ (1). If we use inequality (1) for the case $U=ImQR$,$V=ImQ$,$X=P$ we obtain that $dim(KerP|_{ImQR})\\leq dimImQ-dimImPQ$ (2). On the other hand it is readily that $dim(KerP|_{ImQR})=dimImQR-dimImPQR$ (3). So, applying (2) and (3) we get our inequality",
"Solution_4": "A different approach.\n\n\\[\n\\begin{pmatrix}\nPQ & 0\\\\\nQ & QR\n\\end{pmatrix}\n\\]\nMultiply the first column by $-R$ to the right and add it to the second column, then multiply the second column by $-1$,\n\\[\n\\begin{pmatrix}\nPQ & PQR\\\\\nQ & 0\n\\end{pmatrix}\n\\]\nMultiply the second row by $-P$ to the left, and add it to the first row\n\\[\n\\begin{pmatrix}\n0 & PQR\\\\\nQ & 0\n\\end{pmatrix}\n\\]\n\n$rank(PQ)+rank(QR)=rank\n\\begin{pmatrix}\nPQ & 0\\\\\n0 & QR\n\\end{pmatrix}\n\\leq rank\n\\begin{pmatrix}\nPQ & 0\\\\\nQ & QR\n\\end{pmatrix}\n=rank\n\\begin{pmatrix}\n0 & PQR\\\\\nQ & 0\n\\end{pmatrix}\n=rank(PQR)+rank(Q)\n$\n\nwhere $rank(M)=rang(M)=r(M)$ denotes the rank of a matrix $M$."
}
{
"Tag": [
"articles"
],
"Problem": "Right, here is the news story.\r\nhttp://news.yahoo.com/s/ap/20070410/ap_on_sp_co_ne/bkc_texas_durant\r\nComments?",
"Solution_1": "Yeah, who's Kevin Durant? :D",
"Solution_2": "lol\r\nhes a really good freshman at UT",
"Solution_3": "Expected.\r\n\r\nHe is the next Kevin Garnett (but with a better shooting touch)\r\n\r\nUT could mean University of Tennessee...\r\nbut yeah\r\nHe's a spectacular freshman prospect from the University of Texas.\r\n\r\nA great guy that's going to wow lots of people.",
"Solution_4": "Anyone think that he's going to get picked first in the draft?\r\nI think it's highly reasonable, but it definitely depends on the team.",
"Solution_5": "Good riddance.",
"Solution_6": "Alot of teams will want this kid\r\nHe's a swingman (means he can play Shooting Guard and Small Forward) and could play Power Forward if he develops his body into it.\r\nOnly teams stacked at Shooting Guard, Small Forward, and Power Forward won't pick him\r\nTeams like:\r\nDallas Mavericks (Jason Terry, Josh Howard, Jerry Stackhouse, Dirk Nowitzki)\r\nDetroit Pistons (Richard Hamilton, Tashyuan (sp?) Prince, Rasheed Wallace, Chris Webber, Antonio McDeyss (sp?)\r\nChicago Bulls (Ben Gordon, Luol Deng, Tyrus Thomas, P.J. Brown)\r\nSan Antonio (Manu Ginobli, Bruce Bowen, Tim Duncan)\r\naren't going to pick Durant\r\nbut these teams are all Playoff Contenders\r\n\r\nso yeah\r\nHE WILL GO NUMBER ONE! :D",
"Solution_7": "Agreed. All the teams with decent big men, minus center, are strong playoff contenders... and all the teams headed for the draft are just the opposite. So yeah, Kevin Durant for #1.",
"Solution_8": "this should go to G&FF\r\n\r\n[i][Moderator says: So it is written, so it shall be done.][/i]",
"Solution_9": "Oden will go first. He has much more franchise value than Durant. Good centers are hard to come by, and swingmen aren't. Even if Durant is better, he won't get picked first based on how much he can improve a team over Oden.",
"Solution_10": "[quote=\"besttate\"]Oden will go first. He has much more franchise value than Durant. Good centers are hard to come by, and swingmen aren't. Even if Durant is better, he won't get picked first based on how much he can improve a team over Oden.[/quote]\r\n\r\nAre you even a sports fan?\r\nOden will have to develop his game\r\nHe gets WAY TOO many fouls during games\r\nThis renders him useless in the NBA\r\nGood centers will not win games\r\nSan Antonio Spurs don't have a good center\r\nMiami has an old center that's now a shadow of what he use to be\r\nDetroit had a undrafted Center\r\nOden will never equal Shaq's dominance\r\nSwingman can easily improve teams\r\nCenters take TOO LONG to develop\r\n\r\nDurant's franchise value > Oden's franchise value",
"Solution_11": "I agree with besttate. I think Durant is more athletic and a better fit for the NBA, but that shouldn't take away any credit from Oden.\r\nPlus, teams like the Celtics or the Hawks who might get top-3 lotto picks could definitely use Oden.\r\nOn a side note, DJ Augustin went to my school last year, and man, he was hoss. :) People went to the games just to see him play (he was an All-American!). Our basketball team is consistently good, so I'm surprised we didn't get further in the playoffs last year. Wonder where he will end up in 3 years! Anyway, go UT! w00t.",
"Solution_12": "eh...dont blame really good players to go pro early. same with lendale white and a bunch of other football players.",
"Solution_13": "Oden should stay in college to refine his skillz",
"Solution_14": "[quote=\"mustafa\"]Oden should stay in college to refine his skillz[/quote]\r\n\r\nlol guess you didnt see Oden in the championship game-and he was playing against two lottery picks and another possible NBA player...\r\n\r\nplus yea big guys are hard to come by, even if Durant is more talented than Oden, you still have to go with Oden because its pretty rare to find a seven footer with that much talent, as Oden can change a game without even touching the ball-he will draw double teams that can free up his teamates and hes a lot better defensive player than Durant, as he will make people change how they will shoot the ball when driving for a layup because those people dont want to get the shot blocked (if you look at the history of the NBA draft, most of the first overall picks have been guy over 6'10\"-Bariangi (sp?), Bogut, Yao, Kwame Brown, Shaq, Elton Brand, Hakeem, Ewing.. and so one..not all of them have worked out but Oden is a lot better than people like Kwame Brown and stuff..)",
"Solution_15": "[quote=\"Art of Owna\"][quote=\"mustafa\"]Oden should stay in college to refine his skillz[/quote]\n\nlol guess you didnt see Oden in the championship game-and he was playing against two lottery picks and another possible NBA player...\n\nplus yea big guys are hard to come by, even if Durant is more talented than Oden, you still have to go with Oden because its pretty rare to find a seven footer with that much talent, as Oden can change a game without even touching the ball-he will draw double teams that can free up his teamates and hes a lot better defensive player than Durant, as he will make people change how they will shoot the ball when driving for a layup because those people dont want to get the shot blocked (if you look at the history of the NBA draft, most of the first overall picks have been guy over 6'10\"-Bariangi (sp?), Bogut, Yao, Kwame Brown, Shaq, Elton Brand, Hakeem, Ewing.. and so one..not all of them have worked out but Oden is a lot better than people like Kwame Brown and stuff..)[/quote]\r\n\r\n1) Oden was playing against Horford and Noah\r\n2) Noah was HORRIBLE in the championship game. Noah will never survive in the NBA\r\n3) Horford is undersized\r\n3) Oden gets in too much foul trouble so the double teams won't come that often and the block shots won't come often\r\n4) Oden will never be able to match up with other playoff team big men and make a difference.\r\n5) Durant plays decent defense\r\n\r\nIf Oden stays in college, he will become a better player than durant\r\nbut that's only if\r\n\r\nDurant will become remarkable player",
"Solution_16": "[quote=\"Aznwithabrain\"][quote=\"Art of Owna\"][quote=\"mustafa\"]Oden should stay in college to refine his skillz[/quote]\n\nlol guess you didnt see Oden in the championship game-and he was playing against two lottery picks and another possible NBA player...\n\nplus yea big guys are hard to come by, even if Durant is more talented than Oden, you still have to go with Oden because its pretty rare to find a seven footer with that much talent, as Oden can change a game without even touching the ball-he will draw double teams that can free up his teamates and hes a lot better defensive player than Durant, as he will make people change how they will shoot the ball when driving for a layup because those people dont want to get the shot blocked (if you look at the history of the NBA draft, most of the first overall picks have been guy over 6'10\"-Bariangi (sp?), Bogut, Yao, Kwame Brown, Shaq, Elton Brand, Hakeem, Ewing.. and so one..not all of them have worked out but Oden is a lot better than people like Kwame Brown and stuff..)[/quote]\n\n1) Oden was playing against Horford and Noah\n2) Noah was HORRIBLE in the championship game. Noah will never survive in the NBA\n3) Horford is undersized\n3) Oden gets in too much foul trouble so the double teams won't come that often and the block shots won't come often\n4) Oden will never be able to match up with other playoff team big men and make a difference.\n5) Durant plays decent defense\n\nIf Oden stays in college, he will become a better player than durant\nbut that's only if\n\nDurant will become remarkable player[/quote]\r\n\r\nBut Oden is leaving for the NBA (announced today), but hey look at Hakeem Olajuwan..he was pretty raw (just came from Nigeria where he never played basketball) like Oden and he would also get in foul trouble and didnt have that much stamina....when he came out for the NBA draft in 1984..hes one of the greatest centers 25 year later looking back on it (voted one of the top 50 NBA greatest players), but then again Michael Jordan was the 3rd pick that draft...hmm could Kevin Durant be the MJ of this draft..",
"Solution_17": "[quote=\"Art of Owna\"][/quote][quote=\"Aznwithabrain\"][quote=\"Art of Owna\"][quote=\"mustafa\"]Oden should stay in college to refine his skillz[/quote]\n\nlol guess you didnt see Oden in the championship game-and he was playing against two lottery picks and another possible NBA player...\n\nplus yea big guys are hard to come by, even if Durant is more talented than Oden, you still have to go with Oden because its pretty rare to find a seven footer with that much talent, as Oden can change a game without even touching the ball-he will draw double teams that can free up his teamates and hes a lot better defensive player than Durant, as he will make people change how they will shoot the ball when driving for a layup because those people dont want to get the shot blocked (if you look at the history of the NBA draft, most of the first overall picks have been guy over 6'10\"-Bariangi (sp?), Bogut, Yao, Kwame Brown, Shaq, Elton Brand, Hakeem, Ewing.. and so one..not all of them have worked out but Oden is a lot better than people like Kwame Brown and stuff..)[/quote]\n\n1) Oden was playing against Horford and Noah\n2) Noah was HORRIBLE in the championship game. Noah will never survive in the NBA\n3) Horford is undersized\n3) Oden gets in too much foul trouble so the double teams won't come that often and the block shots won't come often\n4) Oden will never be able to match up with other playoff team big men and make a difference.\n5) Durant plays decent defense\n\nIf Oden stays in college, he will become a better player than durant\nbut that's only if\n\nDurant will become remarkable player[/quote][quote=\"Art of Owna\"]\n\nBut Oden is leaving for the NBA (announced today), but hey look at Hakeem Olajuwan..he was pretty raw (just came from Nigeria where he never played basketball) like Oden and he would also get in foul trouble and didnt have that much stamina....when he came out for the NBA draft in 1984..hes one of the greatest centers 25 year later looking back on it (voted one of the top 50 NBA greatest players), but then again Michael Jordan was the 3rd pick that draft...hmm could Kevin Durant be the MJ of this draft..[/quote]\r\n\r\nYou do make some good points like:\r\nOden's foul troubles are similiar to Hakeem Oajuwan\r\nDurant could be a bit similiar to Jordan\r\n\r\nI agree with 3/4 of it\r\nbut\r\nI'm pretty sure Olajuwan didn't come out of college in his freshman year\r\nand\r\nI'm certain Olajuwan developed his game in college\r\nI remember reading numerous articles of him practicing in gyms with Houston Rocket players in collge\r\nI think he came out in his junior or senior year\r\n\r\nDon't forget:\r\nJordan came out as a Junior (certain)\r\nNone of those two came out as freshman",
"Solution_18": "7 years later durant second draft pick and oden first draft pick durant went to the finals already and has a mvp award",
"Solution_19": "Don't bump seven year old threads... that is an example of how NEVER to use the forums..."
}
{
"Tag": [
"calculus",
"integration",
"calculus computations"
],
"Problem": "Evaluate\r\n\r\n\\[ \\int \\frac{x^2 e^{\\arctan x}}{\\sqrt{1\\plus{}x^2}}\\,dx\\]",
"Solution_1": "I first differentiated \r\n\r\n$ \\sqrt{1\\plus{}x^{2}}e^{\\arctan x }$ to get $ \\frac{(x\\plus{}1)e^{\\arctan x}}{\\sqrt{1\\plus{}x^{2}}}$\r\n\r\nand then\r\n\r\n$ x\\sqrt{1\\plus{}x^{2}}e^{\\arctan x}$ to get $ \\frac{(2x^{2} \\plus{} x \\plus{} 1)e^{\\arctan x}}{\\sqrt{1\\plus{}x^{2}}}$.\r\n\r\nSubtracting, these we have\r\n\r\n$ \\frac{2x^{2}e^{\\arctan x}}{\\sqrt{1\\plus{}x^{2}}}$\r\n\r\nand so $ \\int \\frac{x^{2}e^{\\arctan x}}{\\sqrt {1\\plus{}x^{2}}} dx \\equal{} \\frac{ (x\\minus{}1)\\sqrt{1\\plus{}x^{2}}e^{\\arctan x}}{2} \\plus{}c$."
}
{
"Tag": [],
"Problem": "Given $ 9^6 \\equal{} 531441$, how would you represent $ 531440$ in base 9?\r\n\r\n[hide=\"Answer\"][b]Answer:[/b] $ 888888_9$[/hide]\r\nI thought I had this one.. :(",
"Solution_1": "just like 10^3=1000, 1000-1=999. \r\n\r\nSo, when ur in base 9, one less then 9^6 should equal 88888. \r\n\r\nNot sure how good the explanation is...",
"Solution_2": "[hide]\n$ 531441\\equal{}1000000_{9}$\n$ 531441\\equal{}1000000_{9}\\minus{}1_{9}$\nPerforming the subtraction, we get $ 888888_{9}$.\n[/hide]",
"Solution_3": "Okay. I got it. Thanks",
"Solution_4": "[quote=\"Quickster94\"]just like 10^3=1000, 1000-1=999. \n\nSo, when ur in base 9, one less then 9^6 should equal 88888. \n\nNot sure how good the explanation is...[/quote]\r\n\r\nIt should have 6 8s; I'm not sure if that's a typo or not..."
}
{
"Tag": [
"calculus",
"limit",
"trigonometry",
"function",
"derivative",
"complex analysis",
"calculus computations"
],
"Problem": "And this? how calculate?\r\n\r\n$ \\mathop {\\lim }\\limits_{x \\to 0} \\left( {\\frac{1}{x} \\minus{} \\frac{1}{{\\cos x \\cdot senx}}} \\right)$\r\n\r\nThanks\r\n\r\nAntonio",
"Solution_1": "$ \\lim_{x\\to 0} \\left ( \\frac{1}{x} \\minus{} \\frac{1}{\\sin x\\cos x} \\right )$\r\n$ \\equal{}\\lim_{x\\to 0} \\left ( \\frac{\\sin x\\cos x}{x\\sin x\\cos x} \\minus{} \\frac{1}{\\sin x\\cos x} \\right )$\r\nUse $ \\lim_{x\\to 0} \\frac{\\sin x}{x}\\equal{}1$\r\n$ \\equal{}\\lim_{x\\to 0} \\left ( \\frac{\\cos x\\minus{}1}{\\sin x\\cos x} \\right )$\r\nMultiply by $ \\frac{\\cos x\\plus{}1}{\\cos x\\plus{}1}$\r\n$ \\equal{}\\lim_{x\\to 0} \\left ( \\frac{\\cos ^2x\\minus{}1}{\\sin x\\cos x(\\cos x\\plus{}1)} \\right )$\r\n$ \\equal{}\\lim_{x\\to 0} \\left ( \\frac{\\minus{}\\sin ^2x}{\\sin x\\cos x(\\cos x\\plus{}1)} \\right )$\r\n$ \\equal{}\\lim_{x\\to 0} \\left ( \\frac{\\minus{}\\sin x}{\\cos x(\\cos x\\plus{}1)} \\right )$\r\n$ \\equal{}\\frac{0}{1(2)}$\r\n$ \\equal{}0$",
"Solution_2": "[quote=\"EastyMoryan\"]$ \\lim_{x\\to 0} \\left ( \\frac {1}{x} \\minus{} \\frac {1}{\\sin x\\cos x} \\right )$\n$ \\equal{} \\lim_{x\\to 0} \\left ( \\frac {\\sin x\\cos x}{x\\sin x\\cos x} \\minus{} \\frac {1}{\\sin x\\cos x} \\right )$\nUse $ \\lim_{x\\to 0} \\frac {\\sin x}{x} \\equal{} 1$\n$ \\equal{} \\lim_{x\\to 0} \\left ( \\frac {\\cos x \\minus{} 1}{\\sin x\\cos x} \\right )$\n. \n.\n.\n.\n[/quote]\r\n\r\nI have a doubt in this ... [b]when you apply the limit, shouldn't it be done for the entire function and not just a part of the function[/b]?",
"Solution_3": "This is a wonderful problem for the Taylor series approach: $ \\frac {1}{\\cos x} \\equal{} 1 \\plus{} O(x^2)$. $ \\frac {x}{\\sin x} \\equal{} 1 \\plus{} O(x^2)$. So we have $ \\lim_{x \\to 0} \\frac {1}{x} \\minus{} \\left(\\frac {1}{x} \\plus{} O(x)\\right) \\equal{} \\lim_{x \\to 0} O(x) \\equal{} 0.$\r\n\r\nIf you replace $ \\sec x$ with a function $ f$ such that $ f(0) \\equal{} 1$ and $ f'(0) \\equal{} a$ (I don't think I need any more niceness for $ f$ than the existence of this derivative at 0 -- is that right?) then the answer becomes $ \\minus{}a$.",
"Solution_4": "Can you please explain how is $ \\frac {1}{\\cos x} \\equal{} 1 \\plus{} O(x^2)$? Sorry I m weak in this kind of approximations.",
"Solution_5": "The linear approximation (degree-1 Taylor series approximation) of $ \\sec x$ at $ x \\equal{} 0$ is $ \\sec 0 \\plus{} \\sec' 0 x$, and this is just equal to 1, so $ \\sec x \\equal{} 1 \\plus{} O(x^2)$.",
"Solution_6": "$ \\equal{} \\lim_{x\\to 0}\\left (\\frac {\\sin x\\cos x}{x\\sin x\\cos x} \\minus{} \\frac {1}{\\sin x\\cos x}\\right )$\r\n$ \\equal{} \\lim_{x\\to 0}\\left (\\frac {\\sin x\\cos x}{x\\sin x\\cos x} \\minus{} \\frac {1}{\\sin x\\cos x}(1)\\right )$\r\n$ \\equal{} \\lim_{x\\to 0}\\left [\\frac {\\sin x\\cos x}{x\\sin x\\cos x} \\minus{} \\frac {1}{\\sin x\\cos x}\\left (\\lim_{x\\to 0}\\frac {\\sin x}{x}\\right )\\right ]$\r\n$ \\equal{} \\lim_{x\\to 0}\\left (\\frac {\\sin x}{x}\\right )\\left (\\frac {\\cos x \\minus{} 1}{\\sin x\\cos x}\\right )$\r\n$ \\equal{} \\lim_{x\\to 0}\\left (\\frac {\\cos x \\minus{} 1}{\\sin x\\cos x}\\right )$\r\n\r\nWe could also use L'Hopital's Rule:\r\n\r\n$ \\lim_{x\\to 0} \\frac {\\sin x\\cos x \\minus{} x}{x\\sin x\\cos x}$\r\n$ \\equal{} \\lim_{x\\to 0} \\left (\\frac {\\sin 2x \\minus{} 2x}{x\\sin 2x}\\right )$\r\nHere:\r\n$ \\equal{} \\lim_{x\\to 0} \\left (\\frac {2\\cos 2x \\minus{} 2}{2x\\cos 2x \\plus{} \\sin 2x}\\right )$\r\nAnd here:\r\n$ \\equal{} \\lim_{x\\to 0} \\left (\\frac { \\minus{} 4sin2x}{2(\\cos 2x \\minus{} 2x\\sin 2x) \\plus{} 2\\cos 2x}\\right )$\r\n$ \\equal{} \\left (\\frac {0}{4}\\right )$\r\n$ \\equal{} 0$"
}
{
"Tag": [
"\\/closed"
],
"Problem": "Which online subject classes are extremely useful to take in preparation for the amc 12/aime. I will probably only be able to take about 1-3 subject classes. Also, I will have completed the aops series and part of art and craft of problem solving after.",
"Solution_1": "Go for your weakest areas. That's the general advice.\r\n\r\nI think the Problem Series may help you the most though if your goal is, for example, to qualify for USAMO."
}
{
"Tag": [
"geometry",
"3D geometry",
"pyramid",
"summer program",
"PROMYS"
],
"Problem": "I learned about this game at (of all places) a computer camp in Colorado...\r\n\r\nAnyways, I propose that we play it here, on AoPS. However, I'm not sure how we could do it on a forum, since people may want to join in mid-game, or forget to login for days.\r\n\r\nWhatever.\r\n\r\n[b]Blue[/b]!\r\n\r\nEdit: For those of you who don't know how to play, here are the rules:\r\n-On your turn, you must either say a color or country.\r\n-If you say the name of a country, you lose.",
"Solution_1": "Um... could you... explain how to play?",
"Solution_2": "[color=red]RED![/color]",
"Solution_3": "Oooh, tough one.... hm...\r\n\r\n[color=green]Green[/color]!",
"Solution_4": "If this is anything like Mornington Crescent...then I'm out.",
"Solution_5": "Tajikistan. :D \r\n\r\nOops I am out! :oops:",
"Solution_6": "so we're just naming colors?",
"Solution_7": "Oh are we? Very well then, I'm back in with turquoeise!",
"Solution_8": "crimson......s' that fine??\r\n :huh:",
"Solution_9": "Should be. \r\nOh you have reached hodge conjecture I see.\r\n\r\nAnyway pinkish white.",
"Solution_10": "Egypt.......................ian pyramid tan!\r\n\r\nWhew, close call :)",
"Solution_11": "Maceroni-and-cheese (I've seen it on crayola crayons) :D",
"Solution_12": "blue ",
"Solution_13": "You lose, blue is a country.\r\n\r\n\r\nyellow",
"Solution_14": "[color=darkblue]Deep sea blue![/color]\r\n\r\nI have lost 0 times so far! Wooo!",
"Solution_15": "blue ",
"Solution_16": "cadet brown",
"Solution_17": "blue ",
"Solution_18": "chinese red",
"Solution_19": "[quote=\"Klebian\"]magentaish chartruce[/quote]\r\n\r\nI know this is an old post, but it's chartreuse :D\r\n\r\nThe color of a dead man's face",
"Solution_20": "raspberry!",
"Solution_21": "reddish-(reddish-(reddish-(reddish-purple)))",
"Solution_22": "[quote=\"not_trig\"][quote=\"Klebian\"]magentaish chartruce[/quote]\n\nI know this is an old post, but it's chartreuse :D\n\nThe color of a dead man's face[/quote]\r\n\r\nHaha, now that I think about it, I can't understand why I put a 2nd c in there, I know that at that time I couldn't remember how to spell it, but I know it doesn't have a c, and I've always known that...\r\n\r\nHmm... Europe! Yay!",
"Solution_23": "that's neither a color nor a country. you have to go again.",
"Solution_24": "The colour of Ivory Coast on the page entitled \"Africa - Political\" in the first standard school atlas you can lay your hands on.",
"Solution_25": "The color of the colors of the American flag mixed together!",
"Solution_26": "[quote=\"nutz_for2.718281828\"]The color of the colors of the American flag mixed together![/quote]\r\nThat would make a light purple. Maybe lavendar?",
"Solution_27": "Angola - I like it!",
"Solution_28": "silver?",
"Solution_29": "golden wooden gold!"
}
{
"Tag": [
"geometry",
"incenter",
"rectangle",
"cyclic quadrilateral",
"geometry proposed",
"Angle Chasing"
],
"Problem": "If $ABCD$ is a cyclic quadrilateral, then prove that the incenters of the triangles $ABC$, $BCD$, $CDA$, $DAB$ are the vertices of a rectangle.",
"Solution_1": "By angle chasing. \r\n\r\nYou can show that the incentres of ABC (K), BCD (L), B and C are cyclic because angle CLB = 180 - DCB/2 - DBC/2 = 90 + BDC/2, and similarly, angle CKB = 90 + BAC/2. But BAC = BDC. Thus CLB = CKB and so B,C,K,L are concyclic. Also if we call the incentre of ADC (M), D,C,L,M is cyclic. angle KLC = 180 - angle LBC, angle MLC = 180 - MDC, thus angle KLM = angle KBC + angle MDC = (ADC+ABC)/2 = 90, which is what we want.\r\n\r\nPS: Can someone give me a crashcourse on latex? or point to me where?",
"Solution_2": "It can be proved by complex numbers... See also\r\n\r\n [url=http://www.math.fau.edu/yiu/Geometry.html]Paul Yiu, [i]Notes on Euclidean Geometry[/i][/url], paragraph 10.2.6 (page 154);\r\n [url=http://groups.yahoo.com/group/Hyacinthos/message/8154]Barry Wolk, [i]Hyacinthos message #8154[/i][/url],\r\n [url=http://groups.yahoo.com/group/Hyacinthos/message/8178]Alexey A.Zaslavsky, [i]Hyacinthos message #8178[/i][/url].\r\n\r\n darij",
"Solution_3": "[url=http://www.mathlinks.ro/Forum/viewtopic.php?t=77259]There is the converse of this fact[/url] (from Russian Summer TST 2005):\r\nLet $ABCD$ be a quadrilateral. The incentres of the triangles $ABC, BCD, CDA, DAB$ form a rectangle. Proove, that $ABCD$ is cyclic.\r\nIt's very interesting and rather difficult problem.",
"Solution_4": "Dear Mathlinkers,\n\nhttp://jl.ayme.pagesperso-orange.fr/Docs/Le%20rectangle%20de%20Ryokan%20Maruyama.pdf\n\nSincerely\nJean-Louis",
"Solution_5": "We use complex numbers. Plot the points on the unit circle. Then we can choose $a,b,c,d$ such that $A=a^2,B=b^2,C=c^2,D=d^2$, and \\[I_{ABC}=-bc-ca-ab,I_{BCD}=-cd-db-bc,I_{CDA}=da+ac-cd,I_{DAB}=-ab+bd+da.\\] Then we can compute \\[I_{ABC}+I_{CDA}=-ab-bc-cd+da=I_{BCD}+I_{DAB}\\] and\n\\begin{align*}\n\\left|I_{ABC}-I_{CDA}\\right|^2&=\\left|-bc-2ca-ab-da+cd\\right|^2=\\frac{1}{abcd}\\left(-bc-2ca-ab-da+cd\\right)\\left(-ad-2bd-cd-bc+ab\\right)\\\\\n&=\\frac{1}{abcd}\\left(-cd-2db-bc-da+ab\\right)\\left(-ab-2ac-ad-bc+cd\\right)=\\left|-cd-2db-bc-da+ab\\right|^2=\\left|I_{BCD}-I_{DAB}\\right|^2\n\\end{align*}\nso the incenters form a rectangle.",
"Solution_6": "Labels are on the diagram.\nBy Fact 5 $AMNB$ is cyclic. And so is $AMQD$.\n$\\angle NMQ=\\angle B/2 +\\angle D/2=90^\\circ$. And similarly the others.\n",
"Solution_7": "i only could find those right angles on the rectangle by drawing external tangents \nand since it is a quadrilateral \nit must be a square or rectangle\nand by length chasing interestingly we could say that that square/rectangle is definitely composed of rectangles",
"Solution_8": "We use complex numbers on this one.\nWe just easily prove this a parallelogram and then just show perpendicularity once or prove diagonals equal in length. :blush: ",
"Solution_9": "My solution at [url]https://artofproblemsolving.com/community/c6t48f6h24483_abcd_cyclic_gt_incenters_of_bcd_etc_form_rectangle[/url]\nI hope you will enjoy it!\n\nBest regards,\nsunken rock",
"Solution_10": "[url]https://groups.io/g/euclid/topic/81006484#1476[/url]",
"Solution_11": "Another property. \n[url]https://geometry-diary.blogspot.com/2020/11/1521.html[/url]",
"Solution_12": "Also China 1986 TsT"
}
{
"Tag": [
"geometry",
"circumcircle",
"incenter",
"Euler",
"projective geometry",
"cyclic quadrilateral",
"geometry proposed"
],
"Problem": "Let $ w$ be the circumcircle of $ \\triangle ABC$ and let $ M\\in w$ so that $ BC$ separates $ A$ , $ M$ .\r\n\r\nDenote $ \\|\\begin{array}{ccc}\r\nP\\in (AB) & , & \\widehat {AMP}\\equiv\\widehat {BMP}\\\\\\\r\nN\\in (AC) & , & \\widehat {AMN}\\equiv\\widehat {CMN}\\end{array}$ . Prove that the incenter $ I\\in NP$ .",
"Solution_1": "Let $ MP,MN$ meet $ w$ at $ X,Y$ respectively. Then $ I\\equal{}BY\\cap CX$. By Pascal theorem, $ P,I,N$ are collinear.",
"Solution_2": "This problem was a test IMAR in 2005 (Romania) and it is well-known, generally,\r\nwith the short and nice [b]Leonhard Euler[/b]'s proof. Exists another proof of it ?\r\nWithout the [b]Pascal[/b]'s theorem ? I know one.",
"Solution_3": "A [b]Pascal[/b]-less proof for dear [b]Virgil Nicula[/b]\r\nlet B' and C' be midpoints of arcs AC and AB which doesn't contain B and C respectively.\r\nnow let $ MP\\cap BI \\equiv Y\\ ,\\ MN\\cap CI\\equiv Z$ we have that $ MYIZ$ is cyclic.\r\nhence $ \\angle IYZ \\equal{} \\angle IMZ \\equal{} \\angle B'IN\\ ,\\ \\angle IZY \\equal{} \\angle IMY \\equal{} \\angle C'IP$ hence $ IN\\parallel{}YZ\\ ,\\ IP\\ \\parallel{}YZ$ hence $ N,I,P$ collinear.",
"Solution_4": "[color=darkred][i][b]Leonhard Euler[/b][/i] and [i][b]Amir S.[/b][/i] , here is [u]another[/u] proof without [b]Pascal's theorem[/b] :[/color]\r\n\r\n[color=darkblue][b][u]Proof.[/u][/b] Apply the [b]theorem of bisector[/b] in $ \\{\\begin{array}{cc}\n\\triangle\\ AMB\\ : & \\frac {PB}{PA}=\\frac {MB}{MA}\\\\\\\\\n\\triangle\\ AMC\\ : & \\frac {NC}{NA}=\\frac {MC}{MA}\\end{array}\\ \\ (1)$ .\n\nApply the [b]Ptolemeu's theorem[/b] to the cyclic quadrilateral $ ABMC$ :\n\n$ b\\cdot MB+c\\cdot MC=a\\cdot MA$ $ \\implies$ $ b\\cdot\\frac {MB}{MA}+c\\cdot\\frac {MC}{MA}=a$ . Using \n\nthe relations $ (1)$ obtain $ b\\cdot\\frac {PB}{PA}+c\\cdot\\frac {NC}{NA}=a$ , what characterizes that $ I\\in NP$ .[/color]"
}
{
"Tag": [
"MATHCOUNTS",
"AMC 8"
],
"Problem": "Here all of us can introduce ourselves and kind of give a little \"math profile\". You do not need to give your real name if you don't want to. I'll start. I am kstan013. I am from Mosinee, WI and have taken 3rd place and 1st place at MathCounts at state in recent years. At nationals I have placed 121st and 78th. I also happen to know picounts and mustafa personally.",
"Solution_1": "9th grade, Madison\r\nI know mustafa and picounts, and have met kstan...",
"Solution_2": "I'm from Madison, I know kstan, ciaccona, and picounts, I got 3rd at mathcounts state last year, and 63rd at nats. I also got the best score in wisconsin on the AMC 8 last year, but my school is the only school I've heard of taking the AMC 8, so I don't know if that really counts as anything special at all :roll: .",
"Solution_3": "Well I am from Michigan, right next door, sort of.\r\n\r\nDetroit area.",
"Solution_4": "so mustafa was 3rd and kstan was 1st in state of Wisconsin.\r\n\r\nThen Mustafa placed 63rd, higher than kstan during the Nats.",
"Solution_5": "kstan got to first from 4th or 5th during the countdown round, on the written, I got 2nd and kstan was 4th or 5th.",
"Solution_6": "I am in 8th grade from Appleton, WI. I believe I will get to know kstan....being on the WI team.",
"Solution_7": "I am in 8th grade from Mosinee, WI. I believe I will get to know football....being on the WI team. :rotfl: :D",
"Solution_8": "Too bad nobody from Madison made it, then I could have come to your practices.",
"Solution_9": "But all of us won't have to drive to Madison now.."
}
{
"Tag": [
"calculus",
"algebra unsolved",
"algebra"
],
"Problem": "Hi, I have problem so that. Find x, y if:\r\n\r\n$\\left\\{ \\begin{array}{l} 3x^2+2xy +y^2= 11 \\\\ x^3 + 2xy + 3y^2=17\\end{array} \\right.$\r\n\r\nOh! it's very easy if we change $x^3$ by $x^2$. But it's exact $x^3$.\r\nCan you hepl me? Thanks you very much!\r\n :wallbash_red: :help:",
"Solution_1": "[quote=\"phamvanquy\"]Hi, I have problem so that. Find x, y if:\n\n$\\left\\{ \\begin{array}{l} 3x^2+2xy +y^2= 11 \\\\ x^3 + 2xy + 3y^2=17\\end{array} \\right.$\n\nOh! it's very easy if we change $x^3$ by $x^2$. But it's exact $x^3$.\nCan you hepl me? Thanks you very much!\n :wallbash_red: :help:[/quote]\nSo :\n$9x^2+6xy+3y^2=33$\n$x^3+2xy+3y^2=17$\n\nSubtracting, we get $x^3-9x^2+16=4xy$ and, since $x=0$ is never solution, $y=\\frac{x^3-9x^2+16}{4x}$\n\nPlugging this in first equation, we get $x^6-10x^5+57x^4+32x^3-336x^2+256=0$\n\n$x=1$ is a trivial root, and so $y=2$ and a first solution $\\boxed{\\text{S1 : }(x,y)=(1,2)}$\n\nIt remains to solve $x^5-9x^4+48x^3+80x^2-256x-256=0$\n\nI'm sure your teacher tought you how to easily solve this quintic. Mine did not.\n\nWaiting for a clever olympiad level solution by anybody, numeric calculus easily gives :\n\n$\\boxed{\\text{S2 : }(x,y)=(\\sim 2.34277416990797, \\sim -2.19171663906611)}$\n\n$\\boxed{\\text{S3 : }(x,y)=(\\sim -2.04113089637945, \\sim 3.7440041721502)}$\n\n$\\boxed{\\text{S4 : }(x,y)=(\\sim -0.909043992934478, \\sim -2.14828785654138)}$"
}
{
"Tag": [
"geometry",
"3D geometry",
"trigonometry"
],
"Problem": "this one should be fun. I have a vexlar that I use for ice fishing ( it basically tells the depth and if there are any fish under you by showing a little suspended blip on the readout). Anyway, here is my question..\r\n\r\nThe transducer reads the depth. Lets say the depth is 10 feet and the angle it reads is 12 degrees, what is the equasion to find the surface area that it is covering at a depth of 10 feet (the area the vexlar can pick up fish movement on the bottom of the lake if it reads 12 degrees)..? \r\n\r\nif you do not understand what I am trying to say I can put it like this. if the transducer is sitting on top of a traffic cone what is the surface area at the bottom of that cone given the cone under the transducer reads 12 degrees.",
"Solution_1": "[hide=\"answer\"]\n$4.944\\pi$\n\nThis is pretty easy...\n\nUsing Trigonometry......\n\n$tan 12=\\frac{r}{10}$ , $0.212*10=r$ , $r=2.12$, then\n$\\pi(2.12)^2=4.944\\pi feet squared$\n[/hide]",
"Solution_2": "[quote=\"ajai\"][hide=\"answer\"]\n$44\\pi$\n\nThis is pretty easy...\n\nUsing Trigonometry......\n\n$tan 12=\\frac{r}{10}$ , $0.212*10=r$ , $r=2.12$, then\n$\\pi(2.12)^2=4.944\\pi feet squared$\n[/hide][/quote]\r\n\r\nhow comes the two answers you gave aren't consistent.\r\ni guess the \"4.9\" part was edited out by accident",
"Solution_3": "i changed it.... thanx :)",
"Solution_4": "Please don't post trig problems in the middle school forums."
}
{
"Tag": [
"algebra",
"polynomial",
"number theory theorems",
"number theory"
],
"Problem": "hi\r\n\r\ncan someone please explain about cyclotomic polynomials and how they are used in solving problems involving primes.",
"Solution_1": "Perhaps you could start with http://mathworld.wolfram.com/CyclotomicPolynomial.html\r\nor http://perso.wanadoo.fr/yves.gallot/papers/cyclotomic.html\r\n\r\nHTH",
"Solution_2": "A useful fact about Zp: The roots of phip-1(x) in Zp are exactly the primitive roots mod p (p prime)."
}
{
"Tag": [
"ratio"
],
"Problem": "Hello all! \r\n\r\nI need some idees for the following thing. My english teacher said that once a week, durring the english class we have to be \"the teacher\" :D. I think that's a good way for us to experss ourselfs better and that could help us at the oral exams. Some of my classmates already did that but they picked some topics that weren't so exciting and I say that by the way the rest of my collegues responded to the choosen subject. If you have any ideeas it would be nice of you. I thought to teach some math but they don't pay attetion sometimes at the math class itself :(. Thanks!",
"Solution_1": "Some Fibonacci or golden ratio always helps :)",
"Solution_2": "What about Pithagorean Theorem proof??? Quite simple, but it'd work...",
"Solution_3": "Force them to pay attention! :lol: Our class had to do that in english to in my last year at high school, and me and my friend stubbornly wanted to teach english grammar (which is hated because over its abundance of rules about the choice of suitable tense). The teacher really became anxious and in the days before it, he was quite nervous to learn when we were gonna say our real subject. However i copied some notes to be filled in, and most of the class thought they were somehow forced to fill them in.\r\n\r\nSeriously, there a few basic things about teaching math even I have encountered already : \r\n\r\nYour audience is never up to speed as you would wish them to be.\r\nYou always go slower than you thought you would.\r\n\r\n\r\nAlso what I (as a movie lover) would really appreciate is that when you talk about the subject , you point out where it pops up. For instance adjacency matrices pop up in Good Will Hunting, taylor approximations (somewhat) are used to beat a chinese store owner in eternal sunset...",
"Solution_4": "[quote=\"Jos\u00e9\"]What about Pithagorean Theorem proof??? Quite simple, but it'd work...[/quote]\r\n\r\nI'm in 11th grade :lol: We learn now Analysis and Linear Algebra.\r\n\r\nWell, I'd like to choose a original topic. Not math only, in fact no math at all. My teacher is allergic to math so it would be improper. I know you're smart guys so c`mon :D !!!",
"Solution_5": "Maybe you should try something connected with Computer Science. Perhaps talk about how a game is created (you can learn about that on Google :) ) ...",
"Solution_6": "Well, an entire world of more knowledge opened up with [url]http://www.wikipedia.org[/url] for me\r\n\r\nYou could teach about some not well known historic event, like the third indochina war or the us grenada intervention, meiji revolution\r\n\r\nor you could make a nice copyright infringing explanation of historic pictures (always nice) : \r\n\r\n[url]http://en.wikipedia.org/wiki/Category:Memorable_photographs[/url]\r\n\r\n\r\ni'm just brainstorming, but i like recent history\r\n\r\n is also nice : [url]http://en.wikipedia.org/wiki/Shibboleth[/url]",
"Solution_7": "Explain something interesting, such as current events. Just bring along a newspaper!"
}
{
"Tag": [],
"Problem": "You are at a game show. There are three doors in front of you. You know that one has treasure behind it and the other two have nothing behind them, but you don't know which one has the treasure. You want to get the treasure. So you pick a door. But, before looking behind it, the game show host opens one of the two other doors and you find that the room behind it is empty. The game show host now gives you the option of switching to the third remaining door. The question is, would you have a better chance of getting the treasure by switching?",
"Solution_1": "ahhh...the monty hall problem",
"Solution_2": "Just a small recomendation for K81o7.\r\n\r\nPlease, try to call problems in order that everyone could understand its essence from the title.\r\nAlso you can read http://www.mathlinks.ro/viewtopic.php?p=140098\r\n\r\nThanks",
"Solution_3": "does it matter? your chance is $.5$",
"Solution_4": "Oh...sorry about that. I didn't know that there was a name for this problem...\r\n\r\n[hide=\"Hint\"]Think about the show host's options. He has to open a door with nothing behind it right? And that selection has to be made from the two doors remaining after you choose.[/hide]\n\n[hide=\"Answer\"]Well think about it. If you chose a door with nothing behind it (the chance of doing so is 2/3), then the other door with nothing behind it is revealed, and so the third door has the treasure. If you chose the one with treasure behind it, then either of the other two doors could have been opened but then switching would get you a door with nothing behind it. Overall, when you switch, your chance of getting the treasure ends up as 2/3(as opposed to 1/3 from not switching). So you would want to switch. [/hide][/hide]"
}
{
"Tag": [],
"Problem": "{$ a_{n}$} $ (n \\geq 0)$ is a sequence with $ a_{i} > 0 , (i\\equal{}1,2,...,n)$ and $ a_{n}\\equal{}\\frac{a_{n\\minus{}1}}{m a_{n\\minus{}2}}$ $ (n\\equal{}2,3,...)$ $ m\\in\\Re \\plus{}$ , If $ a_{2009}\\equal{}\\frac{a_{0}}{a_{1}}$ , find $ m$",
"Solution_1": "[hide=\"Hint\"]The recurrence formula telescopes in products[/hide]",
"Solution_2": "An important condition is\r\n\r\n$ a_{n}\\equal{}a_{n\\minus{}6}$",
"Solution_3": "Thanks for helping"
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "1) Let $ a,$ $ b$ and $ c$ are non-negative numbers. Prove that\r\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\leq\\sqrt {\\sum_{cyc}(5a^2 \\plus{} 4ab)}\r\n\\]\r\nFor the collection. \r\nWith the same conditions prove that:\r\n2)\r\n\\[ \\sum_{cyc}\\sqrt {4a^2 \\plus{} ab \\plus{} 4b^2}\\leq\\sqrt {\\sum_{cyc}(22a^2 \\plus{} 5ab)}\r\n\\]\r\n3)\r\n\\[ \\sum_{cyc}\\sqrt {9a^2 \\minus{}2ab \\plus{} 9b^2}\\leq2\\sqrt {\\sum_{cyc}(13a^2 \\minus{}ab)}\r\n\\]",
"Solution_1": "[quote=\"arqady\"]1) Let $ a,$ $ b$ and $ c$ are non-negative numbers. Prove that\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\leq\\sqrt {\\sum_{cyc}(5a^2 \\plus{} 4ab)}\n\\]\nFor the collection. \nWith the same conditions prove that:\n2)\n\\[ \\sum_{cyc}\\sqrt {4a^2 \\plus{} ab \\plus{} 4b^2}\\leq\\sqrt {\\sum_{cyc}(22a^2 \\plus{} 5ab)}\n\\]\n3)\n\\[ \\sum_{cyc}\\sqrt {9a^2 \\minus{} 2ab \\plus{} 9b^2}\\leq2\\sqrt {\\sum_{cyc}(13a^2 \\minus{} ab)}\n\\]\n[/quote]\r\n1) By Cauchy Schwarz Inequality, we have\r\n\\[ LHS^2 \\le 2(a \\plus{} b \\plus{} c)\\sum \\frac {a^2 \\plus{} ab \\plus{} b^2}{a \\plus{} b} \\equal{} 2\\sum (a^2 \\plus{} ab \\plus{} b^2) \\plus{} 2 \\sum \\frac {c(a^2 \\plus{} ab \\plus{} b^2)}{a \\plus{} b} \\equal{} 4\\sum a^2 \\plus{} 6\\sum ab \\minus{} 2abc\\sum \\frac {1}{a \\plus{} b}\r\n\\]\r\nIt suffices to prove that\r\n\\[ \\sum a^2 \\plus{} 2abc\\sum \\frac {1}{a \\plus{} b} \\ge 2\\sum ab\r\n\\]\r\n\r\n\\[ \\Leftarrow \\sum a^2 \\plus{} \\frac {9abc}{a \\plus{} b \\plus{} c} \\ge 2\\sum ab\r\n\\]\r\n\r\n\\[ \\Leftrightarrow \\sum a(a \\minus{} b)(a \\minus{} c) \\ge 0\r\n\\]\r\nwhich is true by Schur's Inequality. :)\r\n\r\nCYH is so strong! :)\r\n\r\nP/s: 2) and 3) can be proved by a \"similar\" way. :)",
"Solution_2": "[quote=\"can_hang2007\"][quote=\"arqady\"]1) Let $ a,$ $ b$ and $ c$ are non-negative numbers. Prove that\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\leq\\sqrt {\\sum_{cyc}(5a^2 \\plus{} 4ab)}\n\\]\n\n[/quote]\n1) By Cauchy Schwarz Inequality, we have\n\\[ LHS^2 \\le 2(a \\plus{} b \\plus{} c)\\sum \\frac {a^2 \\plus{} ab \\plus{} b^2}{a \\plus{} b}\n\\]\n\n[/quote]\r\nYes! :lol: \r\nIt's also my proof.",
"Solution_3": "[quote=\"arqady\"]1) Let $ a,$ $ b$ and $ c$ are non-negative numbers. Prove that\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\leq\\sqrt {\\sum_{cyc}(5a^2 \\plus{} 4ab)}\n\\]\n[/quote]\r\n\r\nArqady, do you think the following inequalities are also true with the same conditions ?\r\n\r\n1/ $ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{3\\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2}$\r\n\r\nand\r\n\r\n2/ $ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{9\\sum_{cyc}(2a^2 \\plus{} bc)^2}$",
"Solution_4": "dduclam, the first is Jensen and the second follows from the first. :wink:",
"Solution_5": "[quote=\"dduclam\"][quote=\"arqady\"]1) Let $ a,$ $ b$ and $ c$ are non-negative numbers. Prove that\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\leq\\sqrt {\\sum_{cyc}(5a^2 \\plus{} 4ab)}\n\\]\n[/quote]\n\nArqady, do you think the following inequalities are also true with the same conditions ?\n\n1/ $ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{3\\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2}$\n\nand\n\n2/ $ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{9\\sum_{cyc}(2a^2 \\plus{} bc)^2}$[/quote]\r\nProblem 1.\r\n\\[ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{3\\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2} \\leftrightarrow \\sum_{cyc}\\sqrt [3]{3(a^2 \\plus{} ab \\plus{} b^2)^2}\\le \\sqrt [3]{9\\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2}\r\n\\]\r\nBy Holder ;\r\n\\[ \\sum_{cyc}\\sqrt [3]{3(a^2 \\plus{} ab \\plus{} b^2)^2} \\leq \\sqrt [3]{9(\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2}\r\n\\]\r\nIt's remain to prove\r\n\\[ \\sqrt [3]{9(\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2} \\le \\sqrt [3]{9\\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2}\r\n\\]\r\nwhich is obvious.(because $ (\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2 \\equal{} \\left(\\sum_{cyc}(2a^2 \\plus{} bc)\\right)^2$)\r\nProblem 2.\r\n\\[ \\sum_{cyc}\\sqrt [3]{(a^2 \\plus{} ab \\plus{} b^2)^2}\\le\\sqrt [3]{9\\sum_{cyc}(2a^2 \\plus{} bc)^2} \\longleftrightarrow \\sum_{cyc}\\sqrt [3]{3(a^2 \\plus{} b^2 \\plus{} c^2)^2} \\le 3\\sqrt [3]{\\sum_{cyc}(2a^2 \\plus{} bc)^2}\r\n\\]\r\nBy Holder ;\r\n\\[ \\sum_{cyc}\\sqrt [3]{3(a^2 \\plus{} b^2 \\plus{} c^2)^2} \\le \\sqrt [3]{(9)(\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2}\r\n\\]\r\nIt's remain to prove that\r\n\\[ \\sqrt [3]{9(\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2} \\le 3\\sqrt [3]{\\sum_{cyc}(2a^2 \\plus{} bc)^2}\r\n\\]\r\n\r\n\\[ \\longleftrightarrow (\\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2))^2 \\le 3\\sum_{cyc}(2a^2 \\plus{} bc)^2\r\n\\]\r\nwhich is obviously true by Cauchy-Schwarz ; $ 3\\sum_{cyc}(2a^2 \\plus{} bc)^2 \\geq (\\sum_{cyc}(2a^2 \\plus{} bc))^2 \\equal{} \\sum_{cyc}(a^2 \\plus{} ab \\plus{} b^2)^2$\r\nEdit : arqady is faster :blush:",
"Solution_6": "[quote=\"arqady\"]With the same conditions prove that:\n2)\n\\[ \\sum_{cyc}\\sqrt {4a^2 \\plus{} ab \\plus{} 4b^2}\\leq\\sqrt {\\sum_{cyc}(22a^2 \\plus{} 5ab)}\n\\]\n3)\n\\[ \\sum_{cyc}\\sqrt {9a^2 \\minus{} 2ab \\plus{} 9b^2}\\leq2\\sqrt {\\sum_{cyc}(13a^2 \\minus{} ab)}\n\\]\n[/quote]\r\n\r\nI cannot prove them using the same method of can_hang2007. :( \r\n\r\nA hint for me please, Arqady.\r\n\r\n\r\nThank you very much.",
"Solution_7": "[quote=\"manlio\"][quote=\"arqady\"]With the same conditions prove that:\n2)\n\\[ \\sum_{cyc}\\sqrt {4a^2 \\plus{} ab \\plus{} 4b^2}\\leq\\sqrt {\\sum_{cyc}(22a^2 \\plus{} 5ab)}\n\\]\n3)\n\\[ \\sum_{cyc}\\sqrt {9a^2 \\minus{} 2ab \\plus{} 9b^2}\\leq2\\sqrt {\\sum_{cyc}(13a^2 \\minus{} ab)}\n\\]\n[/quote]\n\nI cannot prove them using the same method of can_hang2007. :( \n\nA hint for me please, Arqady.\n\n\nThank you very much.[/quote]\r\nThe hint to the $ 3).$\r\nBy Cauchy Schwartz for all $ k\\geq0$ we obtain: \r\n$ \\sum_{cyc}\\sqrt {9a^2 \\minus{} 2ab \\plus{} 9b^2}\\leq\\sqrt {\\sum_{cyc}\\frac {9a^2 \\minus{} 2ab \\plus{} 9b^2}{a \\plus{} b \\plus{} kc}\\cdot\\sum_{cyc}(a \\plus{} b \\plus{} kc)}.$\r\nHence, it remains to find a value of $ k$ for which the following inequality is true.\r\n$ (k \\plus{} 2)(a \\plus{} b \\plus{} c)\\sum_{cyc}\\frac {9a^2 \\minus{} 2ab \\plus{} 9b^2}{a \\plus{} b \\plus{} kc}\\leq 4\\sum_{cyc}(13a^2 \\minus{} ab).$\r\nAn equality in the original inequality holds also when $ a \\equal{} b \\equal{} 1$ and $ c \\equal{} 0,$ which gives $ k \\equal{} \\frac {1}{2}.$ :wink:",
"Solution_8": "$ (k \\plus{} 2)(a \\plus{} b \\plus{} c)\\sum_{cyc}\\frac {9a^2 \\minus{} 2ab \\plus{} 9b^2}{a \\plus{} b \\plus{} kc}\\leq 4\\sum_{cyc}(13a^2 \\minus{} ab).$\r\n\r\n\r\nDo you solve it by full expansion and pqr method, Arqady? \r\n\r\nThank you very much. :)",
"Solution_9": "[quote=\"manlio\"]$ (k \\plus{} 2)(a \\plus{} b \\plus{} c)\\sum_{cyc}\\frac {9a^2 \\minus{} 2ab \\plus{} 9b^2}{a \\plus{} b \\plus{} kc}\\leq 4\\sum_{cyc}(13a^2 \\minus{} ab).$\n\n\nDo you solve it by full expansion and pqr method, Arqady? \n\nThank you very much. :)[/quote]\r\nFor $ k\\equal{}\\frac{1}{2}$ after full expansion we obtain an inequality, which is true by Schur and Muirhead.",
"Solution_10": "[quote=\"arqady\"]dduclam, the first is Jensen and the second follows from the first. :wink:[/quote]\r\n\r\nI see, arqady :lol: And I think the following inequality is also true and hard :\r\n\r\n$ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{b^3 \\plus{} c^3})$\r\n\r\nDo you think so ?",
"Solution_11": "[quote=\"dduclam\"][quote=\"arqady\"]dduclam, the first is Jensen and the second follows from the first. :wink:[/quote]\n\nI see, arqady :lol: And I think the following inequality is also true and hard :\n\n$ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{b^3 \\plus{} c^3})$\n\nDo you think so ?[/quote]\r\nIt is still very easy, my friend ;)\r\nSince we can easily prove that (by Cauchy Schwarz)\r\n\\[ LHS \\le 3\\sqrt{a^2\\plus{}b^2\\plus{}c^2} \\le RHS\\]\r\n:)\r\nMaybe you meant:\r\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{a^3 \\plus{} b^3})\\]\r\n?",
"Solution_12": "[quote=\"dduclam\"][quote=\"arqady\"]dduclam, the first is Jensen and the second follows from the first. :wink:[/quote]\n\nI see, arqady :lol: And I think the following inequality is also true and hard :\n\n$ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{b^3 \\plus{} c^3})$\n\nDo you think so ?[/quote]\r\nIt is still very easy, my friend ;)\r\nSince we can easily prove that (by Cauchy Schwarz)\r\n\\[ LHS \\le 3\\sqrt{a^2\\plus{}b^2\\plus{}c^2} \\le RHS\\]\r\n:)\r\nMaybe you meant:\r\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{a^3 \\plus{} b^3})\\]\r\n?",
"Solution_13": "[quote=\"can_hang2007\"]\nMaybe you meant:\n\\[ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^4}{a^3 \\plus{} b^3})\n\\]\n?[/quote]\r\n\r\nOh, yes. I'm sorry for my wrong type :(\r\n\r\nBesides, the following inequalities also seems correct :\r\n\r\n$ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^6}{a^5 \\plus{} b^5})$\r\n\r\n$ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^7}{a^6 \\plus{} b^6})$\r\n\r\nAnd $ \\sum_{cyc}\\sqrt {a^2 \\plus{} ab \\plus{} b^2}\\le 2\\sqrt3(\\sum_{cyc}\\frac {a^8}{a^7 \\plus{} b^7})$ wrong allready :)",
"Solution_14": "Tajikistan TST 2014!"
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Let $ a,b,c\\geq 0$, s.t. $ a\\plus{}b\\plus{}c\\equal{}3$.Prove that:\r\n$ \\frac{1}{6\\minus{}ab}\\plus{}\\frac{1}{6\\minus{}bc}\\plus{}\\frac{1}{6\\minus{}ca}\\leq \\frac{3}{5}$.",
"Solution_1": "[quote=\"Inequalities Master\"]Let $ a,b,c\\geq 0$, s.t. $ a \\plus{} b \\plus{} c \\equal{} 3$.Prove that:\n$ \\frac {1}{6 \\minus{} ab} \\plus{} \\frac {1}{6 \\minus{} bc} \\plus{} \\frac {1}{6 \\minus{} ca}\\leq \\frac {3}{5}$.[/quote]\r\n$ <\\equal{}>16(ab\\plus{}bc\\plus{}ca) \\plus{}a^2b^2c^2 \\le 36\\plus{}13abc<\\equal{}>$\r\n$ a^2(b\\plus{}c)\\plus{}b^2(c\\plus{}a) \\plus{}c^2(a\\plus{}b)\\plus{}\\frac{3}{4}a^2b^2c^2 \\le a^3\\plus{}b^3\\plus{}c^3\\plus{}\\frac{15}{4}abc$( which is true by Schur and $ abc \\le 1$); :)",
"Solution_2": "who can resolve by Chebychev? :wink:"
}
{
"Tag": [],
"Problem": "In a 12-hour period of time, for how many minutes is at least one ``6'' displayed on a digital clock?",
"Solution_1": "A \"6\" only occurs at x:06, x:16, x:26, x:36, x:46, or x:56. Or when the first digit is 6.\r\n\r\nSo we have 11*6=66, add 60=126 minutes.",
"Solution_2": "There are 59 times with a 6 in the hours, and $ 5 \\cdot 12\\equal{}60$ times for the minutes. However, there are $ 5$ times when we overcount, so $ 59\\plus{}60\\minus{}5\\equal{}114$.",
"Solution_3": "I think you're wrong - for one, there's 60 times with a 6 in the hours: 6:00-6:59.\r\n\r\nAlso, it's not 5*12, its 6*12, b/c we have x:06, x:16, x:26, x:36, x:46, and x:56.\r\n\r\nSo using your method, we have 60+72-6=126 minutes.",
"Solution_4": "I thought a digital clock shows seconds too?...",
"Solution_5": "It depends on which digital clock you're using, I suppose. Mine doesn't.",
"Solution_6": "ok thanks\n"
}
{
"Tag": [
"probability"
],
"Problem": "What is the probbility that each player has a hand containing an ace when the 52 cards of a standard deck are dealt to four players , i.e. each player has 13 cards.",
"Solution_1": "Um\r\n[hide]$\\frac{13^4}{52\\cdot 51 \\cdot 50 \\cdot 49}$[/hide]",
"Solution_2": "but isnt there $4!$ ways of distributing the aces?\r\n\r\nbut anyway how did you get what you got?\r\n\r\n$\\frac{24\\cdot 13^4}{52\\cdot 51 \\cdot 50 \\cdot 49}$",
"Solution_3": "[hide] Yeah, I think he forgot to include the order of the aces. Consider it using permutations. There are a total of $52!$ ways to distribute the cards. For each player, there are $13$ possible slots to put the ace which gives $13^4$, and then a total of $48!$ ways to distribute the rest.\n\nprobability then equals: $\\frac{4!\\times13^4\\times48!}{52!}=\\frac{24\\times13^4}{52\\times51\\times50\\times49}$.\n[/hide]"
}
{
"Tag": [
"function",
"algebra unsolved",
"algebra"
],
"Problem": "Find all continuous functions $ f: \\mathbb{R} \\to \\mathbb{R}$ such that \\[ f\\left( {\\frac {{x \\plus{} y}} {2}} \\right) \\equal{} \\sqrt {f(x)f(y)},\\] for all reals $x$ and $y$.",
"Solution_1": "[quote=\"Xaenir\"]Find all continuous functions $ f: \\mathbb{R} \\to \\mathbb{R}$such that $ f\\left( {\\frac {{x \\plus{} y}} {2}} \\right) \\equal{} \\sqrt {f(x)f(y)}$ , $ \\forall x,y$[/quote]\r\n\r\nIf we can find $ x$ such that $ f(x)\\equal{}0$, then put it into the equation we get $ f(x)\\equal{}0$ for all $ x$ as a solution.\r\n\r\nIf not, then put $ p\\equal{}\\frac{x\\plus{}y}{2}$ we get $ f(p)$ is positive for all $ p$\r\n\r\n\r\nNow $ \\sqrt{f(x\\plus{}y)f(0)}\\equal{}f(\\frac{x\\plus{}y}{2})\\equal{}\\sqrt{f(x)f(y)}$\r\n\r\nSo $ f(x\\plus{}y)f(0)\\equal{}f(x)f(y)$\r\n$ \\Leftrightarrow In (f(x\\plus{}y))\\plus{}In (f(0))\\equal{}In (f(x))\\plus{}In(f(y))$\r\n\r\nNow let $ g(x)\\equal{}In (f(x))\\minus{}In (f(0))$ the equation becomes:\r\n\r\n$ g(x)\\plus{}g(y)\\equal{}g(x\\plus{}y)$\r\n\r\nAs $ g(x)$ is continuous (because $ f(x)$ is continuous), $ g(x)\\equal{}cx$ (as Cauchy) , where $ c$ is a constant.\r\n\r\nLet $ b\\equal{}f(0)$, we have $ f(x)\\equal{}e^{cx\\plus{}In(f(0))}$\r\n\r\nLet $ k\\equal{}e^c$, $ t\\equal{}f(0)$\r\n\r\n\r\nWe get $ f(x)\\equal{}tk^x$ is the solution after checking, where $ t$ is any real constant and $ k$ is any non-zero constant.",
"Solution_2": "[quote=\"Xaenir\"]Find all continuous functions $ f: \\mathbb{R} \\to \\mathbb{R}$such that $ f\\left( {\\frac {{x \\plus{} y}} {2}} \\right) \\equal{} \\sqrt {f(x)f(y)}$ , $ \\forall x,y$[/quote]\r\n\r\nWe clearly have $ f(x)\\geq 0$\r\nIf $ f(x_0)\\equal{}0$ for some $ x_0$, we have obviously $ f(x)\\equal{}f(\\frac{x_0\\plus{}(2x\\minus{}x_0)}{2})\\equal{}\\sqrt{f(x_0)f(2x\\minus{}x_0)}\\equal{}0$ $ \\forall x$\r\n\r\nConsider now $ f(x)>0$ $ \\forall x$\r\n\r\n$ f(\\frac{(x\\plus{}y)\\plus{}0}{2})\\equal{}\\sqrt{f(x\\plus{}y)f(0)}$ and so $ f(x\\plus{}y)\\equal{}\\frac{1}{f(0)}f(x)f(y)$\r\n\r\nThis is a classical equation whose the only continuous solutions are $ f(x)\\equal{}a\\cdot e^{bx}$\r\n\r\nPlugging back in the original equation, we get the solutions :\r\n\r\n$ f(x)\\equal{}a\\cdot e^{bx}$ $ \\forall a\\geq 0$",
"Solution_3": "you can take log two side, then you have Cauchy equation",
"Solution_4": "[quote=\"hqthao\"]you can take log two side, then you have Cauchy equation[/quote]\r\n\r\nYou must first show that $ f(x)\\neq 0$\r\n\r\nAnd, btw, this is exactly what Stephencheng did. :wink:",
"Solution_5": "[quote=pco][quote=\"Xaenir\"]Find all continuous functions $ f: \\mathbb{R} \\to \\mathbb{R}$such that $ f\\left( {\\frac {{x \\plus{} y}} {2}} \\right) \\equal{} \\sqrt {f(x)f(y)}$ , $ \\forall x,y$[/quote]\n\nWe clearly have $ f(x)\\geq 0$\nIf $ f(x_0)\\equal{}0$ for some $ x_0$, we have obviously $ f(x)\\equal{}f(\\frac{x_0\\plus{}(2x\\minus{}x_0)}{2})\\equal{}\\sqrt{f(x_0)f(2x\\minus{}x_0)}\\equal{}0$ $ \\forall x$\n\nConsider now $ f(x)>0$ $ \\forall x$\n\n$ f(\\frac{(x\\plus{}y)\\plus{}0}{2})\\equal{}\\sqrt{f(x\\plus{}y)f(0)}$ and so $ f(x\\plus{}y)\\equal{}\\frac{1}{f(0)}f(x)f(y)$\n\nThis is a classical equation whose the only continuous solutions are $ f(x)\\equal{}a\\cdot e^{bx}$\n\nPlugging back in the original equation, we get the solutions :\n\n$ f(x)\\equal{}a\\cdot e^{bx}$ $ \\forall a\\geq 0$[/quote]\n\nwhat happen if x0 =0 ??? then we cannot divide by f(0) to get solution that you have arrived....\n\n",
"Solution_6": "[quote=Understandingmathematics]what happen if x0 =0 ??? then we cannot divide by f(0) to get solution that you have arrived....[/quote]\nI never divided by $f(x_0)$. I just trivially proved that $f(x_0)=0$ implies solution $f(x)=0$ $\\forall x$ (whatever is $x_0$)\n\n\nRead more carefully.\nThere are two parts in my proof :\nPart 1 : If $\\exists x_0$ such that $f(x_0)=0$\nThen solution is $f\\equiv 0$\n\nPart 2 : If no such $x_0$ and so $f(x)>0$ $\\forall x$ ..\nThen ...\nAnd in this part, I can divide by $f(0)$ since we know that $f(x)>0$ $\\forall x$",
"Solution_7": "this means that there are only one solution $f(x)=0$ or $f(x)$=a e^(bx)",
"Solution_8": "[quote=Understandingmathematics]this means that there are only one solution $f(x)=0$ or $f(x)=$a$e^(bx)$[/quote]\nThis means that we have :\nEither $f(x)=0$ $\\forall x$\nEither $f(x)>0$ $\\forall x$ and then $f(x)=a.e^{bx}$, whatever are $a>0$, and $b$\n\nAnd I merged the two solutions by allowing $a>0$ in the second to be $a\\ge 0$\n\nAnd so general solution $f(x)=a.e^{bx}$ $\\forall x$, whatever are $a\\ge 0$ and $b$\n\n",
"Solution_9": "thankyou very much"
}
{
"Tag": [
"SFFT",
"special factorizations"
],
"Problem": "3 distinct unit fractions are added together to make $\\frac{6}{7}$. What is the least possible sum of the denominators?\r\n\r\nThis was a #30 on the sprint, it prevented me from getting a 22 on the sprint :( .",
"Solution_1": "[quote=\"Ubemaya\"]3 distinct unit fractions are added together to make $\\frac{6}{7}$. What is the least possible sum of the denominators?\n\nThis was a #30 on the sprint, it prevented me from getting a 22 on the sprint :( .[/quote]\r\n[hide]$\\frac{1}{x}+\\frac{1}{y}+\\frac{1}{z}$\nWe need to minimize x y and z. Try x=2 and this is now $\\frac{1}{y}+\\frac{1}{z}=\\frac{5}{14}$ Try x=3. We got $\\frac{1}{42}$. This is kinda big, I really doubt it minimizes. [/hide]\r\nI'm about 85% sure that this is wrong.",
"Solution_2": "$\\frac{1}{x}+\\frac{1}{y}+\\frac{1}{z}=\\frac{6}{7}$\r\n$\\frac{xy+yz+xz}{xyz}=\\frac{6}{7}$\r\n\r\nWell, by analysis, $xyz$ in the denominator will have a factor of 7, so let $x=7$. Then:\r\n\r\n$\\frac{7y+7z+yz}{7yz}=\\frac{6}{7}$\r\n$\\frac{1}{z}+\\frac{1}{y}+\\frac{1}{7}= \\frac{6}{7}$\r\n$\\frac{1}{z}+\\frac{1}{y}= \\frac{5}{7}$\r\nSimon's favorite factoring trick won't work on this...\r\n\r\nSo try $x=14$ the next factor of 7. And keep trying x of the form $x=7r$ until something works, and that will be the minimum???\r\n\r\nIs that a good plan? It seems kind of long-winded for a sprint.",
"Solution_3": "[hide]1/7+2/7+3/7=6/7 so 21?[/hide]",
"Solution_4": "That works, but the problem asks for unit fractions, which are of the form $\\frac{1}{x}$ where x is an integer.",
"Solution_5": "[hide]Use the Greedy Algorithm.\n\nFirst go with $\\frac{1}{2}$\n\nThen go with $\\frac{1}{3}$\n\nThen the last one is $\\frac{1}{42}$\n\nSo the answer is $47$[/hide]",
"Solution_6": "[quote=\"mysmartmouth\"]$\\frac{1}{x}+\\frac{1}{y}+\\frac{1}{z}=\\frac{6}{7}$\n$\\frac{xy+yz+xz}{xyz}=\\frac{6}{7}$\n\nWell, by analysis, $xyz$ in the denominator will have a factor of 7, so let $x=7$. Then:\n\n$\\frac{7y+7z+yz}{7yz}=\\frac{6}{7}$\n$\\frac{1}{z}+\\frac{1}{y}+\\frac{1}{7}= \\frac{6}{7}$\n$\\frac{1}{z}+\\frac{1}{y}= \\frac{5}{7}$\nSimon's favorite factoring trick won't work on this...\n\nSo try $x=14$ the next factor of 7. And keep trying x of the form $x=7r$ until something works, and that will be the minimum???\n\nIs that a good plan? It seems kind of long-winded for a sprint.[/quote]\r\nOh, you can use simon's favorite factoring trick on that. \r\n$(y-1.4)(z-1.4)=1.96$\r\n$(5y-7)(5z-7)=49$\r\nClearly 2.8/2.8 won't work. Neither with the other thing. Both Neil and my solution are the same, but I have doubts about it being the smallest.",
"Solution_7": "[hide=\"Solution\"]Assume that the denominators $x$, $y$, and $z$ are in the order $x < y < z$. Choosing $x = 1$ makes the sum too big, because $1 > \\frac{6}{7}$. Choosing $x \\ge 3$ makes the sum too small, because\n\\[\\frac{1}{3}+\\frac{1}{4}+\\frac{1}{5}= \\frac{47}{60}< \\frac{6}{7}\\, . \\]\nThus $x = 2$. That means\n\\[\\frac{1}{y}+\\frac{1}{z}= \\frac{5}{14}\\, . \\]\nChoosing $y \\ge 6$ makes the sum too small, because\n\\[\\frac{1}{6}+\\frac{1}{7}= \\frac{13}{42}< \\frac{5}{14}\\, . \\]\nThus $y$ is either 3, 4, or 5. If $y = 3$, then $z = 42$. If $y = 4$, then $z = \\frac{28}{3}$, which is not an integer. If $y = 5$, then $z = \\frac{70}{11}$, which is not an integer. \n\nTherefore the only triple that works (up to reordering) is $(2, 3, 42)$. The sum is indeed $\\boxed{47}$.[/hide]",
"Solution_8": "the answer sheet indeed says that it is 47."
}
{
"Tag": [],
"Problem": "hello,\r\nProve that $ (x,y)\\equal{}(2,1)$ is the only solution of the system : $ \\quad \\large (x,y) \\in {\\mathbb R}^2 ,\\quad \\left\\{\\begin{array}{l} y(x^3 \\minus{} y^3) \\equal{} 7 \\\\\r\ny(x \\plus{} y)^2 \\equal{} 9 \\end{array} \\right.$",
"Solution_1": "it's so hard ?",
"Solution_2": "hello, from equation (2) we get $ x_1\\equal{}\\minus{}\\frac{y^{\\frac{3}{2}}\\minus{}3}{\\sqrt{y}}$ and $ x_2\\equal{}\\minus{}\\frac{y^{\\frac{3}{2}}\\plus{}3}{\\sqrt{y}}$, inserting $ x_1$ in equation (1) and simplifying we get\r\n$ \\minus{}2y^{\\frac{11}{2}}\\plus{}9y^4\\minus{}27y^{\\frac{5}{2}}\\plus{}27y\\minus{}7y^{\\frac{3}{2}}\\equal{}0$. This equation has only the real root $ y\\equal{}1$ and from here we get $ x\\equal{}2$.\r\nBy inserting $ x_2$ in equation (1) we get\r\n$ \\minus{}2y^{\\frac{11}{2}}\\minus{}9y^4\\minus{}27y^{\\frac{5}{2}}\\minus{}27y\\minus{}7y^{\\frac{3}{2}}<0$ for $ y>0$, hence $ x\\equal{}2,y\\equal{}1$ is the only solution.\r\nSonnhard."
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "Let $ a,b,c,d\\in [0,1]$. Find the maximum value of A\r\n $ A\\equal{}\\frac{a}{1\\plus{}bcd}\\plus{}\\frac{b}{1\\plus{}cda}\\plus{}\\frac{c}{1\\plus{}dab}\\plus{}\\frac{d}{1\\plus{}abc}$\r\n :D",
"Solution_1": "[quote=\"toanIneq\"]Let $ a,b,c,d\\in [0,1]$. Find the maximum value of A\n $ A \\equal{} \\frac {a}{1 \\plus{} bcd} \\plus{} \\frac {b}{1 \\plus{} cda} \\plus{} \\frac {c}{1 \\plus{} dab} \\plus{} \\frac {d}{1 \\plus{} abc}$\n :D[/quote]\r\nWe have:\r\n$ bcd\\plus{}1 \\geq abcd\\plus{}1$\r\n$ cda\\plus{}1 \\geq abcd\\plus{}1$\r\n$ dab\\plus{}1 \\geq abcd\\plus{}1$\r\n$ abc\\plus{}1 \\geq abcd\\plus{}1$\r\n=>$ A \\leq \\frac{a\\plus{}b\\plus{}c\\plus{}d}{abcd\\plus{}1} \\leq \\frac{1\\plus{}ab\\plus{}1\\plus{}cd}{1\\plus{}abcd}\\leq \\frac{1\\plus{}1\\plus{}1\\plus{}abcd}{1\\plus{}abcd} \\leq \\frac{3\\plus{}3abcd}{1\\plus{}abcd} \\equal{}3$\r\nEquality occurs when $ a;b;c\\equal{}1;d\\equal{}0$ and cyc..",
"Solution_2": "Similar problem:\r\nLet $ a,b,c,d;e \\in [0,1]$. Find the maximum value of M\r\n $ M \\equal{} \\frac {a}{1 \\plus{} bcde} \\plus{} \\frac {b}{1 \\plus{} cdea} \\plus{} \\frac {c}{1 \\plus{} deab} \\plus{} \\frac {d}{1 \\plus{} eabc}\\plus{}\\frac{e}{1\\plus{}abcd}$"
}
{
"Tag": [
"geometry",
"circumcircle",
"ratio",
"cyclic quadrilateral",
"geometry proposed"
],
"Problem": "Let $ABC$ be a triangle and $K$ be its symmedian point. Let $A', B', C'$ be the intersection of $AK, BK, CK$ respectively and the circumcircle. Let $U$ be the intersection of the tangent at $A'$ and the tangent at $B$. Call $M$ and $N$ the intersections of $UK$ with the circumcircle. Prove that $M, N, C, C'$ form a harmonic quadrilateral.",
"Solution_1": "What is a harmonic quadrilateral? In- and circumscriptible?",
"Solution_2": "This site http://www.2dcurves.com/line/lineq.html says that it is a cyclic quadrilateral for which the products of opposite sides are the same.",
"Solution_3": "That's probably not the most fortunate description here. Another definition is: a cyclic quadrilateral $ABCD$ s.t. the tangents in $A,C$ to the circumcircle meet on $BD$ (and the same can be said if the pair $A,C$ is interchanged with the pair $B,D$).\r\n\r\nI'll just sketch it: if $V$ is the intersection between the tangents in $A,B'$ to the circumcircle, then $MN=UV$. We must thus prove that the tangents in $C,C'$ meet on $UV$, or, equivalently, since the polaes of the tangents in $C,C'$ are $C$ and $C'$ respectively, we must show that $P=AB'\\cap BA'$, which is the pole of $UV$, lies on $CC'$. \r\n\r\nThis follows from the following general fact: given $XYZT$, a circumscribed quadrilateral, which touches its incircle in $W_1,W_2,W_3,W_4$ (these points lie on $XY,YZ,ZT,TX$ respectively), $W_1W_4,W_2W_3,YT$ are concurrent. We apply this to the circumscribed quadrilateral determined by the tangents to the circumcircle in $A,B,A',B'$, by also using the fact that the tangents in $A,B$ meet on $CC'$, and so do those in $A',B'$, so $CC'$ is one of the diagonals of this quadrilateral.\r\n\r\nSorry for the lack of detail.",
"Solution_4": "[quote=\"blahblahblah\"] [harmonic quadrilateral is] a cyclic quadrilateral for which the products of opposite sides are the same.[/quote]\r\n\r\nAh! That explains the word \"harmonic\", because it is equivalent to the cross-ratio of the 4 vertices (under projection to any line) being equal to -1."
}
{
"Tag": [
"topology"
],
"Problem": "Let $ n$ be a positive integer and let $ K$ be a convex and compact subset of $ \\mathbb{R}^{n}$. Is it true that $ K$ is always homeomorphic to a $ k$-dimensional closed disc?\r\n\r\nIf the interior of $ K$ is nonempty we can even state an [i]explicit[/i] homeormorphism between $ K$ and $ \\mathbb{D}^{n}$. What can be said in general? :)\r\n\r\nLinks and pdf's are welcome, too.",
"Solution_1": "The general case can be reduced to the special one:\r\nwlog we can assume that $ 0\\in K$. Now we consider the linear subspace $ V$ spanned by $ K$. It equals $ \\bigcup_{n\\in\\mathbb{N}} n(K\\minus{}K)$ as one can show. Now $ K$ has non-empty interior with respect to $ V$. So we can also assume that $ int K \\neq \\emptyset$ and if suffices prove that $ (K,\\partial K)\\cong (D^n,\\partial D^n)$. And this one seems to be clear to you if I understand you correctly.",
"Solution_2": "I have two questions on the notations in your post, Gockel.\r\n\r\n1. What do you exactly mean by [b]K-K[/b]?\r\n2. Could please rewrite the last line of your post? I'm not sure I followed you there... Especially when you say [i]if suffices prove that[/i]... Also, what do you mean by the pair [b](K,dK)[/b]?\r\n\r\nThanks again.",
"Solution_3": "By $ K\\minus{}K$ I meant the set of all differences: $ K\\minus{}K: \\equal{}\\{k\\minus{}k' | k,k'\\in K\\}$\r\n\r\nI wanted to write \"[i]it[/i] suffices [i]to[/i] prove that ...\"\r\nSorry. :blush: \r\n\r\nAnd the pair $ (K,\\partial K)$ means exactly what it says: the pair $ (K,\\partial K)$.\r\nA homeomorphism $ (K,\\partial K)\\to (D^n,\\partial D^n)$ is a homeomorphism of topological pairs, i.e. a bijective map $ f: K\\to D^n$ which is continouos in both directions and maps $ \\partial K$ to $ \\partial D^n$.\r\nNow that I think a bit more about the subject I think the construction is even stronger than that. There is a homeomorphism of $ \\mathbb{R}^n$ onto itself which maps $ K$ to $ D^n$ (and hence $ \\partial K$ to $ \\partial D^n$).",
"Solution_4": "[quote=\"Gockel\"] ... prove that $ (K,\\partial K)\\cong (D^n,\\partial D^n)$.[/quote]\r\n\r\nShouldn't it be $ \\displaystyle (\\mathrm{K}, \\partial \\mathrm{K}) \\cong (\\mathrm{D}^{\\dim \\mathrm{V}}, \\partial \\mathrm{D}^{\\dim \\mathrm{V}})$? ;)",
"Solution_5": "That's the same... $ V\\equal{}\\mathbb{R}^n$.",
"Solution_6": "[quote=\"Gockel\"]$ V \\equal{} \\mathbb{R}^n$.[/quote]\r\n\r\nAlways? What if, for example, $ K\\equal{}\\{0\\}$ and $ n\\equal{}1$? Does it follow that $ V \\equal{} \\langle K \\rangle \\equal{} \\mathbb{R}$? :(",
"Solution_7": "But $ \\{0\\}$ has empty interior. $ (K,\\partial K)\\cong(D^n,\\partial D^n)$ holds iff $ int(K)\\neq\\emptyset$.",
"Solution_8": "This is the post I'm referring to, dear [b]Gockel[/b]... ;)\r\n\r\n[quote=\"Gockel\"]The general case can be reduced to the special one:\nwlog we can assume that $ 0\\in K$. Now we consider the linear subspace $ V$ spanned by $ K$. It equals $ \\bigcup_{n\\in\\mathbb{N}} n(K \\minus{} K)$ as one can show. Now $ K$ has non-empty interior with respect to $ V$. So we can also assume that $ int K \\neq \\emptyset$ and if suffices prove that $ (K,\\partial K)\\cong (D^n,\\partial D^n)$. And this one seems to be clear to you if I understand you correctly.[/quote]",
"Solution_9": "I know... what is unclear about it? If $ K$ has non-empty interior then $ V\\equal{}\\mathbb{R}^n$ and $ (K,\\partial K)\\cong (D^n,\\partial D^n)$. That's exactly what I wrote...",
"Solution_10": "[quote=\"Gockel\"] What is unclear about it? If $ K$ has non-empty interior then $ V \\equal{} \\mathbb{R}^n$...[/quote]\r\n\r\nI thought Riquelme had told us to forget about that case... I also thought that in your first post you were taking K to be arbitrary, possibly with an empty interior. Weren't you trying to show us how to proceed in that [i]scenario[/i]?",
"Solution_11": "And I did...\r\nConstruct $ V$ as I did. Then $ K$ has non-empty interior in $ V$ and wlog we can assume that $ V\\equal{}\\mathbb{R}^n$. So we're again in the case which is clear to Riquelme."
}
{
"Tag": [
"algebra",
"polynomial",
"algebra unsolved"
],
"Problem": "(a) A polynomial $ P$ has degree at most $ n \\minus{} 10\\sqrt {n}$. Prove that\r\n\r\n$ |P(0)|\\leq\\frac {1}{10}(\\binom{n}{1}|P(1)| \\plus{} \\binom{n}{2}|P(2)| \\plus{} ... \\plus{} \\binom{n}{n}|P(n)|)$\r\n\r\n(b) Prove that for any sufficiently large integer $ n$, there exist a polynomial $ P$ with degree at most $ n\\minus{}10\\sqrt{n}$ such that\r\n\r\n$ |P(0)|>\\frac {1}{10}(\\binom{n}{1}|P(1)| \\plus{} \\binom{n}{2}|P(2)| \\plus{} ... \\plus{} \\binom{n}{n}|P(n)|)$",
"Solution_1": "$ a$ contradicts $ b$, doesn't it?",
"Solution_2": "Indeed, how stupid I am.\r\n\r\nCorrect version for (b):\r\n\r\n(b) Prove that for any sufficiently large integer $ n$, there exist a polynomial $ P$ with degree at most $ n \\minus{} \\frac{1}{10}\\sqrt {n}$ such that\r\n\r\n$ |P(0)| > \\frac {1}{10}(\\binom{n}{1}|P(1)| \\plus{} \\binom{n}{2}|P(2)| \\plus{} ... \\plus{} \\binom{n}{n}|P(n)|)$"
}
{
"Tag": [
"geometry",
"geometric transformation",
"rotation"
],
"Problem": "I have one doubt.All bodies in the universe from electrons and quarks to galaxies and stars are rotating about some axis.Why do all the bodies rotate?What sets them into rotation?Also some bodies rotate faster than others.Can anyone explain this?Is it due to the magnetic field possessed by the body?",
"Solution_1": "No replies? :(",
"Solution_2": "This is a very fundamental question, something perhaps no one can answer. It is akin to asking, \"Why do objects move? Once moving, why do they stay moving?\" They move because they are pushed into motion, just as things rotate because they are turned. Newton's first law states that things tend to stay moving when they are already moving, and Newton's first law for rotation says the same for rotating bodies. This is an experimental fact, deduced by observation, and cannot be derived from first principles.\r\n\r\nJust as some objects move faster than others, some can rotate faster than others. If they are turned harder, they will move faster. This is also an experimental finding, and gives rise to Newton's second law for rotation.\r\n\r\nRotation, in general, has nothing to do with magnetic fields. A non-charged, non-magnetic object can still rotate. The concept of rotation does not stem from magnetism; rather, it is the other way round, that rotation often generates magnetism due to moving charges.",
"Solution_3": "Then why do neutron stars and similar bodies with strong magnetic fields rotate so fast?",
"Solution_4": "You've got cause and effect mixed up, I think.\r\nNeutron stars rotate very fast, due to conservation of angular momentum. Hence, the moving charges inside them create extremely large magnetic fields."
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "Let $S_{n}$ be the summation :\r\n\r\n$S_{n}=\\sum_{i=1}^{n}\\frac{1}{n+i}$ and $n\\geq 2$\r\n\r\nProve that :\r\n\r\n[size=150]$S_{n}\\leq \\frac{n(3n+1)^{n-1}}{2^{n-1}.{A_{2n}^{n}}}$[/size]",
"Solution_1": "What is $A_{2n}^{n}$?",
"Solution_2": "$A_{n}^{a}=n(n-1)(n-2)...(n-a+1)$ for every integer $a\\leq n$",
"Solution_3": "Thank you for your prompt reply! :)"
}
{
"Tag": [
"calculus",
"integration",
"quadratics",
"number theory proposed",
"number theory",
"Diophantine equation"
],
"Problem": "Determine all non-negative integral pairs $ (x, y)$ for which\r\n\\[ (xy \\minus{} 7)^2 \\equal{} x^2 \\plus{} y^2.\\]",
"Solution_1": "let's rewrite the equation as $ 48\\plus{}(x^2\\minus{}1)(y^2\\minus{}1)\\equal{}14xy\\qquad (*)$.\r\n\r\nIf $ x\\equal{}0$ we get $ y\\equal{}7$. putting aside the solutions $ (0,7),(7,0)$ we can assume that $ x,y$ are positive.\r\n\r\nIf $ x\\equal{}1$ it turns out that $ y\\equal{}\\frac{14\\pm\\sqrt{61}}{3}\\not\\in \\mathbb{Z}$\r\n\r\nif $ x,y$ are both at least $ 5$ then $ x^2\\minus{}4x\\plus{}1$ and $ y^2\\minus{}4y\\plus{}1$ are both positive so from $ (*)$ it follows that\r\n\r\n$ 14xy\\equal{}48\\plus{}(x^2\\minus{}1)(y^2\\minus{}1)>4x\\cdot 4y\\equal{}16xy$ a contradiction.\r\n\r\nSo let's assume $ x\\in \\{3,4\\}$\r\n\r\nIf $ x\\equal{}3$ we get $ y\\in \\{4,\\frac54\\}$\r\n\r\nIf $ x\\equal{}4$ we get $ y\\in \\{3,\\frac{11}{5}\\}$\r\n\r\nSo in conclusion the only solutions are $ (x,y)\\in \\{(0,7),(7,0),(3,4),(4,3)\\}$",
"Solution_2": "$ (xy\\minus{}7)^2\\equal{}x^2\\plus{}y^2$\r\n\r\n$ x^2y^2\\plus{}49\\minus{}12xy\\equal{}(x\\plus{}y)^2$\r\n\r\n$ (xy\\minus{}6)^2\\plus{}13\\equal{}(x\\plus{}y)^2$\r\n\r\n$ 13\\equal{}(x\\plus{}y\\minus{}xy\\plus{}6)(x\\plus{}y\\plus{}xy\\minus{}6)$\r\n\r\nNow there are 4 cases, which solved give the solutions.",
"Solution_3": "Can anyone provide an approach using pythagorean triplets?",
"Solution_4": "Let us Rearrange the Equation as\n$x^2(y^2-1)-(14y).x+(49-y^2)=0$ which is a quadratic in $x$\nHence , $\\Delta =(2y^2-1)+195=t^2$ and the rest is trivial to figure out :lol:",
"Solution_5": "[quote=Rijul saini]Can anyone provide an approach using pythagorean triplets?[/quote]\n\nSeeing as no one gave a solution using Pythagorean triplets \nThe question undoubtedly boils down to finding the number of triangles for which $x,y and (xy-7)$ are sides of a given triangle \n\nNow, we may or may not know the algorithm given by pythagoras himself that states that the sides of the Pythagorean triangle are of the form $2n^2 + 2n, 2n+1 and 2n^2 + 2n + 1$ \nPlugging the given for the given values \nSuch that $x = 2n^2 + 2n$\n$y = 2n+1$\nAnd $(xy - 7) = 2n^2 + 2n +1$ \n\nWe find that $n^3 + n^2 = 2$ \nOr $n^2 \\cdot (n+1) = 2$ \nThis gives us $n$ as $1$ only (for positive integral values)\n\nThus this gives $x =4$\n$y= 3$ \nAnd thus $xy-7 = 5$\nThus since the pairs are ordered we have $(3,4)$ and $(4,3)$ \n\nAlso since they say \u201cnon negative\u201d \nPlugging in 0 in place of $x$ or $y$, we again get the solutions $(7,0)$ and $(0,7)$ \n\nHence the solutions are $(3,4),(4,3),(7,0)$ and $(0,7)$",
"Solution_6": "[quote=Syngates]Now, we may or may not know the algorithm given by pythagoras himself that states that the sides of the Pythagorean triangle are of the form $2n^2 + 2n, 2n+1 and 2n^2 + 2n + 1$[/quote]\n\nWhat about the Pythagorean triple $(8,15,17)$? This doesn't seem to be of that form.",
"Solution_7": "The approach using Pythagorean triples: \nLet assume $x , y > 0$, so $x=a^2-b^2; y=2ab; xy-7=a^2+b^2$, where $a>b>0$. Obviously $a \\ge 2, b \\ge 1$\nThus, $2ab(a^2-b^2)=a^2+b^2+7$\n$2ab = ab + ab \\ge a + b + 1$ and $a-b \\ge 1$\nSo $a^2+b^2+7=2ab(a-b)(a+b) \\ge (a+b+1)(a+b)$\n$7 \\ge 2ab + a + b \\ge 2ab + 2 + 1$\n$2 \\ge ab$\nThus $a=2, b=1$. This gives us $x=3, y=4$. ",
"Solution_8": "Storage \n\n\\begin{align*}&(xy-7)^2=x^2+y^2 \\\\ \\Longleftrightarrow \\;& x^2y^2-12xy+49=x^2+2xy+y^2 \\\\ \\Longleftrightarrow \\;& (x+y)^2-(xy-6)^2=13\\\\ \\Longleftrightarrow \\;& (x+y-xy+6)(x+y+xy-6)=13\\end{align*}\n\n$\\mathbf{Case\\; 1:}$\n$$ (x+y-xy+6)=13 \\quad \\quad \\text{and} \\quad \\quad (x+y+xy-6)=1$$which gives $x+y=7$ and $xy=0$ which then yields the solutions $(7,0),(0,7)$\n\n$\\mathbf{Case\\; 2:}$\n$$ (x+y+xy-6)=13 \\quad \\quad \\text{and} \\quad \\quad (x+y-xy+6)=1$$which gives $x+y=7$ and $xy=12$ which then yields the solutions $(4,3),(3,4)$\n\nHence all the solutions are $\\boxed{(7,0),(0,7),(4,3),(3,4)}$",
"Solution_9": "Our aim is two transform the equation to the desirable form.\\\\\n$(xy-7)^2=x^2+y^2 \\; \\Rightarrow \\; (xy)^2+49-14xy=x^2+y^2$\\\\\n$\\Rightarrow (xy)^2+49-12xy=x^2+y^2+2xy \\; \\Rightarrow \\; (xy-6)^2+13=(x+y)^2$ \\\\\n$\\Rightarrow 13 \\cdot 1=(x+y)^2-(xy-6)^2 \\; \\Rightarrow \\; 13 \\cdot 1=(x+y+xy-6)(x+y-xy+6)$ \\\\\nNow it becomes mere calculation... There are two cases to look on.\\\\\n$\\mathtt{Case\\; 1:}$\\\\\n$(x+y-xy+6)=13 \\quad \\quad \\; \\quad \\quad (x+y+xy-6)=1$\\\\\nSolving for $(x+y)$ and $(xy)$ gives us $7$ and $0$ respectively which henceforth yields $(0,7)$ and $(7,0)$ as Solution.\\\\\\\\\n$\\mathtt{Case\\; 2:}$\\\\\n$(x+y+xy-6)=13 \\quad \\quad \\; \\quad \\quad (x+y-xy+6)=1$\\\\\nSolving for $(x+y)$ and $(xy)$ gives us $7$ and $12$ respectively which henceforth yields $(3,4)$ and $(4,3)$ as Solutions.\\\\\\\\\nTherefore, $(3,4)$, $(4,3)$, $(0,7)$, $(7,0)$ are the integral Solutions to the equation.",
"Solution_10": "First, we rewrite with algebra: $(xy-7)^2=x^2+y^2 \\leftrightarrow x^2y^2-14xy+49=x^2+y^2 \\leftrightarrow (xy-6)^2-(x+y)^2=-13 \\leftrightarrow [xy-6-(x + y)][xy-6+(x + y)] = -13$.\n\n\nNow, we can break into cases:\n\n[i]Case 1 $(xy-x-y-6=-1, xy+x+y-6=13)$[/i]:\nAdding the equations and simplifying, we get $xy=12$. Subtracting the equations and simplifying, we get $x+y=7$. Solving the system, we get the solutions $(4,3), (3,4)$.\n\n[i]Case 2 $(xy-6-(x + y) =-13, xy-6+(x + y)=1)$[/i]:\nManipulating as above, we get $xy=0$ and $x+y=7$. Solving yields $(0,7), (7,0)$ as solutions.\n\nIn summary, our solutions are $\\boxed{(4,3),(3,4),(0,7),(7,0)}$.",
"Solution_11": "\\( (xy - 7)^2 = x^2 + y^2\\Rightarrow x^2y^2-14xy+49=x^2+y^2 \\Rightarrow x^2+y^2+12xy+36+13=x^2+y^2+2xy \\Rightarrow (xy-6)^2-(x+y)^2=-13\\)\n\\(\\Rightarrow (x+y+xy-6)(x+y-xy+6)=13\\). Let $(x+y+xy-6)=a$, $(x+y-xy+6)=b$. Thus we get 4 cases:\n\n$(i)$ Case $1$: Observe that $a<0 \\Leftrightarrow b<0$. If $a<0 \\Rightarrow b<0 \\Rightarrow a+b<0 \\Rightarrow 2(x+y)<0$, which is not possible as $x,y$ are non-negative integers. Thus, this case is not possible.\n\n$(ii)$ Case $2$: If $a=1$, $b=13$ $\\Rightarrow a+b=2(x+y)=14 \\Rightarrow (x+y) =7$, $a-b=2(xy-6)=-12 \\Rightarrow xy=0 \\Rightarrow x=0$ or $y=0$, yielding the solutions $(0,7)$ and $(7,0)$ from this case.\n\n$(iii)$ Case $3$: If $a=13$, $b=1$ $\\Rightarrow a+b=2(x+y)=14 \\Rightarrow (x+y)=7$, $a-b= 2(xy-6)=13 \\Rightarrow xy=12$. We form a polynomial $p(x_0)= x_0^2-7x_0+12=0$. The roots of this equation are $x$ and $y$. By factorising, $p(x_0)=(x_0-4)(x_0-3)$. Thus the roots are $4$ and $3$. We get one more pair of solutions: $(4,3)$ and $(3,4)$.\n\nSolutions: $\\boxed{(7,0),(0,7),(3,4),(4,3)}$"
}
{
"Tag": [
"\\/closed"
],
"Problem": "Does anyone know how to make a link to reveal hidden content?? I have seen numerous people do this, but I have no earthly idea how yet. \r\n\r\n\r\n :roll:",
"Solution_1": "Write (hide=text you want as the title) stuff you want hidden (/hide).\r\nExcept [] instead of (), becomes [hide=\"text you want as the title\"] stuff you want hidden [/hide].\r\nAlso, if you just go (hide) stuff you want hidden (/hide), the text will be \"Click to reveal hidden content.\"\r\n\r\nIf you want to do something you see someone else writing, just quote them to see the code. :D",
"Solution_2": "Please ask questions such as this in the questions forum, next time. Thanks.\r\n\r\nAnyway, to create hide codes, use the following code:\r\n\r\n[code][hide=text or smilies to click here]YOUR TEXT HERE\nMORE TEXT\nEVEN MORE TEXT[/hide][/code]",
"Solution_3": "does this feature work on other kinds of forums too? or just AoPS?\r\n\r\njust out of curiousity",
"Solution_4": "Depends on the forum, there is no general answer. Just try.",
"Solution_5": "Some forums have the features. Others don't. There is a PHPBB mod to create that, but most don't to answer your question. Some do. :D"
}
{
"Tag": [
"algebra",
"polynomial",
"induction",
"algebra unsolved"
],
"Problem": "determinate coefficient $x^{2}$ of polinomial $P(x) =(((...((x-2)^{2}-2)^{2}-....2)^{2}$",
"Solution_1": "Let $P_{0}(x)=x,P_{n+1}(x)=(P_{n}(x)-2)^{2}$. Then by induction $P_{n}(x)=4-4^{n}x+4^{n-1}\\frac{4^{n}-1}{3}x^{2}(mod \\ x^{3})$."
}
{
"Tag": [],
"Problem": "If M/N = 6/5,then 3M+2N = ?",
"Solution_1": "[hide=\" Have I missed something?\"]\n\n$ \\frac{M}{N}\\equal{}\\frac{6}{5}$\n\n$ 5M\\equal{}6N$\n\n$ 2N\\equal{} \\frac{5M}{3}$\n\n$ 3M\\plus{}2N\\equal{} 3M \\plus{} \\frac{5M}{3}\\equal{} \\frac{14M}{3}$.\n\n[/hide]",
"Solution_2": "hello, we get\r\n$ M\\equal{}\\frac{6}{5}N$\r\nand plugging this in the term above we get\r\n$ \\frac{18}{5}N\\plus{}\\frac{10}{5}N\\equal{}\\frac{28}{5}N$\r\nSonnhard.",
"Solution_3": "$ \\frac {14M}{3} \\equal{} \\frac {14}{3}\\cdot \\frac{6N}{5} \\equal{} \\frac {28N}{5}$. \r\n\r\nI still think there's something sri340 forgot to tell us."
}
{
"Tag": [
"algebra",
"polynomial"
],
"Problem": "Prove $(1^{k}+2^{k}+...+n^{k})$ is divisible by $\\frac{n(n+1)}{2}$ where $k$ is an odd positive integer.\r\n\r\n[I posted this earlier under a different post here but didn't get any response.]",
"Solution_1": "We have\r\n\r\n\\begin{eqnarray*}2(1^{k}+2^{k}+\\dots+n^{k}) &=& (1^{k}+n^{k})+(2^{k}+(n-1)^{k})+\\dots+((n-1)^{k}+2^{k})+(n^{k}+1^{k})\\\\ &=& (n+1)p_{1}(n)+(n+1)p_{2}(n)+\\dots+(n+1)p_{n}(n)\\\\ &=& (n+1)p(n)\\end{eqnarray*}\r\n\r\nwhere $p_{i}(n)$ are some polynomials and $p(n)$ is their sum. (We're using the fact that $a+b|a^{m}+b^{m}$ for an odd natural $m$.) Therefore, $2(1^{k}+2^{k}+\\dots+n^{k})$ is always divisible by $n+1$. Subsequently, $2(1^{k}+\\dots+(n-1)^{k})$ is always divisible by $n$, meaning that $2(1^{k}+\\dots+(n-1)^{k})+2n^{k}$ is also divisible by $n$. Hence, $2(1^{k}+\\dots+n^{k})$ is divisible by both $n$ and $n+1$, but as those two factors are coprime, it must be also divisible by their product:\r\n\r\n$2(1^{k}+\\dots+n^{k})=n(n+1)q(n)$\r\n\r\nwhere $q(n)$ is some polynomial, but since $n(n+1)$ is always even, we can write\r\n\r\n$1^{k}+\\dots+n^{k}=\\frac{n(n+1)}{2}q(n)$, hence the conclusion.",
"Solution_2": "You can easily derive that \\[\\sum_{i=1}^{n}i^{k}=\\sum_{i=1}^{k}S(k,i)\\binom{n+1}{i+1}i!\\] Then you easily see that $\\frac{n(n+1)}{2}| \\sum_{i=1}^{k}S(k,i) \\binom{n+1}{i+1}i!$.\r\n\r\nQ.E.D.\r\n\r\nMasoud Zargar",
"Solution_3": "[quote=\"boxedexe\"]You can easily derive that \\[\\sum_{i=1}^{n}i^{k}=\\sum_{i=1}^{k}S(k,i)\\binom{n+1}{i+1}i!\\] Then you easily see that $\\frac{n(n+1)}{2}| \\sum_{i=1}^{k}S(k,i) \\binom{n+1}{i+1}i!$.\n\nQ.E.D.\n\nMasoud Zargar[/quote]What is the notation $S(k,i)$?",
"Solution_4": "Stirling number of the second type. $S(n,k)$ is the number of ways to partition an n-set into k parts.",
"Solution_5": "[quote=\"boxedexe\"]You can easily derive that \\[\\sum_{i=1}^{n}i^{k}=\\sum_{i=1}^{k}S(k,i)\\binom{n+1}{i+1}i!\\] Then you easily see that $\\frac{n(n+1)}{2}| \\sum_{i=1}^{k}S(k,i) \\binom{n+1}{i+1}i!$.\n\nQ.E.D.\n\nMasoud Zargar[/quote]\r\n[b]boxedexe,[/b] I cannot \"easily\" derive what you just wrote and I am sure very few people in this community can derive the same easily unless they have seen the formula beforehand, in which case the formula is obviously not easy.\r\n\r\nAlso, when you talk about Stirling numbers, I would appreciate it if you direct others like me to an online article that can explain what these numbers are. Without such help, it becomes very hard for me to comprehend what you write.",
"Solution_6": "Well then I am sorry for not doing so. \r\n\r\n1. [url=http://mathworld.wolfram.com/StirlingNumberoftheSecondKind.html]Stirling Numbers of the Second Kind[/url]\r\n2. The proof of the formula is counting in two ways; trying to represent the number of k+1-tuples such that the k+1-th term is $x_{k+1}>\\max_{i\\max_{irank(A^{n}),rank(B^{n\\minus{}1})>rank(B^n),AB\\equal{}BA.$\r\nProve that : $ (A\\plus{}B)^n\\equal{}0$\r\n--------- :P ----------",
"Solution_1": "$ dim(ker(A^{n\\minus{}1}))rank(A^{n}),rank(B^{n\\minus{}1})>rank(B^n)$ (well, loup blanc didn't use them either).\r\n\r\n darij",
"Solution_6": "the rank considerations were necessary to prove that a,b were nilpotent. The rest of the problem follows as you pointed out darij",
"Solution_7": "note to self: buy glasses\r\n\r\n darij",
"Solution_8": "Also Darij, it is not enough to prove that A+B is nilpotent, we must have it be nilpotent of order n. I'm pretty self taught and the only proof I know that commutative rings admit a nilradical doesn't preserve orders: i.e., if A^n=B^n=0 then A+B is nilpotent but not necessarily (A+B)^n=0\r\n\r\n\r\nedit: i just looked up the \"standard\" proof of the existence of the nilradical for the first time and it too only shows that A^n=B^m=0 implies (A+B)^(n+m-1)=0\r\n\r\nSo, your cited result doesn't make progress on the problem unless you are able to prove that if A+B is nilpotent, then its order is not greater than n.",
"Solution_9": "Hi me@home,\r\na free lesson for you:\r\nlet $ A\\in\\mathcal{M}_n(K)$ s.t. $ A^k\\equal{}0$; then $ spectrum(A)\\equal{}\\{0\\}$ . We can triangularize $ A$ over $ K$. Then we may assume that $ A$ is a strictly upper triangular matrix. Now calculate $ A^2,A^3,\\cdots$ and you will see that $ A^n\\equal{}0$.",
"Solution_10": "me@home: it seems you need new glasses, too ;)\r\n\r\n[quote=\"darij grinberg\"]If a $ n\\times n$ matrix $ X$ is nilpotent, then $ X^n \\equal{} 0$ (this follows from rank considerations, or from Cayley-Hamilton and minimal polynomials, or is simply well-known).[/quote]\r\n\r\nloup blanc: triangulation is great, but let us kill harder problems using it ;)\r\n\r\n darij"
}
{
"Tag": [
"inequalities",
"symmetry"
],
"Problem": "For $a,b,c,d \\in \\mathbb{R}$ such that \\[ a^2+b^2+c^2+d^2=1 \\]\r\nProve that \\[ a(1+d-a)+b(1+d-b)+c(1+d-c)-d-1 \\leq 0 \\]",
"Solution_1": "We are to prove that $a+b+c+d\\leq 2.$",
"Solution_2": "[hide]The LHS is equivalent to $(d+1)(a+b+c-1)-a^2-b^2-c^2$. This, in turn, can be simplified to $(d+1)(a+b+c-1)+d^2-1=(d+1)(a+b+c+d-2)$. By Cauchy-Schwartz, $(a^2+b^2+c^2+d^2)*4\\ge (a+b+c+d)^2$ so $2\\ge |a+b+c+d|$. Then subtracting two from a+b+c+d we have a nonpositive. However, since $|d|\\le 1$, d+1 is nonnegative, so the given inequality must be true.[/hide]",
"Solution_3": "Here is another approach of the proof for $a+b+c+d\\leq 2.$\r\n\r\n$\\left(a-\\frac{1}{2}^\\right)^2+\\left(b-\\frac{1}{2}^\\right)^2+\\left(c-\\frac{1}{2}^\\right)^2+\\left(d-\\frac{1}{2}^\\right)^2\\geq 0.$\r\n\r\n$\\Longleftrightarrow a^2+b^2+c^2+d^2-(a+b+c+d)+1\\geq 0.$\r\n\r\n$\\Longleftrightarrow a+b+c+d\\leq 2.\\ Q.E.D.$",
"Solution_4": "Nice\r\n\r\nOk kunny, maybe this indeed was a bit easy, like you said. What about this? (another self-made one.) :D \r\n\r\nFor $a,b,c \\in \\mathbb{R}^+$ with $a\\geq b \\geq c$\r\nso that $a+b+c=1$\r\n\r\nProve that $ab(3a-1)+ac(3b-1)+bc(3c-1) \\geq 0$",
"Solution_5": "Thanks, I will prove that. :)",
"Solution_6": "[quote=\"Jan\"]\nFor $a,b,c \\in \\mathbb{R}$ so that $a+b+c=1$\nProve that $ab(3a-1)+ac(3b-1)+bc(3c-1) \\geq 0$[/quote]\r\n\r\nJan, sorry are you sure this inequality? I think may be $ab(3a-1)+bc(3b-1)+ca(3c-1) \\geq 0.$\r\n\r\nkunny",
"Solution_7": "No, I'm pretty sure about my inequality, but there may be indeed a problem with it. \r\nI forgot an extra thing: $a \\geq b \\geq c$.\r\nI'm not very familiar with these sort of inequalities, so it's possible that this is not needed in the inequality, but just to be sure, i'll add it in the question.\r\nSorry",
"Solution_8": "No problem, Jan. I thought your ineqality hasn't the symmetry of $a,b,c.$\r\nThat is to say $ab\\rightarrow bc\\rightarrow ca.$\r\n\r\nkunny",
"Solution_9": "Yes, I know, it's not symmetric, so you can not assume that $a \\geq b \\geq c$, thus it needs to be given. \r\nPity, now there is a hint in the question, else it would be much more difficult. :P\r\n\r\nanyone has any ideas for solving this problem? I'm very curious.",
"Solution_10": "That's true Arne, but it's not necessary.\r\nI had something else in mind when I created the problem. :D",
"Solution_11": "[hide=\"my solution\"]\nExpand and re-arrange the inequality to obtain\n$-a^2 - b^2 - c^2 + (d + 1)a + (d + 1)b + (d+1)c - (d + 1) \\leq 0$\nNow substitute $a^2 + b^2 + c^2 = 1 - d^2$ to get\n$d^2 - 1 + (d + 1)a + (d + 1)b + (d+1)c - (d + 1) \\leq 0$\ndividing through by $d + 1$, (which we can do WLOG because $d \\geq -1$, and the inequality is obviously satisfied if $d=1$) we get\n$d - 1 + a + b + c - 1 \\leq 0 \\Rightarrow a + b + c + d \\leq 2$\nApplying Cauchy-Schwartz, we get\n$(a^2 + b^2 + c^2 + d^2)(4) \\geq (a + b + c + d)^2$\nand the result follows.\n[/hide]",
"Solution_12": "[quote=\"Jan\"]Nice\n\nOk kunny, maybe this indeed was a bit easy, like you said. What about this? (another self-made one.) :D \n\nFor $a,b,c \\in \\mathbb{R}$ with $a\\geq b \\geq c$\nso that $a+b+c=1$\n\nProve that $ab(3a-1)+ac(3b-1)+bc(3c-1) \\geq 0$[/quote]\r\nI don't think this inequality holds. For example, try $a = 3, b = -1, c = -1$, and you find $ab(3a-1)+ac(3b-1)+bc(3c-1) = -16 < 0$.",
"Solution_13": "?!? Wow, you're right. It's wrong. Can somebody tell me what's wrong with my method?\r\n\r\nChebyshev inequality tells us that \\[ \\frac{a+b+c}{3}.\\frac{ab+ac+bc}{3} \\leq \\frac{a^2b+abc+bc^2}{3} \\]\r\n\\[ \\iff \\frac{1}{3}.\\frac{ab+ac+bc}{3}\\leq \\frac{a^2b+abc+bc^2}{3} \\]\r\n\\[ \\iff ab+ac+bc \\leq 3a^2b+3abc+3bc^2 \\]\r\n\\[ \\iff 0 \\leq ab(3a-1) + ac(3b-1) + bc(3c-1) \\]\r\n\r\nLooks right to me... As far as I know Chebyshev holds for real numbers.",
"Solution_14": "First I also thought the use of Chevichev's inequality, but why can we use that?",
"Solution_15": "[quote=\"Jan\"]?!? Wow, you're right. It's wrong. Can somebody tell me what's wrong with my method?\n\nChebyshev inequality tells us that...[/quote]\r\nChebyshev's inequality only applies with two series that both have $a_1 \\geq a_2 \\geq ... \\geq a_n$. So, for that inequality to hold, you'd have to have $ab \\geq ac \\geq bc$, which doesn't necessarily follow from $a \\geq b \\geq c$. As you'll note, my counter-example doesn't follow this inequality, and so Chebyshev's Inequality breaks down.\r\n\r\nI believe if you added the condition that $a,b,c \\geq 0$, then your inequality would hold.",
"Solution_16": "I agree with you.",
"Solution_17": "You are right zanttrang, I didn't look very careful when I thought $ab \\geq ac \\geq bc$\r\nOf course it is not always true.\r\nThis is something I should have seen. But the inequality isn't lost, I only have to say that $a,b,c$ are positive. Then the inequality holds and the problem is solvable.\r\n\r\nSorry all, I corrected the question.\r\nBut, now the part I was curious of, does anyone have proofs without Chebyshev? \r\n\r\nFor example, as Arne suggested, homogenizing:\r\n\r\nWe rewrite the equality as \\[ 3a^b+3abc+3bc^2 \\geq ab+bc+ac \\]\r\nNow we homogenize: \\[ 3a^b+3abc+3bc^2 \\geq ab+bc+ac (a+b+c) = a^2b+ab^2 +3abc + b^2c+bc^2+a^2c+ac^2 \\]\r\nSo we get \\[ 2a^2b+2bc^2 \\geq ab^2 + b^2c+ac^2+a^2c \\]\r\n\r\nHow do we prove that's true? \r\n\r\nAny other solutions?",
"Solution_18": "[quote=\"Jan\"]\nSo we get \\[ 2a^2b+2bc^2 \\geq ab^2 + b^2c+ac^2+a^2c \\]\n\nHow do we prove that's true? \n\nAny other solutions?[/quote]\r\nSplit this up in the following way:\r\n\r\n$(ba^2 + bc^2 - ab^2 - ac^2) +(ba^2 + bc^2 - ca^2 - cb^2) \\geq 0$\r\n\r\nWe can prove that each of these two terms is greater than or equal to zero separately, through two applications of the re-arrangement inequality. Note that by the re-arrangement inequality,\r\n\r\n$a(ab) + b(c^2) \\geq b(ab) + a(c^2)$\r\n\r\nand\r\n\r\n$b(a^2) + c(bc) \\geq c(a^2) + b(bc)$\r\n\r\nSo, both of the terms are greater than or equal to zero, and our inequality is proven (this wouldn't work if some of the terms were negative, because I rely on the fact that $ab \\geq c^2$). This is the first time I've actually used the re-arrangement inequality, so I may have used it incorrectly :blush: \r\nAlso, because Chebyshev's Inequality is proven using the re-arrangement inequality, this is probably functionally equivalent to using Chevyshev.",
"Solution_19": "[quote]$(ba^2 + bc^2 - ab^2 + ac^2) +(ba^2 + bc^2 - ac^2 - cb^2) \\geq 0$[/quote]\r\nI'm sorry, but this is not the same, you should have $(ba^2 + bc^2 - ab^2 - a^2c) +(ba^2 + bc^2 - ac^2 - cb^2) \\geq 0$",
"Solution_20": "[quote=\"Jan\"][quote]$(ba^2 + bc^2 - ab^2 + ac^2) +(ba^2 + bc^2 - ac^2 - cb^2) \\geq 0$[/quote]\nI'm sorry, but this is not the same, you should have $(ba^2 + bc^2 - ab^2 - a^2c) +(ba^2 + bc^2 - ac^2 - cb^2) \\geq 0$[/quote]\r\nI made a typesetting error - the second $ac^2$ should really be $a^2 c$, and the first $ac^2$ should have a negative sign in front of it; then my solution should proceed normally (I should have checked what I wrote more closely - however, I believe I used the right form in the rest of my solution).",
"Solution_21": "But in fact, your idea to use rearrangement is good.\r\n\r\nSo i'll write down your idea for a solution: \r\n\\[ 2a^2b+2bc^2 \\geq ab^2 + b^2c+ac^2+a^2c \\]\r\n\r\nBy rearrangement, we get: \\[ a(ab)+b(c^2) \\geq ac^2 + ab^2 \\]\r\nAnd: \\[ a^2(b) + bc (c) \\geq b^2c + a^2c \\]\r\nAdding these two inequalities gives us \\[ 2a^2b+2bc^2 \\geq ab^2 + b^2c+ac^2+a^2c \\]\r\n\r\nQED :D \r\n\r\nOther ways to prove it?"
}
{
"Tag": [
"geometry",
"geometry solved"
],
"Problem": "can enyone tell me what is the fermat point in geometry.\r\nthanks for helping.",
"Solution_1": "Do you mean the Fermat point?\r\n\r\nCan someone move this to the appropriate section?"
}
{
"Tag": [
"calculus",
"MIT",
"college"
],
"Problem": "If the college I want to go to does not recognize any AP credits, is it worth it to spend time self-studying one or two AP subjects and then take the exams?",
"Solution_1": "Yes, it is worth it to learn something and show that you did.",
"Solution_2": "Thanks! :lol:",
"Solution_3": "if you don't know for sure that you will end going to that college, it's worth taking APs in case you end up going somewhere else that does give AP credit. Also, some colleges use AP scores for placement (esp calc), even if they don't use them for credit.",
"Solution_4": "also, APs look good while you're applying- they show you've taken the initiative to take harder classes when you didn't have to!",
"Solution_5": "Thanks for all the help! :) \r\n\r\nI still got one question here: \r\nMy school only offers AP Calculus and not any other courses. I am planning to take the AP Statistics exam this year and then AP Biology in Gr. 12. If I am not enrolled in an AP class, how can I show them that I am taking the AP Biology exam in May?",
"Solution_6": "Colleges will want you to list your current courses when applying, so you can list \"AP biology by self-study\" if you like.",
"Solution_7": "The other nice thing about AP classes is the GPA boost. I'm going to take as many as possible... I [b]need[/b] that GPA boost... :D",
"Solution_8": "[quote]The other nice thing about AP classes is the GPA boost. I'm going to take as many as possible... I need that GPA boost... [/quote]\r\n\r\nCan anyone please explain how AP classes help in universtiy GPA boost?\r\n\r\nAlso, I am planing to take AP Calculus and may be AP Statistics next year when I am in grade 12. You think those two AP credits will make it easier for me to get into a well known university like MIT? A lot of students that I know take AP courses in grade 11 while I am planning to take them in grade 12, and also I am only going to take 2 AP course; do i need to take more AP courses to get into MIT?",
"Solution_9": "How high school grade-point averages are calculated depends on the arbitrary policies of each high school. In SOME high schools, a grade of A in an AP course counts for more than a grade of A in some ordinary course. Other high schools do not have this policy. If yours does not, don't worry about it. College admission officers have their own separate ways of evaluating an applicant's high school record to see if an applicant took challenging courses in high school. Each college may have a slightly different way of doing this, and no college I am aware of publishes its methodology for comparing high school courses.",
"Solution_10": "[quote=\"satyam\"][quote]The other nice thing about AP classes is the GPA boost. I'm going to take as many as possible... I need that GPA boost... [/quote]\n\nCan anyone please explain how AP classes help in universtiy GPA boost?\n\nAlso, I am planing to take AP Calculus and may be AP Statistics next year when I am in grade 12. You think those two AP credits will make it easier for me to get into a well known university like MIT? A lot of students that I know take AP courses in grade 11 while I am planning to take them in grade 12, and also I am only going to take 2 AP course; do i need to take more AP courses to get into MIT?[/quote]\r\n\r\n\r\nLast year my son attended a meeting with several recruiters from top colleges. I was talking to an Alumni recruiter from MIT. He said that three AP scores were baseline among applicants, and they even had applicants with 20+ AP Scores. I shot back with \"Do you have any idea how elitist that is??\" He immediately backed down - and told me that a kid who only has a couple because their school only offers a couple will look just as good as a kid who has many because they go to an excellent prep school. \r\n\r\nBTW, if you are taking Calc BC, that counts as two - you get an AB and BC score.\r\n\r\nW/R/T to the GPA boost - on a scale of 0-4 GPA, you get an extra point for an AP class, so if you got an A in an AP course, it is basically 5 out of 4. I can average out that B in English :-)",
"Solution_11": "[quote=\"lfm\"]\nBTW, if you are taking Calc BC, that counts as two - you get an AB and BC score.\n[/quote]\r\n\r\nI don't think this is true... you do get an AB subscore, but AFAIK it does not count as an [i]additional[/i] AP credit. Not that the sheer number of AP credits is a big deal. Also, I'm pretty sure most top colleges only accept BC anyway."
}
{
"Tag": [
"number theory",
"prime factorization"
],
"Problem": "Find all composite integers M such that the product of 2005xM has exactly eight divisors",
"Solution_1": "[hide=\"hint\"]\n$2005=5*401$ thus it has $4$ devisors itself.\nnow, find all numbers in a form of :$M=pq$ \n\nedit: but doesn't this have infinite number of answers?\nyou can choose any two primes to be $p,q$\n[/hide]",
"Solution_2": "[hide]Well any prime which isn't already a factor would work because that would give us 2*2*2 factors, but it has to be composite, so...\n$25$ or $160801$\nThis would make the prime factorization $5^3*401$ or $5*401^3$ making the number of factors 2*4=8[/hide]",
"Solution_3": "oh, I didnt realize those were the same problem. My solution is in the other thread if anyone wants to read it :D"
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Books Pham Kim Hung - Secrets in Inequalities \r\nPages 23\r\n\r\n\r\n$ a^2b^2 \\plus{} b^2c^2 \\plus{} c^2a^2 \\leq2$\r\nEquality holds for $ a \\equal{} b \\equal{} 1,c \\equal{} 0$\r\n\r\nBut...\r\nwhen $ a \\equal{} b \\equal{} 1,c \\equal{} 0$\r\n$ a^2b^2 \\plus{} b^2c^2 \\plus{} c^2a^2 \\equal{} 1$\r\n\r\nI think it can't take $ 2$\r\n\r\n$ a \\plus{} b \\plus{} c \\equal{} 2$\r\nwe assume that $ max(a,b,c) \\equal{} a$\r\n$ a^2b^2 \\plus{} b^2c^2 \\plus{} c^2a^2 \\leq a^2b^2 \\plus{} 2a^2bc \\plus{} a^2c^2 \\equal{} a^2(b \\plus{} c)^2 \\equal{} a^2(2 \\minus{} a)^2$\r\n\r\n$ \\leq (\\frac {a \\plus{} b \\plus{} c}{2})^2 \\equal{} 1$",
"Solution_1": "Written inequality is obviously wrong. You just have to put 1 instead 2 on the RHS. \r\nIn thesolution you gave us it's also proved that it's lower or equal than 1 but not 2. :wink:"
}
{
"Tag": [
"combinatorics proposed",
"combinatorics"
],
"Problem": "Six distinct symbols are transmitted through a communication channel. A total of 18 banks are to be inserted between the symbols with at least two blanks between every pair of symbols. In how many ways can the symbols and blank arranged?",
"Solution_1": "[hide=\"Hint1\"] the problem is equivalent to: Six distinct symbols are transmitted through a communication channel. A total of EIGHT blanks are to be inserted between the symbols with at least ZERO blanks between every pair of symbols. [/hide]\n[hide=\"Hint2\"] Equal to the number of ways of arranging 8 blanks and 4 separations [/hide]"
}
{
"Tag": [],
"Problem": "\u03a4\u03b5\u03bb\u03b5\u03af\u03c9\u03c3\u03b5 \u03b7 \u03c6\u03b5\u03c4\u03b9\u03bd\u03ae \u039c\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03c7\u03c1\u03bf\u03bd\u03b9\u03ac \u03ba\u03b1\u03b9 \u03ba\u03b1\u03bb\u03ac \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03b3\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03ad\u03bd\u03b1\u03c2 \u03b1\u03c0\u03bf\u03bb\u03bf\u03b3\u03b9\u03c3\u03bc\u03cc\u03c2 \u03ba\u03b1\u03b9 \u03b7 \u03b1\u03c0\u03b1\u03b9\u03c4\u03bf\u03cd\u03bc\u03b5\u03bd\u03b7 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03c0\u03c1\u03bf\u03ba\u03b5\u03b9\u03bc\u03ad\u03bd\u03bf\u03c5 \u03bd\u03b1 \u03b4\u03b9\u03bf\u03c1\u03b8\u03c9\u03b8\u03bf\u03cd\u03bd \u03ba\u03ac\u03c0\u03bf\u03b9\u03b1 \u03bb\u03ac\u03b8\u03b7 \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03b3\u03af\u03bd\u03bf\u03c5\u03bc\u03b5 \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03bf\u03b9 . \r\n\u039f \u03ba.\u0391\u03bb\u03b5\u03be\u03b1\u03bd\u03b4\u03c1\u03ae\u03c2 (\u03c4\u03ad\u03c9\u03c2 \u03c0\u03c1\u03cc\u03b5\u03b4\u03c1\u03bf\u03c2 \u03c4\u03b7\u03c2 \u0395\u039c\u0395) \u03c3\u03b5 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b1 \u03bf\u03bc\u03b9\u03bb\u03af\u03b1 \u03c4\u03bf\u03c5 \u03b5\u03af\u03c0\u03b5 \u03cc\u03c4\u03b9 \u03b5\u03bc\u03b5\u03af\u03c2 \u03b1\u03c3\u03c7\u03bf\u03bb\u03bf\u03cd\u03bc\u03b1\u03c3\u03c4\u03b5 \u00ab\u03b5\u03c1\u03b1\u03c3\u03b9\u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac\u00bb \u03bc\u03b5 \u03c4\u03bf\u03c5\u03c2 \u039c\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03bf\u03cd\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2 \u03c3\u03b5 \u03b1\u03bd\u03c4\u03af\u03b8\u03b5\u03c3\u03b7 \u03bc\u03b5 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b5\u03c2 \u03ac\u03bb\u03bb\u03b5\u03c2 \u03c7\u03ce\u03c1\u03b5\u03c2 .\r\n\u0388\u03c4\u03c3\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 .\u038c\u03c7\u03b9 \u03b1\u03c0\u03bb\u03ac \u00ab\u03b5\u03c1\u03b1\u03c3\u03b9\u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac\u00bb \u03b1\u03bb\u03bb\u03ac \u03b8\u03b1 \u03ad\u03bb\u03b5\u03b3\u03b1 \u03ba\u03b1\u03b9 \u03bb\u03af\u03b3\u03bf \"\u03ba\u03bf\u03c1\u03bf\u03b9\u03b4\u03b5\u03c5\u03c4\u03b9\u03ba\u03ac\" \r\n\r\n\u039a\u03b1\u03b9 \u03b5\u03be\u03b7\u03b3\u03bf\u03cd\u03bc\u03b1\u03b9 .\r\n\r\n1) \u03a3\u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03bf\u03c5 \u03c3\u03c5\u03bc\u03bc\u03b5\u03c4\u03ad\u03c7\u03bf\u03c5\u03bd \u03c3\u03c4\u03bf\u03c5\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2 (\u03b5\u03c0\u03af \u03c4\u03b7 \u03b5\u03c5\u03ba\u03b1\u03b9\u03c1\u03af\u03b1 \u03b8\u03ad\u03bb\u03c9 \u03bd\u03b1 \u03c3\u03c5\u03b3\u03c7\u03b1\u03c1\u03ce \u03cc\u03bb\u03b1 \u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03bf\u03c5 \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03b1\u03bd \u03ba\u03b1\u03b9 \u03c4\u03b9\u03c2 \u03b4\u03cd\u03bf \u03b5\u03b8\u03bd\u03b9\u03ba\u03ad\u03c2 \u03bc\u03b1\u03c2 \u03bf\u03bc\u03ac\u03b4\u03b5\u03c2 \u03b3\u03b9\u03b1\u03c4\u03af \u03c7\u03c9\u03c1\u03af\u03c2 \u03ba\u03b1\u03bc\u03af\u03b1 \u03b2\u03bf\u03ae\u03b8\u03b5\u03b9\u03b1 \u03ba\u03b1\u03c4\u03ac\u03c6\u03b5\u03c1\u03b1\u03bd \u03b1\u03c5\u03c4\u03ac \u03c0\u03bf\u03c5 \u03ba\u03b1\u03c4\u03ac\u03c6\u03b5\u03c1\u03b1\u03bd) \u03b4\u03b5\u03bd \u03b4\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u0395\u039c\u0395 \u03ba\u03b1\u03b8\u03cc\u03bb\u03bf\u03c5 \u03b2\u03b9\u03b2\u03bb\u03b9\u03bf\u03b3\u03c1\u03b1\u03c6\u03af\u03b1 .\r\n\u0391\u03c0\u03bf\u03c1\u03af\u03b1\u03c2 \u03ac\u03be\u03b9\u03bf , \u03b3\u03b9\u03b1\u03c4\u03af \u03c4\u03bf \u03ba\u03ac\u03bd\u03bf\u03c5\u03bd .\r\n\u0394\u03b5\u03bd \u03be\u03ad\u03c1\u03bf\u03c5\u03bd \u03c4\u03af\u03c4\u03bb\u03bf\u03c5\u03c2 \u03b2\u03b9\u03b2\u03bb\u03af\u03c9\u03bd ? \r\n\u0394\u03b5\u03bd \u03b8\u03ad\u03bb\u03bf\u03c5\u03bd \u03bd\u03b1 \u03b4\u03ce\u03c3\u03bf\u03c5\u03bd ? \r\n\u03a4\u03b1 \u03b4\u03af\u03bd\u03bf\u03c5\u03bd \u03c3\u03b5 \u03bb\u03af\u03b3\u03bf\u03c5\u03c2 ? \r\n\r\n2) \u0394\u03b5\u03bd \u03b3\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03bc\u03b1\u03b8\u03ae\u03bc\u03b1\u03c4\u03b1 .\u0391\u03c5\u03c4\u03ac \u03c4\u03b1 \u03bb\u03af\u03b3\u03b1 \u03c0\u03bf\u03c5 \u03b3\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c3\u03ae\u03bc\u03b1\u03bd\u03c4\u03b1 (\u03b3\u03b9\u03b1 \u03bd\u03b1 \u03c0\u03bf\u03c5\u03bd \u03cc\u03c4\u03b9 \u03ba\u03ac\u03c4\u03b9 \u03ba\u03ac\u03bd\u03bf\u03c5\u03bd) .\r\n\u03a3\u03b5 \u03ac\u03bb\u03bb\u03b5\u03c2 \u03c7\u03ce\u03c1\u03b5\u03c2 \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd trainers \u03ba\u03b1\u03b9 \u03c4\u03b1 \u03bc\u03b1\u03b8\u03ae\u03bc\u03b1\u03c4\u03b1 \u03b3\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03cc\u03bb\u03b7 \u03c4\u03b7 \u03c7\u03c1\u03bf\u03bd\u03b9\u03ac \u03b4\u03b9 \u03b1\u03bb\u03bb\u03b7\u03bb\u03bf\u03b3\u03c1\u03b1\u03c6\u03af\u03b1\u03c2 (!!!) .\r\n\u0397 \u03c0\u03c1\u03cc\u03c4\u03b1\u03c3\u03b7 \u03c0\u03bf\u03c5 \u03ad\u03ba\u03b1\u03bd\u03b5 \u03bf nickolas (\u03bc\u03b5 \u03b5\u03c0\u03b9\u03c3\u03c4\u03bf\u03bb\u03ae \u03c0\u03bf\u03c5 \u03ad\u03c3\u03c4\u03b5\u03b9\u03bb\u03b5 \u03c3\u03c4\u03bf\u03bd \u00ab\u03ac\u03c6\u03b1\u03bd\u03c4\u03bf\u00bb \u03c0\u03c1\u03cc\u03b5\u03b4\u03c1\u03bf \u03c4\u03b7\u03c2 \u0395\u039c\u0395 \u03bf\u03cd\u03c4\u03b5 \u03c3\u03c4\u03b7\u03bd \u03c4\u03b5\u03bb\u03b5\u03c4\u03ae \u03c4\u03bf\u03c5 \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 \u03ae\u03c4\u03b1\u03bd \u03c0\u03b1\u03c1\u03ce\u03bd !!) \u03bd\u03b1 \u03b1\u03bd\u03b1\u03bb\u03ac\u03b2\u03b5\u03b9 \u03c4\u03b7\u03bd \u03c0\u03c1\u03bf\u03b5\u03c4\u03bf\u03b9\u03bc\u03b1\u03c3\u03af\u03b1 \u03bf Vornicu \u03b4\u03b5\u03bd \u03ad\u03b3\u03b9\u03bd\u03b5 \u03b4\u03b5\u03ba\u03c4\u03ae (\u03c6\u03b1\u03bd\u03c4\u03ac\u03b6\u03bf\u03bc\u03b1\u03b9 \u03cc\u03c4\u03b9 \u03bf\u03cd\u03c4\u03b5 \u03bc\u03c0\u03ae\u03ba\u03b1\u03bd \u03c3\u03c4\u03bf\u03bd \u03ba\u03cc\u03c0\u03bf \u03bd\u03b1 \u03c4\u03bf\u03bd \u03b1\u03c0\u03b1\u03bd\u03c4\u03ae\u03c3\u03bf\u03c5\u03bd) .\r\n\u03a4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03ac\u03bd\u03b5 \u03c3\u03c4\u03bf\u03c5\u03c2 \u03b4\u03b9\u03b5\u03b8\u03bd\u03b5\u03af\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2 \u03ad\u03c7\u03bf\u03bd\u03c4\u03b1\u03c2 \u03c4\u03b9\u03c2 \u03b5\u03c5\u03c7\u03ad\u03c2 \u03c4\u03c9\u03bd \u03b3\u03bf\u03bd\u03b9\u03ce\u03bd \u03c4\u03bf\u03c5\u03c2 \u03ba\u03b1\u03b9 \u03c4\u03b7\u03c2 \u03b5\u03ba\u03ac\u03c3\u03c4\u03bf\u03c4\u03b5 \u03b5\u03c0\u03b9\u03c4\u03c1\u03bf\u03c0\u03ae\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03ce\u03bd (\u03bc\u03b5 \u03b9\u03c3\u03cc\u03b2\u03b9\u03b1 \u03bc\u03ad\u03bb\u03b7 \u03c4\u03bf\u03c5\u03c2 \u03ba.\u03ba.\u03a4\u03c5\u03c1\u03bb\u03ae \u03ba\u03b1\u03b9 \u039a\u03bf\u03bd\u03c4\u03bf\u03b3\u03b9\u03ac\u03bd\u03bd\u03b7 - \u0397 \u03b1\u03c0\u03cc\u03bb\u03c5\u03c4\u03b7 \u03b4\u03c5\u03b1\u03c1\u03c7\u03af\u03b1 \u03c4\u03b7\u03c2 \u0395\u039c\u0395 ). \r\n\u03a0\u03ac\u03bc\u03b5 \u03c4\u03ce\u03c1\u03b1 \u03c3\u03c4\u03bf\u03bd \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 . \r\n\u03a4\u03b1 \u03b8\u03ad\u03bc\u03b1\u03c4\u03b1 \u03c0\u03bf\u03c5 \u03ad\u03b2\u03b1\u03bb\u03b1\u03bd \u03c6\u03ad\u03c4\u03bf\u03c2 \u03ae\u03c4\u03b1\u03bd \u03b1\u03c0\u03b1\u03c1\u03ac\u03b4\u03b5\u03ba\u03c4\u03b1 . \u03a4\u03bf 2\u03bf \u03ba\u03b1\u03b9 \u03c4\u03bf 3\u03bf \u03ae\u03c4\u03b1\u03bd \u03ba\u03b1\u03bb\u03ac \u03bc\u03b5\u03bd \u03b3\u03bd\u03c9\u03c3\u03c4\u03ac \u03b4\u03b5 (\u03ad\u03b3\u03b9\u03bd\u03b5 \u03c4\u03cc\u03c4\u03b5 \u03bc\u03b5\u03b3\u03ac\u03bb\u03b7 \u03c3\u03c5\u03b6\u03ae\u03c4\u03b7\u03c3\u03b7) .\r\n\u03a4\u03cc\u03c4\u03b5 \u03cc\u03bc\u03c9\u03c2 \u03b4\u03b5\u03bd \u03b5\u03b9\u03c0\u03ce\u03b8\u03b7\u03ba\u03b5 \u03cc\u03c4\u03b9 \u03ba\u03b1\u03b9 \u03c4\u03bf 1\u03bf \u03b8\u03ad\u03bc\u03b1 \u03ae\u03c4\u03b1\u03bd \u03b3\u03bd\u03c9\u03c3\u03c4\u03cc\u03c4\u03b1\u03c4\u03bf !\r\n\u038c\u03c0\u03c9\u03c2 \u03bc\u03b5 \u03c0\u03bb\u03b7\u03c1\u03bf\u03c6\u03cc\u03c1\u03b7\u03c3\u03b5 \u03c6\u03bf\u03b9\u03c4\u03b7\u03c4\u03ae\u03c2 \u03c4\u03bf\u03c5 \u039c\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03bf\u03cd \u0399\u03c9\u03b1\u03bd\u03bd\u03af\u03bd\u03c9\u03bd , \u03c4\u03bf \u03b5\u03af\u03c7\u03b5 \u03b2\u03ac\u03bb\u03b5\u03b9 \u03bf \u03ba. \u039c\u03c0\u03cc\u03bb\u03b7\u03c2 \u03c3\u03c4\u03b7\u03bd \u03b5\u03be\u03b5\u03c4\u03b1\u03c3\u03c4\u03b9\u03ba\u03ae \u03c4\u03bf\u03c5 \u0399\u03bf\u03cd\u03bd\u03b9\u03bf\u03c5 \u03c4\u03bf\u03c5 2005 \u03ba\u03b1\u03b9 \u03c4\u03bf\u03c5 \u03a3\u03b5\u03c0\u03c4\u03b5\u03bc\u03b2\u03c1\u03af\u03bf\u03c5 \u03c4\u03bf\u03c5 2005 !!\r\n\u0393\u03b9\u03b1 \u03c4\u03bf 4\u03bf \u03b4\u03b5\u03bd \u03c4\u03bf \u03c3\u03c5\u03b6\u03b7\u03c4\u03ac\u03bc\u03b5 , \u03ae\u03c4\u03b1\u03bd \u03b1\u03c0\u03b1\u03c1\u03ac\u03b4\u03b5\u03ba\u03c4\u03bf \u03ba\u03b1\u03b9 \u03b1\u03c3\u03b1\u03c6\u03ad\u03c2 .\r\n\u03a3\u03c4\u03b7 \u03c3\u03c5\u03bd\u03ad\u03c7\u03b5\u03b9\u03b1 \u03b5\u03af\u03c7\u03b1\u03bc\u03b5 \u03c4\u03b7\u03bd \u03c0\u03b1\u03c1\u03b1\u03af\u03c4\u03b7\u03c3\u03b7 \u03c4\u03bf\u03c5 \u03b1\u03b3\u03b1\u03c0\u03b7\u03c4\u03bf\u03cd \u03ba. \u039c\u03c0\u03cc\u03bb\u03b7 \u03b1\u03c0\u03cc \u03c0\u03c1\u03cc\u03b5\u03b4\u03c1\u03bf\u03c2 \u03c4\u03b7\u03c2 \u03b5\u03c0\u03b9\u03c4\u03c1\u03bf\u03c0\u03ae\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03ce\u03bd , \u03c4\u03b7\u03bd \u03bf\u03c0\u03bf\u03af\u03b1 \u03b7 \u03b5\u03c0\u03af\u03c3\u03b7\u03bc\u03b7 \u0395\u039c\u0395 \u03c6\u03c1\u03cc\u03bd\u03c4\u03b9\u03b6\u03b1\u03bd \u03bd\u03b1 \u03bc\u03b7 \u03b3\u03af\u03bd\u03b5\u03b9 \u03b3\u03bd\u03c9\u03c3\u03c4\u03ae . \r\n\u0393\u03b9\u03b1\u03c4\u03af \u03ac\u03c1\u03b1\u03b3\u03b5 ??\r\n\u039f \u03ba. \u039c\u03c0\u03cc\u03bb\u03b7\u03c2 \u03ba\u03b1\u03bb\u03ac \u03b8\u03b1 \u03ae\u03c4\u03b1\u03bd \u03bd\u03b1 \u03bc\u03b1\u03c2 \u03ad\u03bb\u03b5\u03b3\u03b5 \u03c4\u03bf\u03c5\u03c2 \u03bb\u03cc\u03b3\u03bf\u03c5\u03c2 .\u03a0\u03b9\u03c3\u03c4\u03b5\u03cd\u03c9 \u03cc\u03c4\u03b9 \u03ae\u03c4\u03b1\u03bd \u03c0\u03bf\u03bb\u03cd \u03c3\u03bf\u03b2\u03b1\u03c1\u03bf\u03af !!!!\r\n\u03a3\u03c4\u03bf\u03bd \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 \u03b5\u03c0\u03af\u03c3\u03b7\u03c2 \u03c5\u03c0\u03ac\u03c1\u03c7\u03b5\u03b9 \u03ad\u03bd\u03b1 \u03ba\u03ce\u03bb\u03c5\u03bc\u03b1 \u03c4\u03b7\u03c2 \u0395\u039c\u0395 \u03bc\u03b5 \u03c4\u03b7 \u03b4\u03b9\u03ac\u03c1\u03ba\u03b5\u03b9\u03b1 \u03c4\u03bf\u03c5. \u03a0\u03bf\u03bb\u03bb\u03bf\u03af \u03bc\u03b1\u03b8\u03b7\u03c4\u03ad\u03c2 \u03ba\u03b1\u03b9 \u03b3\u03bf\u03bd\u03b5\u03af\u03c2 \u03b6\u03b7\u03c4\u03bf\u03cd\u03bd \u03bf \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03cc\u03c2 \u03bd\u03b1 \u03ad\u03c7\u03b5\u03b9 4 \u2013 4.30 \u03ce\u03c1\u03b5\u03c2 \u03b4\u03b9\u03ac\u03c1\u03ba\u03b5\u03b9\u03b1 .\u0386\u03b3\u03bd\u03c9\u03c3\u03c4\u03bf \u03b3\u03b9\u03b1\u03c4\u03af \u03bf\u03b9 \u03ba.\u03ba \u03c4\u03b7\u03c2 \u0395\u039c\u0395 \u03b4\u03b5\u03bd \u03c4\u03bf \u03ba\u03ac\u03bd\u03bf\u03c5\u03bd \u03b4\u03b5\u03ba\u03c4\u03cc .\u0395\u03c0\u03af\u03c3\u03b7\u03c2 \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd \u03ba\u03b1\u03c4\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2 \u03b1\u03c0\u03cc \u03b3\u03bf\u03bd\u03b5\u03af\u03c2 \u03cc\u03c4\u03b9 \u03b7 \u03b5\u03c0\u03b9\u03c4\u03ae\u03c1\u03b7\u03c3\u03b7 \u03c3\u03c4\u03bf\u03bd \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03b1\u03bb\u03ae . \u039f\u03b9 \u03bc\u03b1\u03b8\u03b7\u03c4\u03ad\u03c2 \u03ac\u03bd\u03b5\u03c4\u03b1 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03c3\u03c5\u03bd\u03b5\u03bd\u03bd\u03bf\u03b7\u03b8\u03bf\u03cd\u03bd \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03bc\u03b9\u03bb\u03ae\u03c3\u03bf\u03c5\u03bd .\u0398\u03b1 \u03c0\u03b5\u03b9 \u03ba\u03b1\u03bd\u03b5\u03af\u03c2 \u03cc\u03c4\u03b9 \u03b1\u03c5\u03c4\u03cc \u03b4\u03b5\u03bd \u03c4\u03bf\u03bd \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03b5\u03b9 .\u03a3\u03c9\u03c3\u03c4\u03cc .\r\n\u0391\u03bb\u03bb\u03ac \u03b1\u03bd \u03c4\u03bf\u03c5\u03c2 \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03b5\u03b9 ? \u0393\u03b9\u03b1 \u03c3\u03ba\u03b5\u03c6\u03c4\u03b5\u03af\u03c4\u03b5 \u03c4\u03bf . \u039a\u03ac\u03c0\u03bf\u03b9\u03bf\u03b9 \u03b3\u03bf\u03bd\u03b5\u03af\u03c2 \u03b5\u03af\u03c7\u03b1\u03bd \u03ba\u03ac\u03bd\u03b5\u03b9 \u03b5\u03c0\u03ce\u03bd\u03c5\u03bc\u03b5\u03c2 \u03ba\u03b1\u03c4\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2 \u03c3\u03c4\u03bf \u03c0\u03b1\u03c1\u03b5\u03bb\u03b8\u03cc\u03bd . \r\n\u03a3\u03c4\u03b7\u03bd \u039c\u03b5\u03c3\u03bf\u03b3\u03b5\u03b9\u03ac\u03b4\u03b1 \u03c4\u03bf 3\u03b2 \u03b5\u03c1\u03ce\u03c4\u03b7\u03bc\u03b1 \u03ae\u03c4\u03b1\u03bd \u03bb\u03ac\u03b8\u03bf\u03c2 !!! \u039a\u03b1\u03bd\u03b5\u03af\u03c2 \u03b4\u03b5\u03bd \u03b1\u03bd\u03ad\u03bb\u03b1\u03b2\u03b5 \u03c4\u03b7\u03bd \u03b5\u03c5\u03b8\u03cd\u03bd\u03b7 .\u03a4\u03b9 \u03ad\u03b3\u03b9\u03bd\u03b5 , \u03bb\u03ac\u03b8\u03bf\u03c2 \u03c3\u03c4\u03b7 \u03bc\u03b5\u03c4\u03ac\u03c6\u03c1\u03b1\u03c3\u03b7 \u03ae \u03ae\u03c4\u03b1\u03bd \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b1\u03c1\u03c7\u03ae \u03bb\u03ac\u03b8\u03bf\u03c2 .\u039a\u03b1\u03bb\u03ac \u03b4\u03b5\u03bd \u03c4\u03b1 \u03bb\u03cd\u03bd\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b8\u03ad\u03bc\u03b1\u03c4\u03b1 \u03c0\u03c1\u03bf\u03c4\u03bf\u03cd \u03c4\u03b1 \u03b2\u03ac\u03bb\u03bf\u03c5\u03bd ? \r\n\u0395\u03c0\u03af\u03c3\u03b7\u03c2 \u03b8\u03ad\u03c4\u03c9 \u03c5\u03c0\u03cc\u03c8\u03b9\u03bd \u03c3\u03b1\u03c2 \u03b3\u03b9\u03b1 \u03c0\u03c1\u03bf\u03b2\u03bb\u03b7\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03cc \u03ba\u03b1\u03b9 \u03c4\u03bf \u03c0\u03b1\u03c1\u03b1\u03ba\u03ac\u03c4\u03c9 post \u03c4\u03bf\u03c5 \u03ba. Dulac (\u03bf \u03bf\u03c0\u03bf\u03af\u03bf\u03c2 \u03c6\u03b1\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03cc\u03c4\u03b9 \u03be\u03ad\u03c1\u03b5\u03b9 \u03c0\u03bf\u03bb\u03bb\u03ac) . \r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?p=503458&search_author=DuLac&show_results=posts&sort_dir=desc&return_chars=200#503458\r\n \r\n\u0391\u03c5\u03c4\u03ac \u03ba\u03b1\u03c4\u03b1\u03b8\u03ad\u03c4\u03c9 \u03b3\u03b9\u03b1 \u03b1\u03c1\u03c7\u03ae \u03ba\u03b1\u03b9 \u03c3\u03c4\u03b7 \u03c3\u03c5\u03bd\u03ad\u03c7\u03b5\u03b9\u03b1 \u03c4\u03bf\u03c5 \u03b4\u03b9\u03b1\u03bb\u03cc\u03b3\u03bf\u03c5 (\u03ae\u03c1\u03b5\u03bc\u03bf\u03c5 \u03ba\u03b1\u03b9 \u03b4\u03b7\u03bc\u03b9\u03bf\u03c5\u03c1\u03b3\u03b9\u03ba\u03bf\u03cd) \u03c0\u03bf\u03c5 \u03b8\u03b1 \u03c0\u03c1\u03bf\u03ba\u03cd\u03c8\u03b5\u03b9 \u03b8\u03b1 \u03c0\u03c1\u03bf\u03c3\u03b8\u03ad\u03c3\u03c9 \u03ba\u03b1\u03b9 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b5\u03c2 \u03ac\u03bb\u03bb\u03b5\u03c2 \u03c0\u03b1\u03c1\u03b1\u03c4\u03b7\u03c1\u03ae\u03c3\u03b5\u03b9\u03c2 \u03bc\u03bf\u03c5 .\r\n\r\n\u03a4\u03b7\u03bd \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03c4\u03b7\u03bd \u03b5\u03c0\u03b9\u03b6\u03b7\u03c4\u03bf\u03cd\u03bc\u03b5 . \u039c\u03b1\u03c2 \u03b2\u03bf\u03b7\u03b8\u03ac\u03b5\u03b9 \u03bd\u03b1 \u03b3\u03af\u03bd\u03bf\u03c5\u03bc\u03b5 \u03ba\u03b1\u03bb\u03cd\u03c4\u03b5\u03c1\u03bf\u03b9 .",
"Solution_1": "\u03a6\u03af\u03bb\u03b5 \u03a3\u03c4\u03ac\u03c1\u03c4 \u03b1\u03c6\u03bf\u03cd \u03b8\u03ad\u03bb\u03b5\u03b9\u03c2 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03b8\u03b1 \u03ad\u03c7\u03b5\u03b9\u03c2 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03b7 \u03bf\u03c0\u03bf\u03af\u03b1 \u03b8\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03af\u03c3\u03c9\u03c2 \u03ba\u03b9 \u03c0\u03bf\u03bb\u03cd \u03c3\u03ba\u03bb\u03b7\u03c1\u03ae \u03c3\u03b5 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b1 \u03c3\u03b7\u03bc\u03b5\u03af\u03b1 \u03b1\u03bd\u03b1\u03c4\u03c1\u03b5\u03c0\u03c4\u03b9\u03ba\u03ae \u03ba\u03b9 \u03c7\u03b9\u03bf\u03c5\u03bc\u03bf\u03c1\u03b9\u03c3\u03c4\u03b9\u03ba\u03ae \u03c3\u03b5 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b1 \u03ac\u03bb\u03bb\u03b1 \u03b1\u03bb\u03bb\u03ac \u03c0\u03b1\u03bd\u03c4\u03bf\u03cd \u03b5\u03c0\u03bf\u03b9\u03ba\u03bf\u03b4\u03bf\u03bc\u03b7\u03c4\u03b9\u03ba\u03ae \u0391\u03bb\u03bb\u03ac \u03b8\u03b1 \u03ae\u03b8\u03b5\u03bb\u03b1 \u03bd\u03b1 \u03c3\u03bf\u03c5 \u03b8\u03c5\u03bc\u03af\u03c3\u03c9 \u03c4\u03b9 \u03b5\u03af\u03c7\u03b5 \u03b3\u03af\u03bd\u03b5\u03b9 \u03ba\u03b1\u03c4\u03ac \u03c4\u03b7\u03bd \u03b2\u03b1\u03bb\u03ba\u03b1\u03bd\u03b9\u03ac\u03b4\u03b1 \u03cc\u03c0\u03bf\u03c5 \u03b5\u03c3\u03cd \u03b5\u03af\u03c7\u03b5\u03c2 \u03ba\u03ac\u03bd\u03b5\u03b9 \u03ba\u03b1\u03bb\u03ae \u03b1\u03bb\u03bb\u03ac \u03ba\u03b9 \u03c5\u03c0\u03b5\u03c1\u03b2\u03bf\u03bb\u03b9\u03ba\u03ac \u03b5\u03c0\u03b9\u03b8\u03b5\u03c4\u03b9\u03ba\u03ae \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03ba\u03b9 \u03ad\u03b4\u03c9\u03c3\u03b5\u03c2 \u03ad\u03b4\u03b1\u03c6\u03bf\u03c2 \u03c3\u03b5 \u03ad\u03bd\u03b1\u03bd \u03b1\u03bd\u03cd\u03c0\u03b1\u03c1\u03ba\u03c4\u03bf \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03b5\u03af\u03c7\u03b5 \u03c4\u03bf \u03b8\u03ac\u03c1\u03c1\u03bf\u03c2 \u03bd\u03b1 \u03b4\u03ce\u03c3\u03b5\u03b9 \u03c4\u03bf \u03cc\u03bd\u03bf\u03bc\u03ac \u03c4\u03bf\u03c5 \u03bd\u03b1 \u03ba\u03b1\u03c4\u03b7\u03b3\u03bf\u03c1\u03ae\u03c3\u03b5\u03b9 \u03c4\u03bf\u03c5\u03c2 \u03c0\u03ac\u03bd\u03c4\u03b5\u03c2 \u03b5\u03b3\u03ce \u03c4\u03bf\u03c5 \u03b1\u03bd\u03c4\u03b9\u03c3\u03c4\u03ac\u03b8\u03b7\u03ba\u03b1 \u03cc\u03c3\u03bf \u03b3\u03b9\u03bd\u03cc\u03c4\u03b1\u03bd \u03ba\u03b1\u03b8\u03cc\u03c4\u03b9 \u03bc\u03bf\u03c5 \u03b1\u03c0\u03b7\u03cd\u03b8\u03c5\u03bd\u03b5 \u03b2\u03b1\u03c1\u03b9\u03ad\u03c2 \u03b5\u03ba\u03c6\u03c1\u03ac\u03c3\u03b5\u03b9\u03c2 \u03b1\u03bb\u03bb\u03ac \u03c4\u03bf \u03c4\u03b5\u03bb\u03b9\u03ba\u03cc \u03b1\u03c0\u03bf\u03c4\u03ad\u03bb\u03b5\u03c3\u03bc\u03b1 \u03ae\u03c4\u03b1\u03bd \u03bd\u03b1 \u03b1\u03bd\u03b1\u03c3\u03c4\u03b1\u03c4\u03c9\u03b8\u03b5\u03af \u03bd\u03b1 \u03b1\u03c0\u03bf\u03b4\u03b9\u03bf\u03c1\u03b3\u03b1\u03bd\u03c9\u03b8\u03b5\u03af \u03ba\u03b9 \u03bd\u03b1 \u03b6\u03b7\u03bc\u03b9\u03c9\u03b8\u03b5\u03af \u03c4\u03bf \u03b5\u03bb\u03bb\u03b7\u03bd\u03b9\u03ba\u03cc \u03c6\u03cc\u03c1\u03bf\u03c5\u03bc \u03c0\u03bf\u03c5 \u03bf \u03bc\u03cc\u03bd\u03bf\u03c2 \u03c3\u03ba\u03bf\u03c0\u03cc\u03c2 \u03c4\u03bf\u03c5 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03c7\u03c4\u03af\u03b6\u03b5\u03b9 \u03ba\u03b9 \u03cc\u03c7\u03b9 \u03bd\u03b1 \u03ba\u03b1\u03c4\u03b1\u03c3\u03c4\u03c1\u03ad\u03c6\u03b5\u03b9 \u03b8\u03ad\u03bb\u03c9 \u03bd\u03b1 \u03c0\u03b9\u03c3\u03c4\u03b5\u03cd\u03c9 \u03cc\u03c4\u03b9 \u03c4\u03bf \u03ad\u03ba\u03b1\u03bd\u03b5\u03c2 \u03ba\u03b1\u03c4\u03ac \u03bb\u03ac\u03b8\u03bf\u03c2 \u03ba\u03b9 \u03b1\u03c0\u03cc \u03b1\u03c5\u03c4\u03ae\u03bd \u03c4\u03b7\u03bd \u03c0\u03af\u03c3\u03c4\u03b7 \u03bc\u03bf\u03c5 \u03c3\u03bf\u03c5 \u03b5\u03c6\u03b9\u03c3\u03c4\u03ce \u03c4\u03b7\u03bd \u03c0\u03c1\u03bf\u03c3\u03bf\u03c7\u03ae \u03ad\u03c4\u03c3\u03b9 \u03ce\u03c3\u03c4\u03b5 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03b5\u03c0\u03b1\u03bd\u03b1\u03bb\u03b7\u03c6\u03b8\u03b5\u03af \u03b4\u03b9\u03cc\u03c4\u03b9 \u03c3\u03b5 \u03b1\u03c5\u03c4\u03ae\u03bd \u03c4\u03b7\u03bd \u03c0\u03b5\u03c1\u03af\u03c0\u03c4\u03c9\u03c3\u03b7 \u03b8\u03b1 \u03c0\u03b9\u03c3\u03c4\u03ad\u03c8\u03c9 \u03cc\u03c4\u03b9 \u03c4\u03bf \u03ba\u03ac\u03bd\u03b5\u03b9\u03c2 \u03b5\u03c3\u03ba\u03b5\u03bc\u03bc\u03ad\u03bd\u03b1 \u03b4\u03b5\u03bd \u03b8\u03ad\u03bb\u03c9 \u03bd\u03b1 \u03bc\u03b5 \u03c0\u03b1\u03c1\u03b5\u03be\u03b7\u03b3\u03ae\u03c3\u03b5\u03b9\u03c2 \u03cc\u03bc\u03c9\u03c2 \u03b8\u03b1 \u03c0\u03b1\u03bb\u03ad\u03c8\u03c9 \u03bc\u03b5 \u03bd\u03cd\u03c7\u03b9\u03b1 \u03ba\u03b9 \u03bc\u03b5 \u03b4\u03cc\u03bd\u03c4\u03b9\u03b1 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03be\u03b1\u03bd\u03b1\u03c3\u03c5\u03bc\u03b2\u03b5\u03af \u03b1\u03c5\u03c4\u03cc \u03c4\u03bf \u03c3\u03c5\u03bc\u03b2\u03ac\u03bd \u03b3\u03b9\u03b1\u03c4\u03af \u03b5\u03b3\u03ce \u03b5\u03ba\u03c4\u03b9\u03bc\u03ce \u03c0\u03bf\u03bb\u03cd \u03b1\u03c5\u03c4\u03ac \u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03bf\u03c5 \u03b1\u03b3\u03b1\u03c0\u03bf\u03cd\u03bd \u03c4\u03bf \u03c6\u03cc\u03c1\u03bf\u03c5\u03bc \u03c4\u03bf \u03b1\u03b3\u03b1\u03c0\u03ce \u03ba\u03b9 \u03b5\u03b3\u03ce \u03ba\u03b9 \u03c3\u03b5 \u03bc\u03af\u03b1 \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03b1\u03b4\u03cd\u03bd\u03b1\u03bc\u03b7 \u03c3\u03ae\u03bc\u03b5\u03c1\u03b1-\u03bb\u03cc\u03b3\u03c9 \u03c0\u03bf\u03bb\u03b9\u03c4\u03b5\u03af\u03b1\u03c2- \u0395\u03bb\u03bb\u03ac\u03b4\u03b1 \u03c4\u03bf \u03bd\u03b1 \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03bf\u03c5 \u03b1\u03b3\u03b1\u03c0\u03bf\u03cd\u03bd \u03bc\u03b9\u03b1 \u03c4\u03ad\u03c4\u03bf\u03b9\u03b1 \u03b9\u03b4\u03ad\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03be\u03b9\u03bf\u03b8\u03b1\u03cd\u03bc\u03b1\u03c3\u03c4\u03b1 \u03ba\u03b9 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03c0\u03c1\u03bf\u03c3\u03c4\u03b1\u03c4\u03b5\u03cd\u03bf\u03bd\u03c4\u03b1\u03b9\r\n\u03a6\u03c5\u03c3\u03b9\u03ba\u03ac \u03ba\u03b9 \u03b5\u03b3\u03ce \u03b5\u03af\u03bc\u03b1\u03b9 \u03c3\u03c4\u03b7\u03bd \u03b1\u03c0\u03cc\u03bb\u03c5\u03c4\u03b7 \u03b4\u03b9\u03ac\u03b8\u03b5\u03c3\u03b7 \u03cc\u03bb\u03c9\u03bd \u03b3\u03b9\u03b1 \u03bf\u03c0\u03bf\u03b9\u03b1\u03b4\u03ae\u03c0\u03bf\u03c4\u03b5 \u03b5\u03af\u03b4\u03bf\u03c5\u03c2 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03cc\u03c3\u03bf\u03c2 \u03c3\u03ba\u03bb\u03b7\u03c1\u03ae \u03ba\u03b9 \u03b1\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03b4\u03b9\u03cc\u03c4\u03b9 \u03b5\u03b3\u03ce \u03ad\u03c7\u03c9 \u03c3\u03c5\u03bd\u03b7\u03b8\u03af\u03c3\u03b5\u03b9 \u03c3\u03c4\u03b7\u03bd \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03bc\u03ad\u03c7\u03c1\u03b9 \u03ba\u03b9 \u03c4\u03b1 \u03cc\u03c1\u03b9\u03b1 \u03c4\u03b7\u03c2 \u03c3\u03c5\u03ba\u03bf\u03c6\u03b1\u03bd\u03c4\u03af\u03b1\u03c2 \u03c3\u03b5 \u03b2\u03ac\u03c1\u03bf\u03c2 \u03bc\u03bf\u03c5 \u03b1\u03ba\u03cc\u03bc\u03b1 \u03ba\u03b9 \u03b1\u03c0\u03cc \u03ba\u03b1\u03b8\u03b7\u03b3\u03b7\u03c4\u03ae \u039c\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ce\u03bd \u03c4\u03bf\u03c5 \u03c3\u03c7\u03bf\u03bb\u03b5\u03af\u03bf\u03c5 \u03bc\u03bf\u03c5 \u03b7 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03c3\u03b1\u03c2 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b3\u03af\u03bd\u03b5\u03b9 \u03bc\u03ad\u03c3\u03c9 \u03cc\u03c0\u03bf\u03b9\u03bf\u03c5 \u03bc\u03ad\u03c3\u03bf\u03c5 \u03b5\u03c0\u03b9\u03b8\u03c5\u03bc\u03b5\u03af\u03c4\u03b5 \u03b1\u03bb\u03bb\u03ac \u03ba\u03b9 \u03bc\u03ad\u03c3\u03c9 \u03c0\u03c1\u03bf\u03c3\u03c9\u03c0\u03b9\u03ba\u03bf\u03cd \u03bc\u03c5\u03bd\u03ae\u03bc\u03b1\u03c4\u03bf\u03c2 \u03b8\u03b1 \u03ae\u03b8\u03b5\u03bb\u03b1 \u03cc\u03bc\u03c9\u03c2 \u03c4\u03b7\u03bd \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03b5\u03bd\u03b1\u03bd\u03c4\u03af\u03bf\u03bd \u03bc\u03bf\u03c5 \u03bd\u03b1 \u03c4\u03b7\u03bd \u03b4\u03bf\u03cd\u03bd\u03b5 \u03cc\u03bc\u03c9\u03c2 \u03ba\u03b9 \u03cc\u03bb\u03bf\u03b9 \u03b3\u03b9\u03b1 \u03b1\u03c5\u03c4\u03cc \u03b5\u03b3\u03ce \u03b8\u03b1 \u03c4\u03b7\u03bd \u03b4\u03b7\u03bc\u03bf\u03c3\u03b9\u03b5\u03cd\u03c3\u03c9 \u03b1\u03bd \u03bc\u03bf\u03c5 \u03b4\u03bf\u03b8\u03b5\u03af \u03c3\u03b5 \u03c0\u03c1\u03bf\u03c3\u03c9\u03c0\u03b9\u03ba\u03cc \u03bc\u03ae\u03bd\u03c5\u03bc\u03b1 \u03ba\u03b1\u03b8\u03b1\u03c1\u03cc\u03c2 \u03bf\u03c5\u03c1\u03b1\u03bd\u03cc\u03c2 \u03b1\u03c3\u03c4\u03c1\u03b1\u03c0\u03ad\u03c2 \u03b4\u03b5\u03bd \u03c6\u03bf\u03b2\u03ac\u03c4\u03b1\u03b9 \u03a0\u039f\u03a4\u0395 \r\n\u03a3\u03c4\u03bf \u03b4\u03b9\u03b1\u03c4\u03ac\u03c5\u03c4\u03b1 \u03cc\u03bc\u03c9\u03c2 \u03b1\u03c6\u03bf\u03cd \u03b8\u03b5\u03c2 \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u03bf\u03c1\u03af\u03c3\u03c4\u03b5 \r\n1.\t\u03b5\u03c1\u03b1\u03c3\u03b9\u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac \u03b4\u03b5\u03bd \u03b8\u03b1 \u03b4\u03b9\u03b1\u03c6\u03c9\u03bd\u03ae\u03c3\u03c9 \u03b5\u03c0\u03b1\u03c5\u03be\u03ac\u03bd\u03c9 \u03c3\u03b5 \u03b1\u03c5\u03c4\u03cc \u03cc\u03bc\u03c9\u03c2 \u03c4\u03bf \u03ba\u03bf\u03c1\u03bf\u03ca\u03b4\u03b5\u03c5\u03c4\u03b9\u03ba\u03ac \u03b5\u03af\u03bd\u03b1\u03b9 \u03bb\u03af\u03b3\u03bf \u03c5\u03c0\u03b5\u03c1\u03b2\u03bf\u03bb\u03b9\u03ba\u03cc \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03bf\u03c1\u03bf\u03ca\u03b4\u03b5\u03cd\u03bf\u03c5\u03bd \u03ba\u03b1\u03bd\u03ad\u03bd\u03b1 \u03b1\u03c0\u03bb\u03ce\u03c2 \u03b5\u03c0\u03b9\u03b4\u03b5\u03b9\u03ba\u03c4\u03b9\u03ba\u03ac \u03b1\u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03bf\u03cd\u03bd \u03c0\u03bf\u03bb\u03b9\u03c4\u03b5\u03af\u03b1 \u03ba\u03c1\u03ac\u03c4\u03bf\u03c2 \u03ba\u03b9 \u03ac\u03bb\u03bb\u03bf\u03b9 \u03c6\u03bf\u03c1\u03b5\u03af\u03c2 \u03bc\u03b5\u03c4\u03b1\u03be\u03cd \u03b1\u03c5\u03c4\u03ce\u03bd \u03ba\u03b9 \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03b9 \u03b1\u03bb\u03bb\u03ac \u03cc\u03c7\u03b9 \u03cc\u03bb\u03bf\u03b9 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u0395\u039c\u0395 \u03b1\u03bd \u03b8\u03b5\u03c2 \u03bf\u03bd\u03cc\u03bc\u03b1\u03c4\u03b1 \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9\u03c2 \u03c0\u03b1\u03c1\u03ac \u03bd\u03b1 \u03bc\u03b5 \u03c0\u03c1\u03bf\u03ba\u03b1\u03bb\u03ad\u03c3\u03b5\u03b9\u03c2(\u03bc\u03b5 \u03c4\u03b7\u03bd \u03ba\u03b1\u03bb\u03ae \u03c4\u03b7\u03bd \u03ad\u03bd\u03bd\u03bf\u03b9\u03b1 \u03c0\u03ac\u03bd\u03c4\u03b1)\u03ba\u03b9 \u03b5\u03b3\u03ce \u03b8\u03b1 \u03c3\u03bf\u03c5 \u03b1\u03c1\u03b1\u03b4\u03b9\u03ac\u03c3\u03c9 \u03bf\u03bd\u03cc\u03bc\u03b1\u03c4\u03b1 \u03bc\u03b5 \u03b5\u03c0\u03b5\u03be\u03b7\u03b3\u03ae\u03c3\u03b5\u03b9\u03c2 \u03ac\u03c1\u03b1 \u03c3\u03c4\u03bf \u03ba\u03bf\u03c1\u03bf\u03ca\u03b4\u03b5\u03c5\u03c4\u03b9\u03ba\u03ac \u03b5\u03b3\u03ce \u03b8\u03b1 \u03ad\u03b2\u03b1\u03b6\u03b1 {\u03bc\u03b5 \u03c3\u03c7\u03b5\u03c4\u03b9\u03ba\u03ae \u03b1\u03b4\u03b9\u03b1\u03c6\u03bf\u03c1\u03af\u03b1} \r\n2.\t\u03a3\u03c4\u03b9\u03c2 \u03b5\u03c1\u03c9\u03c4\u03ae\u03c3\u03b5\u03b9\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u03b2\u03b9\u03b2\u03bb\u03b9\u03bf\u03b3\u03c1\u03b1\u03c6\u03af\u03b1\r\n3.\t\u03a0\u03c1\u03bf\u03c6\u03b1\u03bd\u03ce\u03c2 \u03ba\u03b9 \u03be\u03ad\u03c1\u03bf\u03c5\u03bd \u03b1\u03bb\u03bb\u03ac \u03b8\u03b5\u03c9\u03c1\u03bf\u03cd\u03bd \u03b4\u03b5\u03b4\u03bf\u03bc\u03ad\u03bd\u03bf \u03cc\u03c4\u03b9 \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03bf\u03cd\u03bc\u03b5 \u03c4\u03b7\u03bd \u03b2\u03b9\u03b2\u03bb\u03b9\u03bf\u03b8\u03ae\u03ba\u03b7 \u03c4\u03b7\u03c2 \u0395\u039c\u0395 (\u03b5\u03b3\u03ce 4 \u03c7\u03c1\u03cc\u03bd\u03b9\u03b1 \u03c0\u03bf\u03c5 \u03c0\u03b7\u03b3\u03b1\u03af\u03bd\u03c9 \u03c3\u03c5\u03c7\u03bd\u03ac \u03bf\u03cd\u03c4\u03b5 \u03c4\u03b7\u03bd \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03af\u03b7\u03c3\u03b1 \u03bf\u03cd\u03c4\u03b5 \u03b5\u03af\u03b4\u03b1 \u03c0\u03bf\u03c4\u03ad \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03bd \u03bd\u03b1 \u03c4\u03b7\u03bd \u03c7\u03c1\u03b7\u03c3\u03b9\u03bc\u03bf\u03c0\u03bf\u03b9\u03b5\u03af)\r\n4.\t\u0394\u03b5\u03bd \u03c4\u03bf\u03c5\u03c2 \u03b1\u03c6\u03bf\u03c1\u03ac \u03b8\u03b5\u03c9\u03c1\u03bf\u03cd\u03bd \u03c5\u03c0\u03bf\u03c7\u03c1\u03ad\u03c9\u03c3\u03b7 \u03c3\u03bf\u03c5 \u03cc\u03bb\u03b1 \u03bd\u03b1 \u03c4\u03b1 \u03be\u03ad\u03c1\u03b5\u03b9\u03c2 \u03cc\u03c3\u03bf\u03bd \u03b1\u03c6\u03bf\u03c1\u03ac \u03c4\u03b7\u03bd \u03b2\u03b9\u03b2\u03bb\u03b9\u03bf\u03b3\u03c1\u03b1\u03c6\u03af\u03b1 \u03ba\u03b9 \u03cc\u03c3\u03bf\u03bd \u03b1\u03c6\u03bf\u03c1\u03ac \u03b1\u03c5\u03c4\u03ae\u03bd \u03ba\u03ac\u03b8\u03b5 \u03b1\u03c5\u03c4\u03ae\u03bd \u03c4\u03b7\u03bd \u03b3\u03bd\u03ce\u03c3\u03b7 \u03b1\u03c5\u03c4\u03cc \u03b5\u03be\u03b7\u03b3\u03b5\u03af\u03c4\u03b1\u03b9 \u03b1\u03c0\u03cc 2 \u03c0\u03b1\u03c1\u03ac\u03b3\u03bf\u03bd\u03c4\u03b5\u03c2 \u03bf \u03c0\u03c1\u03ce\u03c4\u03bf\u03c2 \u03b1\u03c6\u03bf\u03c1\u03ac \u03c4\u03b7\u03bd \u03b5\u03bb\u03bb\u03b7\u03bd\u03b9\u03ba\u03cc \u03b3\u03ad\u03bd\u03bf\u03c2 \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03bd\u03b1 \u03b2\u03bf\u03b7\u03b8\u03ae\u03c3\u03b5\u03b9 \u03c4\u03bf\u03bd \u03c3\u03c5\u03bc\u03c0\u03bf\u03bb\u03af\u03c4\u03b7 \u03ae \u03c4\u03bf\u03bd \u03c3\u03c5\u03bd\u03ac\u03bd\u03b8\u03c1\u03c9\u03c0\u03bf \u03bd\u03b1 \u03b1\u03bd\u03ad\u03b2\u03b5\u03b9 \u03b1\u03bb\u03bb\u03ac \u03c4\u03bf\u03bd \u03c0\u03c1\u03bf\u03c4\u03b9\u03bc\u03ac \u03c7\u03b1\u03bc\u03b7\u03bb\u03ac \u03b3\u03b9\u03b1 \u03bd\u03b1 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03b4\u03b9\u03b1\u03ba\u03c1\u03b9\u03b8\u03b5\u03af \u03bf \u03af\u03b4\u03b9\u03bf\u03c2 \u03ba\u03b9 \u03b4\u03b5\u03cd\u03c4\u03b5\u03c1\u03bf\u03bd \u03bb\u03cc\u03b3\u03c9 \u03c4\u03bf\u03c5 \u00ab\u03b5\u03c1\u03b1\u03c3\u03b9\u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac\u00bb \u03b1\u03c6\u03bf\u03cd \u03b4\u03b5\u03bd \u03c0\u03bb\u03b7\u03c1\u03ce\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c9\u03c2 \u03b3\u03bd\u03ae\u03c3\u03b9\u03bf\u03b9 \u0388\u03bb\u03bb\u03b7\u03bd\u03b5\u03c2 \u03c0\u03ac\u03bd\u03c4\u03b1 \u03b4\u03b5\u03bd \u03c4\u03bf\u03c5\u03c2 \u03b5\u03bd\u03b4\u03b9\u03b1\u03c6\u03ad\u03c1\u03b5\u03b9 \u03bd\u03b1 \u03b2\u03bf\u03b7\u03b8\u03ae\u03c3\u03bf\u03c5\u03bd \u03c3\u03ba\u03ad\u03c6\u03c4\u03bf\u03bd\u03c4\u03b1\u03b9 \u03b1\u03bd \u03b8\u03b5\u03c2 \u03bd\u03b1 \u03b4\u03b9\u03b1\u03ba\u03c1\u03b9\u03b8\u03b5\u03af\u03c2 \u03b2\u03b3\u03ac\u03bb\u03c4\u03b1 \u03c0\u03ad\u03c1\u03b1 \u03bc\u03cc\u03bd\u03bf\u03c2 \u03c3\u03bf\u03c5 \u03c3\u03b5 \u03ad\u03bd\u03b1 13-17 \u03c7\u03c1\u03bf\u03bd\u03ce\u03bd \u03c0\u03b1\u03b9\u03b4\u03af \u03b1\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03c0\u03bf\u03c4\u03ad \u03b4\u03c5\u03bd\u03b1\u03c4\u03cc\u03bd\r\n5.\t\u03a0\u03b9\u03b8\u03b1\u03bd\u03ce\u03c2 \u03bd\u03b1\u03b9 \u03c3\u03c5\u03bc\u03b2\u03b1\u03af\u03bd\u03b5\u03b9 \u03c3\u03b5 \u03b1\u03c1\u03ba\u03b5\u03c4\u03ad\u03c2 \u03c0\u03b5\u03c1\u03b9\u03c0\u03c4\u03ce\u03c3\u03b5\u03b9\u03c2 \u03b1\u03bd \u03ad\u03c7\u03b5\u03b9 \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03c2 \u03c6\u03af\u03bb\u03bf \u03ba\u03b1\u03b8\u03b7\u03b3\u03b7\u03c4\u03ae \u03c3\u03c4\u03b7\u03bd \u0395\u03bb\u03bb\u03ac\u03b4\u03b1 \u03c6\u03af\u03bb\u03bf\u03b9 \u03bc\u03bf\u03c5 \u03b9\u03c3\u03c7\u03cd\u03b5\u03b9 \u03c4\u03bf \u03ba\u03b1\u03b8\u03b5\u03c3\u03c4\u03ce\u03c2 \u03c4\u03bf\u03c5 \u00ab\u03bc\u03ad\u03c3\u03bf\u03c5\u00bb \u03b1\u03bd \u03ad\u03c7\u03b5\u03b9 \u03c6\u03af\u03bb\u03bf \u03b3\u03bd\u03c9\u03c3\u03c4\u03cc \u03c3\u03c5\u03b3\u03b3\u03b5\u03bd\u03ae \u03ba\u03ac\u03c0\u03bf\u03c5 \u03bf\u03c0\u03bf\u03c5\u03b4\u03ae\u03c0\u03bf\u03c4\u03b5 \u03b8\u03b1 \u03b5\u03be\u03c5\u03c0\u03b7\u03c1\u03b5\u03c4\u03b7\u03b8\u03b5\u03af\u03c2 \u03b5\u03b9\u03b4\u03b5\u03bc\u03ae \u03c4\u03c1\u03ad\u03c7\u03b1 \u03b3\u03cd\u03c1\u03b5\u03c5\u03b5 \u03b1\u03c5\u03c4\u03cc \u03af\u03c3\u03c9\u03c2 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf \u03bc\u03bf\u03bd\u03b1\u03b4\u03b9\u03ba\u03cc \u03c0\u03c1\u03ac\u03b3\u03bc\u03b1 \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03c0\u03c1\u03cc\u03ba\u03b5\u03b9\u03c4\u03b1\u03b9 \u03bd\u03b1 \u03b1\u03bb\u03bb\u03ac\u03be\u03b5\u03b9 \u03c0\u03bf\u03c4\u03ad \u03b1\u03c6\u03bf\u03cd \u03bf \u0388\u03bb\u03bb\u03b7\u03bd\u03b1\u03c2 \u03b4\u03cd\u03c3\u03ba\u03bf\u03bb\u03b1 \u03b1\u03bb\u03bb\u03ac\u03b6\u03b5\u03b9 \u03c4\u03b7\u03bd \u03b9\u03b4\u03b9\u03bf\u03c3\u03c5\u03b3\u03ba\u03c1\u03b1\u03c3\u03af\u03b1 \u03c4\u03bf\u03c5 \u03bc\u03b1\u03ba\u03ac\u03c1\u03b9 \u03b7 \u03b4\u03b9\u03ba\u03ae \u03bc\u03b1\u03c2 \u03ba\u03b9 \u03bf\u03b9 \u03b5\u03c0\u03cc\u03bc\u03b5\u03bd\u03b5\u03c2 \u03b3\u03b5\u03bd\u03b9\u03ad\u03c2 \u03bd\u03b1 \u03c4\u03bf \u03b1\u03bb\u03bb\u03ac\u03be\u03bf\u03c5\u03bd \u03ac\u03c1\u03b4\u03b7\u03bd \u03b1\u03c5\u03c4\u03cc\r\n6.\t\u0393\u03b9\u03b1 \u03c4\u03b1 \u03bc\u03b1\u03b8\u03ae\u03bc\u03b1\u03c4\u03b1 \r\n7.\t\u0388\u03c7\u03b5\u03b9\u03c2 \u03b4\u03af\u03ba\u03b9\u03bf \u03cc\u03c4\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bf\u03c5\u03c3\u03b9\u03b1\u03c3\u03c4\u03b9\u03ba\u03ac \u03b1\u03bd\u03cd\u03c0\u03b1\u03c1\u03ba\u03c4\u03b7 \u03b7 \u03b4\u03b9\u03b4\u03b1\u03c3\u03ba\u03b1\u03bb\u03af\u03b1 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b5\u03c4\u03b1\u03b9\u03c1\u03b5\u03af\u03b1 \u03b1\u03bb\u03bb\u03ac \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03c3\u03c4\u03b9\u03b3\u03bc\u03ae \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03c5\u03c0\u03ac\u03c1\u03c7\u03b5\u03b9 \u03b5\u03c0\u03b1\u03b3\u03b3\u03b5\u03bb\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03cc\u03c2 \u03ba\u03b9 \u03cc\u03bb\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03c1\u03b1\u03c3\u03b9\u03c4\u03b5\u03c7\u03bd\u03b9\u03ba\u03ac \u03bd\u03b1 \u03b5\u03af\u03bc\u03b1\u03c3\u03c4\u03b5 \u03b5\u03c5\u03c7\u03b1\u03c1\u03b9\u03c3\u03c4\u03b7\u03bc\u03ad\u03bd\u03bf\u03b9 \u03c0\u03bf\u03c5 \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd \u03ad\u03c3\u03c4\u03c9 \u03ba\u03b9 \u03b1\u03c5\u03c4\u03ac \u03c4\u03b1 \u03bc\u03b1\u03b8\u03ae\u03bc\u03b1\u03c4\u03b1 \u03cc\u03c4\u03b9 \u03bf \u03c0\u03c1\u03cc\u03b5\u03b4\u03c1\u03bf\u03c2 \u03c4\u03b7\u03c2 \u0395\u039c\u0395 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c3\u03b1\u03bd \u03bd\u03b1 \u03bc\u03b7\u03bd \u03c5\u03c0\u03ac\u03c1\u03c7\u03b5\u03b9 \u03c4\u03bf \u03be\u03ad\u03c1\u03c9 \u03ba\u03b9 \u03b5\u03b3\u03ce \u03b4\u03b9\u03cc\u03c4\u03b9 \u03c4\u03cc\u03c3\u03b5\u03c2 \u03c6\u03bf\u03c1\u03ad\u03c2 \u03c0\u03bf\u03c5 \u03c0\u03ae\u03b3\u03b1 \u03b4\u03b5\u03bd \u03c4\u03bf\u03bd \u03b5\u03af\u03b4\u03b1 \u03c0\u03bf\u03c4\u03ad \u03c3\u03b5 \u03b1\u03bd\u03c4\u03af\u03b8\u03b5\u03c3\u03b7 \u03bc\u03b5 \u03c4\u03bf\u03bd \u039a\u03a5\u03a1\u0399\u039f \u03bc\u03b5 \u039a \u03ba\u03b5\u03c6\u03b1\u03bb\u03b1\u03af\u03bf \u0391\u03bb\u03b5\u03be\u03b1\u03bd\u03b4\u03c1\u03ae \u03bf \u03bf\u03c0\u03bf\u03af\u03bf\u03c2 \u03c4\u03b1\u03ba\u03c4\u03b9\u03ba\u03ac \u03b5\u03c0\u03b9\u03c3\u03ba\u03b5\u03c0\u03c4\u03cc\u03c4\u03b1\u03bd \u03c4\u03b1 \u03b3\u03c1\u03b1\u03c6\u03b5\u03af\u03b1 \u03cc\u03c3\u03bf\u03bd \u03b1\u03c6\u03bf\u03c1\u03ac \u03c4\u03bf \u03b8\u03ad\u03bc\u03b1 \u03c4\u03bf\u03c5 \u03c6\u03af\u03bb\u03c4\u03b1\u03c4\u03bf\u03c5 Valentin Vornicu \u03c0\u03b1\u03c1\u03cc\u03c4\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03be\u03b1\u03af\u03c1\u03b5\u03c4\u03bf\u03c2 \u03ba\u03b9 \u03c3\u03c4\u03bf \u03c6\u03cc\u03c1\u03bf\u03c5\u03bc \u03b1\u03bb\u03bb\u03ac \u03ba\u03b9 \u03c3\u03c4\u03bf\u03c5\u03c2 \u03b1\u03b3\u03ce\u03bd\u03b5\u03c2 \u03c3\u03c4\u03bf\u03c5\u03c2 \u03bf\u03c0\u03bf\u03af\u03bf\u03c5\u03c2 \u03ad\u03c6\u03c4\u03b9\u03b1\u03be\u03b5 \u03c4\u03bf \u03cc\u03bd\u03bf\u03bc\u03ac \u03c4\u03bf\u03c5 \u03c0\u03b9\u03c3\u03c4\u03b5\u03cd\u03c9 \u03cc\u03c4\u03b9 \u03b4\u03b5\u03bd \u03ad\u03c0\u03c1\u03b5\u03c0\u03b5 \u03bd\u03b1 \u03b5\u03c0\u03b9\u03bb\u03b5\u03b3\u03b5\u03af \u03b4\u03b9\u03cc\u03c4\u03b9 \u03b8\u03b1 \u03c5\u03c0\u03ae\u03c1\u03c7\u03b5 \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 \u03b5\u03c0\u03b9\u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1\u03c2 \u03b5\u03ba\u03c4\u03cc\u03c2 \u03ba\u03b9 \u03b1\u03bd \u03c5\u03c0\u03ae\u03c1\u03c7\u03b5 \u03ad\u03bd\u03b1\u03c2 \u03c0\u03bf\u03bb\u03cd \u03ba\u03b1\u03bb\u03cc\u03c2 \u03bc\u03b5\u03c4\u03b1\u03c6\u03c1\u03b1\u03c3\u03c4\u03ae\u03c2 \u03c0\u03b1\u03c1\u03cc\u03bb\u03b1 \u03b1\u03c5\u03c4\u03ac \u03ba\u03c1\u03af\u03bd\u03c9 \u03ba\u03b9 \u03b5\u03b3\u03ce \u03b5\u03c0\u03b9\u03c4\u03b1\u03ba\u03c4\u03b9\u03ba\u03ae \u03c4\u03b7\u03bd \u03b1\u03bd\u03ac\u03b3\u03ba\u03b7 \u03b5\u03bd\u03cc\u03c2 \u03c0\u03bf\u03bb\u03cd \u03ba\u03b1\u03bb\u03bf\u03cd \u03c0\u03c1\u03bf\u03c0\u03bf\u03bd\u03b7\u03c4\u03ae \u03b5\u03af\u03c4\u03b5 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u0395\u03bb\u03bb\u03ac\u03b4\u03b1 \u03b5\u03af\u03c4\u03b5 \u03b1\u03c0\u03cc \u03c4\u03bf \u03b5\u03be\u03c9\u03c4\u03b5\u03c1\u03b9\u03ba\u03cc \u03ba\u03b9 \u03c4\u03b7\u03bd \u03ac\u03bc\u03b5\u03c3\u03b7 \u03ba\u03b9 \u03b5\u03bd\u03c4\u03b1\u03c4\u03b9\u03ba\u03ae \u03b4\u03bf\u03c5\u03bb\u03b5\u03b9\u03ac \u03c4\u03bf\u03c5\r\n8.\t\u038c\u03c3\u03bf\u03bd \u03b1\u03c6\u03bf\u03c1\u03ac \u03c4\u03bf\u03c5\u03c2 \u039a\u03cd\u03c1\u03b9\u03bf\u03c5\u03c2 \u03a4\u03c5\u03c1\u03bb\u03ae \u03ba\u03b9 \u039a\u03bf\u03bd\u03c4\u03bf\u03b3\u03b9\u03ac\u03bd\u03bd\u03b7 \r\n9.\t\u039f\u03b9 \u03c0\u03b5\u03c1\u03b9\u03c3\u03c3\u03cc\u03c4\u03b5\u03c1\u03bf\u03b9 \u03b1\u03c0\u03cc \u03c4\u03bf \u03c6\u03cc\u03c1\u03bf\u03c5\u03bc \u03b3\u03bd\u03c9\u03c1\u03af\u03b6\u03bf\u03c5\u03bd \u03cc\u03c4\u03b9 \u03ad\u03c7\u03c9 \u03bc\u03b9\u03b1 \u03ba\u03cc\u03bd\u03c4\u03c1\u03b1 \u03ba\u03b9 \u03b4\u03b5\u03bd \u03c4\u03bf \u03ba\u03c1\u03cd\u03b2\u03c9 \u03cc\u03c0\u03c9\u03c2 \u03b2\u03bb\u03ad\u03c0\u03b5\u03b9\u03c2 \u03bf\u03cd\u03c4\u03b5 \u03ba\u03b9 \u03b5\u03b3\u03ce \u03bc\u03b5 \u03c4\u03bf\u03bd \u039a\u03cd\u03c1\u03b9\u03bf \u039a\u03bf\u03bd\u03c4\u03bf\u03b3\u03b9\u03ac\u03bd\u03bd\u03b7 \u03ba\u03b9 \u03c9\u03c2 \u03b5\u03ba \u03c4\u03bf\u03cd\u03c4\u03bf\u03c5 \u03b4\u03b5\u03bd \u03b8\u03b1 \u03b5\u03ba\u03c6\u03ad\u03c1\u03c9 \u03ba\u03b1\u03bc\u03af\u03b1 \u03b3\u03bd\u03ce\u03bc\u03b7 \u03ba\u03b1\u03b8\u03cc\u03c4\u03b9 \u03b8\u03b1 \u03b8\u03b5\u03c9\u03c1\u03b7\u03b8\u03b5\u03af \u03c5\u03c0\u03bf\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03ba\u03ae \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03b4\u03b5\u03bd \u03bc\u03c0\u03bf\u03c1\u03ce \u03bf\u03cd\u03c4\u03b5 \u03ba\u03b9 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03c4\u03bf \u03c3\u03c7\u03bf\u03bb\u03b9\u03ac\u03c3\u03c9 \u03ba\u03b1\u03b8\u03cc\u03bb\u03bf\u03c5 \u03b8\u03b1 \u03c0\u03c9 \u03cc\u03c4\u03b9 \u03c0\u03ac\u03b5\u03b9 \u03c0\u03ac\u03c1\u03b1 \u03c0\u03bf\u03bb\u03cd \u03b7 \u03b9\u03c3\u03c4\u03bf\u03c1\u03af\u03b1 \u03bc\u03b5 \u03b1\u03c5\u03c4\u03cc\u03bd \u03c4\u03bf\u03bd \u03ba\u03cd\u03c1\u03b9\u03bf \u03bf \u03bf\u03c0\u03bf\u03af\u03bf\u03c2 \u03c4\u03bf \u03bc\u03cc\u03bd\u03bf \u03c0\u03bf\u03c5 \u03be\u03ad\u03c1\u03b5\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03b2\u03c1\u03af\u03b6\u03b5\u03b9 \u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac (\u03ba\u03b1\u03c4\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2 \u03c3\u03c5\u03bc\u03bc\u03b1\u03b8\u03b7\u03c4\u03ce\u03bd \u03ba\u03b9 \u03c6\u03af\u03bb\u03c9\u03bd \u03bc\u03bf\u03c5 \u03ba\u03b9 \u03b7 \u03b4\u03b9\u03ba\u03ae \u03bc\u03bf\u03c5) \u03ba\u03b9 \u03c4\u03bf \u03bc\u03cc\u03bd\u03bf \u03ac\u03bb\u03bb\u03bf \u03c0\u03bf\u03c5 \u03be\u03ad\u03c1\u03b5\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b9\u03c3\u03cc\u03b2\u03b9\u03b1 \u03bc\u03ad\u03bb\u03bf\u03c2 \u03c4\u03b7\u03c2 \u03b5\u03c0\u03b9\u03c4\u03c1\u03bf\u03c0\u03ae\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03ce\u03bd \u03c7\u03ce\u03c1\u03b9\u03c2 \u03bd\u03b1 \u03c3\u03c5\u03bc\u03b2\u03ac\u03bb\u03bb\u03b5\u03b9 \u03ba\u03b9 \u03c0\u03ac\u03c1\u03b1 \u03c0\u03bf\u03bb\u03cd \u03b5\u03ba\u03c4\u03cc\u03c2 \u03b1\u03c0\u03cc \u03ad\u03bd\u03b1 \u03bc\u03ac\u03b8\u03b7\u03bc\u03b1 \u03ba\u03b9 2-3 \u03c3\u03b7\u03bc\u03b5\u03b9\u03ce\u03c3\u03b5\u03b9\u03c2 \u03b3\u03b9\u03b1 \u03c3\u03c5\u03bd\u03b4\u03c5\u03b1\u03c3\u03c4\u03b9\u03ba\u03ae \u03b3\u03b5\u03c9\u03bc\u03b5\u03c4\u03c1\u03af\u03b1 \u03ba\u03b9 \u03b3\u03b5\u03c9\u03bc\u03b5\u03c4\u03c1\u03af\u03b1 \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9 \u03ba\u03ac\u03bd\u03b5\u03b9 \u03ba\u03b9 \u03c4\u03af\u03c0\u03bf\u03c4\u03b1 \u03ac\u03bb\u03bb\u03bf \u03ac\u03c3\u03c7\u03b5\u03c4\u03b1 \u03bc\u03b5 \u03c4\u03b1 \u03b2\u03b9\u03b2\u03bb\u03af\u03b1 \u03c4\u03b1 \u03bf\u03c0\u03bf\u03af\u03b1 \u03b1\u03c2 \u03c3\u03b7\u03bc\u03b5\u03b9\u03c9\u03b8\u03b5\u03af \u03cc\u03c4\u03b9 \u03c4\u03b1 \u03ad\u03ba\u03b1\u03bd\u03b5 \u03c0\u03c1\u03b9\u03bd 25 \u03c7\u03c1\u03cc\u03bd\u03b9\u03b1 ! \u03ac\u03c1\u03b1 \u03b4\u03b5\u03bd \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03c2 \u03bd\u03b1 \u03bc\u03b5 \u03ba\u03b1\u03c4\u03b7\u03b3\u03bf\u03c1\u03ae\u03c3\u03b5\u03b9 \u03b3\u03b9\u03b1 \u03c5\u03c0\u03bf\u03ba\u03b5\u03b9\u03bc\u03b5\u03bd\u03b9\u03c3\u03bc\u03cc \u038c\u03bc\u03c9\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd \u039a\u03cd\u03c1\u03b9\u03bf \u03a4\u03c5\u03c1\u03bb\u03ae \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03c4\u03bf\u03bd \u03be\u03ad\u03c1\u03c9 \u03ba\u03b9 \u03be\u03ad\u03c1\u03c9 \u03c4\u03b9 \u03ad\u03c7\u03b5\u03b9 \u03ba\u03ac\u03bd\u03b5\u03b9 \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u03b5\u03c4\u03b1\u03b9\u03c1\u03b5\u03af\u03b1 \u03c0\u03cc\u03c3\u03bf \u03b1\u03b3\u03b1\u03c0\u03ac\u03b5\u03b9 \u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03cc\u03c3\u03bf \u03b4\u03ad\u03bd\u03b5\u03c4\u03b1\u03b9 \u03bc\u03b5 \u03b1\u03c5\u03c4\u03ac \u03c3\u03b5 \u039a\u0391\u0398\u0395 \u03b1\u03c0\u03bf\u03c3\u03c4\u03bf\u03bb\u03ae \u03c0\u03bf\u03c5 \u03c3\u03c5\u03bc\u03bc\u03b5\u03c4\u03ad\u03c7\u03b5\u03b9 \u03c0\u03cc\u03c3\u03b5\u03c2 \u03ce\u03c1\u03b5\u03c2 \u03c4\u03b7\u03bd \u03b7\u03bc\u03ad\u03c1\u03b1 \u03c4\u03c1\u03ad\u03c7\u03b5\u03b9 \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u03b5\u03c4\u03b1\u03b9\u03c1\u03b5\u03af\u03b1 \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03c4\u03bf\u03c5 \u03b4\u03af\u03bd\u03b5\u03b9 \u03ba\u03ad\u03c1\u03b4\u03bf\u03c2 \u03b5\u03bd\u03ce \u03b8\u03b1 \u03ad\u03c0\u03c1\u03b5\u03c0\u03b5 \u03b8\u03b1 \u03ae\u03b8\u03b5\u03bb\u03b1 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03c4\u03bf\u03bd \u03c3\u03c7\u03bf\u03bb\u03b9\u03ac\u03c3\u03b5\u03b9 \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03c2 \u03ba\u03b1\u03ba\u03ac \u03ba\u03b9 \u03b1\u03bd \u03b8\u03ad\u03bb\u03b5\u03b9 \u03bd\u03b1 \u03c4\u03bf \u03ba\u03ac\u03bd\u03b5\u03b9 \u03b3\u03b9\u03b1 \u03ba\u03ac\u03b8\u03b5 \u03ba\u03b1\u03ba\u03cc \u03c3\u03c7\u03cc\u03bb\u03b9\u03bf \u03b1\u03c2 \u03ad\u03c7\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03c0\u03bf\u03bb\u03cd \u03c4\u03c1\u03b1\u03bd\u03c4\u03b1\u03c7\u03c4\u03ae \u03b4\u03b9\u03ba\u03b1\u03b9\u03bf\u03bb\u03bf\u03b3\u03af\u03b1 \u03ba\u03b9 \u03cc\u03c7\u03b9 \u03bc\u03c0\u03bf\u03c5\u03c1\u03b4\u03bf\u03bb\u03bf\u03b3\u03af\u03b1 \u03cc\u03bc\u03c9\u03c2 \u03b8\u03b1 \u03c3\u03c5\u03bc\u03c6\u03c9\u03bd\u03ae\u03c3\u03c9 \u03c3\u03c4\u03bf\u03bd \u03c7\u03b1\u03c1\u03b1\u03ba\u03c4\u03b7\u03c1\u03b9\u03c3\u03bc\u03cc \u03b7 \u03b1\u03c0\u03cc\u03bb\u03c5\u03c4\u03b7 \u03b4\u03c5\u03b1\u03c1\u03c7\u03af\u03b1 \u03bc\u03cc\u03bd\u03bf \u03c3\u03c4\u03bf \u03c7\u03c1\u03bf\u03bd\u03b9\u03ba\u03cc \u03ba\u03b1\u03b8\u03b1\u03c1\u03ac \u03ba\u03bf\u03bc\u03bc\u03ac\u03c4\u03b9 \u03c4\u03b7\u03c2 \u03c6\u03c1\u03ac\u03c3\u03b7\u03c2 \u03ba\u03b9 \u03cc\u03c7\u03b9 \u03b1\u03c0\u03cc\u03bb\u03c5\u03c4\u03b1 \u03bc\u03cc\u03bd\u03bf \u03ba\u03b1\u03c4\u03ac \u03c4\u03bf \u03ae\u03bc\u03b9\u03c3\u03c5 \u03c3\u03c4\u03bf \u03ba\u03b1\u03c4\u03b1\u03b3\u03b3\u03b5\u03bb\u03c4\u03b9\u03ba\u03cc \u03c0\u03b5\u03c1\u03b9\u03b5\u03c7\u03cc\u03bc\u03b5\u03bd\u03cc \u03c4\u03b7\u03c2 \r\n10.\t\u0393\u03b9\u03b1 \u03c4\u03bf\u03bd \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 \r\n11.\t\u0391\u03bd\u03b1\u03bb\u03c5\u03c4\u03b9\u03ba\u03ac \u03ad\u03c7\u03c9 \u03cc\u03bb\u03b5\u03c2 \u03c4\u03b9\u03c2 \u03c0\u03b7\u03b3\u03ad\u03c2 \u03c4\u03c9\u03bd \u03b8\u03b5\u03bc\u03ac\u03c4\u03c9\u03bd \u03ba\u03b1\u03b9 \u03b1\u03c6\u03bf\u03cd \u03c4\u03b9\u03c2 \u03b3\u03bd\u03c9\u03c3\u03c4\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03c9 \u03b8\u03b1 \u03c4\u03bf\u03c0\u03bf\u03b8\u03b5\u03c4\u03b7\u03b8\u03ce \u0391 \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 \u0395\u03be\u03b5\u03c4\u03b1\u03c3\u03c4\u03b9\u03ba\u03ae \u0399\u03c9\u03b1\u03bd\u03bd\u03af\u03c9\u03bd \u0398\u03b5\u03cc\u03b4\u03c9\u03c1\u03bf\u03c5 \u039c\u03c0\u03cc\u03bb\u03b7 \u0399\u03bf\u03cd\u03bd\u03b9\u03bf\u03c2 2005-\u03a3\u03b5\u03c0\u03c4\u03ad\u03bc\u03b2\u03c1\u03b9\u03bf\u03c2 2005 \u0392 \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 Baltic Way 2003 problem 2 \u0393 \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 Baltic Way 2003 problem 3 \u0394 \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 \u0398\u03b5\u03cc\u03b4\u03c9\u03c1\u03bf\u03c5 \u039c\u03c0\u03cc\u03bb\u03b7 \u03c0\u03c1\u03bf\u03c4\u03b5\u03b9\u03bd\u03cc\u03bc\u03b5\u03bd\u03bf \u03b8\u03ad\u03bc\u03b1 \u03b3\u03b9\u03b1 \u03c4\u03bf \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03cc \u03c4\u03b7\u03c2 \u0391\u03b8\u03ae\u03bd\u03b1\u03c2 \u03ba\u03b9 \u03b1\u03c3\u03b1\u03c6\u03ad\u03c2 \u03cc\u03c0\u03c9\u03c2 \u03c0\u03bf\u03bb\u03cd \u03c3\u03c9\u03c3\u03c4\u03ac \u03bb\u03b5\u03c2 \u03c6\u03af\u03bb\u03b5 \u03a3\u03c4\u03b1\u03c1\u03c4 \u0391\u03c5\u03c4\u03cc \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03b1\u03c4\u03b5\u03be\u03bf\u03c7\u03ae\u03bd \u03bd\u03c4\u03c1\u03bf\u03c0\u03ae \u03b1\u03c6\u03b5\u03c4\u03ad\u03c1\u03bf\u03c5 \u03bb\u03bf\u03b3\u03b9\u03ba\u03cc \u03b1\u03c0\u03cc \u03bc\u03b9\u03b1 \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03b1\u03b4\u03b9\u03ac\u03c6\u03bf\u03c1\u03b7 \u03cc\u03c0\u03c9\u03c2 \u03c0\u03c1\u03bf\u03b5\u03af\u03c0\u03b1\u03bc\u03b5 \u03c7\u03ce\u03c1\u03b1 \u03ba\u03b9 \u03bc\u03cc\u03bd\u03bf \u03c3\u03b5 \u03b5\u03bc\u03ac\u03c2 \u03c0\u03bf\u03c5 \u03b1\u03b3\u03b1\u03c0\u03ac\u03bc\u03b5 \u03c4\u03b1 \u039c\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ac \u03ba\u03b9 \u03c4\u03b7\u03bd \u03c7\u03ce\u03c1\u03b1 \u03bc\u03b1\u03c2 \u03bc\u03ac\u03c2 \u03c0\u03bf\u03bd\u03ac \u03ba\u03b9 \u03bc\u03b1\u03c2 \u03c0\u03bb\u03b7\u03b3\u03ce\u03bd\u03b5\u03b9 \u03b5\u03bb\u03c0\u03af\u03b6\u03c9 \u03ba\u03ac\u03c0\u03bf\u03b9\u03b1 \u03c3\u03c4\u03b9\u03b3\u03bc\u03ae \u03bd\u03b1 \u03c4\u03bf \u03c3\u03c5\u03bd\u03b5\u03b9\u03b4\u03b7\u03c4\u03bf\u03c0\u03bf\u03b9\u03ae\u03c3\u03bf\u03c5\u03bd \u03b1\u03c5\u03c4\u03bf\u03af \u03c0\u03bf\u03c5 \u03b4\u03b9\u03bf\u03b9\u03ba\u03bf\u03cd\u03bd \u03c4\u03b7\u03bd \u0395\u039c\u0395 \u03b1\u03bd \u03ba\u03b9 \u03c4\u03bf \u03b2\u03bb\u03ad\u03c0\u03c9 \u03b4\u03cd\u03c3\u03ba\u03bf\u03bb\u03bf\r\n12.\t\u0393\u03b9\u03b1 \u03c4\u03b7\u03bd \u03c0\u03b1\u03c1\u03b1\u03af\u03c4\u03b7\u03c3\u03b7 \u03c4\u03bf\u03c5 \u03ba\u03c5\u03c1\u03af\u03bf\u03c5 \u039c\u03c0\u03cc\u03bb\u03b7 \r\n13.\t\u03a4\u03bf \u03cc\u03c4\u03b9 \u03bf \u03ba\u03cd\u03c1\u03b9\u03bf\u03c2 \u039c\u03c0\u03cc\u03bb\u03b7\u03c2 \u03c0\u03b1\u03c1\u03b1\u03b9\u03c4\u03ae\u03b8\u03b7\u03ba\u03b5 \u03bc\u03c5\u03c3\u03c4\u03b9\u03ba\u03ac (\u03b3\u03b9\u03b1 \u03c0\u03c1\u03bf\u03c3\u03c9\u03c0\u03b9\u03ba\u03bf\u03cd\u03c2 \u03bb\u03cc\u03b3\u03bf\u03c5\u03c2 \u03cc\u03c0\u03c9\u03c2 \u03b1\u03bd\u03b1\u03ba\u03bf\u03b9\u03bd\u03ce\u03b8\u03b7\u03ba\u03b5) \u03b4\u03b5\u03af\u03c7\u03bd\u03b5\u03b9 \u03c3\u03ba\u03bf\u03c4\u03ac\u03b4\u03b9 \u03ba\u03b9 \u03c0\u03c1\u03bf\u03be\u03b5\u03bd\u03b5\u03af \u03c0\u03bf\u03bd\u03b7\u03c1\u03ad\u03c2 \u03c3\u03ba\u03ad\u03c8\u03b5\u03b9\u03c2 \u03ba\u03b9 \u03c4\u03b9 \u03b8\u03b1 \u03ac\u03bb\u03bb\u03b1\u03b6\u03b5 \u03cc\u03bc\u03c9\u03c2 \u03b1\u03bd \u03c4\u03bf \u03bc\u03b1\u03b8\u03b1\u03af\u03bd\u03b1\u03bc\u03b5? \u0395\u03b3\u03ce \u03c0\u03b9\u03c3\u03c4\u03b5\u03cd\u03c9 \u03cc\u03c4\u03b9 \u03ba\u03b9 \u03b1\u03bd \u03c4\u03bf \u03bc\u03ac\u03b8\u03bf\u03c5\u03bc\u03b5 \u03c0\u03bf\u03c4\u03ad \u03ba\u03b9 \u03b1\u03bd \u03cc\u03c7\u03b9 (\u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03b8\u03b1 \u03c4\u03bf \u03bc\u03ac\u03b8\u03bf\u03c5\u03bc\u03b5 \u03c0\u03bf\u03c4\u03ad \u03c3\u03c4\u03b7\u03bd \u0395\u03bb\u03bb\u03ac\u03b4\u03b1 \u03b6\u03bf\u03cd\u03bc\u03b5 \u03c4\u03ce\u03c1\u03b1) \u03b7 \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7 \u03b8\u03b1 \u03c0\u03b1\u03c1\u03b1\u03bc\u03b5\u03af\u03bd\u03b5\u03b9 \u03b7 \u03af\u03b4\u03b9\u03b1 \u03ac\u03c3\u03c7\u03b7\u03bc\u03b7 \u03b5\u03b9\u03ba\u03cc\u03bd\u03b1 \u03ac\u03c1\u03b1 \u03ba\u03b1\u03bb\u03cc \u03b8\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bd\u03b1 \u03ba\u03bf\u03b9\u03c4\u03ac\u03bc\u03b5 \u03bc\u03c0\u03c1\u03bf\u03c3\u03c4\u03ac \u03c3\u03c4\u03bf \u03bc\u03ad\u03bb\u03bb\u03bf\u03bd \u03ba\u03b9 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03b1\u03c3\u03c7\u03bf\u03bb\u03bf\u03cd\u03bc\u03b1\u03c3\u03c4\u03b5 \u03bc\u03b5 \u03c4\u03bf \u03c0\u03b1\u03c1\u03b5\u03bb\u03b8\u03cc\u03bd \u03a6\u03c5\u03c3\u03b9\u03ba\u03ac \u03cc\u03bc\u03c9\u03c2 \u03ba\u03b9 \u03b8\u03ad\u03bb\u03c9 \u03ba\u03b9 \u03b8\u03b5\u03c9\u03c1\u03ce \u03c5\u03c0\u03bf\u03c7\u03c1\u03ad\u03c9\u03c3\u03b7 \u03c4\u03c9\u03bd \u03c5\u03c0\u03b5\u03c5\u03b8\u03cd\u03bd\u03c9\u03bd \u03ba\u03b9 \u03c4\u03bf\u03c5 \u039a \u039c\u03c0\u03cc\u03bb\u03b7 \u03bd\u03b1 \u03b5\u03bd\u03b7\u03bc\u03b5\u03c1\u03ce\u03c3\u03b5\u03b9 \u03c4\u03bf\u03c5\u03c2 \u03bb\u03cc\u03b3\u03bf\u03c5\u03c2 \u03c4\u03b7\u03c2 \u03b1\u03c0\u03bf\u03c7\u03ce\u03c1\u03b7\u03c3\u03ae\u03c2 \u03c4\u03bf\u03c5\r\n14.\t\u0393\u03b9\u03b1 \u03c4\u03b7\u03bd \u03b4\u03b9\u03ac\u03c1\u03ba\u03b5\u03b9\u03b1 \u03c4\u03bf\u03c5 \u0391\u03c1\u03c7\u03b9\u03bc\u03ae\u03b4\u03b7 \u03ba\u03b9 \u03c4\u03b7\u03bd \u03b5\u03c0\u03b9\u03c4\u03ae\u03c1\u03b7\u03c3\u03b7 \r\n15.\t\u03a0\u03c1\u03bf\u03c3\u03c9\u03c0\u03b9\u03ba\u03ac \u03b8\u03b5\u03c9\u03c1\u03ce \u03cc\u03c4\u03b9 \u03bf\u03b9 3 \u03ce\u03c1\u03b5\u03c2 \u03b5\u03af\u03bd\u03b1\u03b9 \u03b5\u03bd\u03b4\u03b5\u03b9\u03ba\u03c4\u03b9\u03ba\u03ad\u03c2 \u03ba\u03b9 \u03cc\u03c4\u03b9 \u03b4\u03b5\u03bd \u03b8\u03b1 \u03b1\u03bb\u03bb\u03ac\u03be\u03bf\u03c5\u03bd \u03ba\u03b9 \u03c0\u03bf\u03bb\u03bb\u03ac \u03b1\u03bd \u03b1\u03c5\u03be\u03b7\u03b8\u03b5\u03af \u03bc\u03b5 \u03c4\u03b7\u03bd \u03b4\u03b9\u03ac\u03c1\u03ba\u03b5\u03b9\u03b1 \u03c4\u03bf\u03c5 \u03b4\u03b9\u03cc\u03c4\u03b9 \u03b1\u03c2 \u03bc\u03b7\u03bd \u03bb\u03ad\u03bc\u03b5 \u03cc,\u03c4\u03b9 \u03b8\u03ad\u03bb\u03bf\u03c5\u03bc\u03b5 \u03b1\u03bd \u03b4\u03b5\u03bd \u03bc\u03c0\u03bf\u03c1\u03b5\u03af\u03c2 \u03bd\u03b1 \u03bb\u03cd\u03c3\u03b5\u03b9\u03c2 \u03c4\u03b1 \u03b8\u03ad\u03bc\u03b1\u03c4\u03b1 \u03c3\u03b5 3 \u03ce\u03c1\u03b5\u03c2 \u03b4\u03b5\u03bd \u03b8\u03b1 \u03bc\u03c0\u03bf\u03c1\u03ad\u03c3\u03b5\u03b9\u03c2 \u03b5\u03cd\u03ba\u03bf\u03bb\u03b1 \u03c4\u03bf\u03c5\u03bb\u03ac\u03c7\u03b9\u03c3\u03c4\u03bf\u03bd \u03bd\u03b1 \u03c4\u03bf \u03ba\u03ac\u03bd\u03b5\u03b9\u03c2 \u03c3\u03b5 1-1.5 \u03ce\u03c1\u03b1 \u03ad\u03be\u03c4\u03c1\u03b1 \u03c7\u03c1\u03cc\u03bd\u03bf \u03c0\u03b9\u03b8\u03b1\u03bd\u03ce\u03c2 \u03ad\u03c4\u03c3\u03b9 \u03c3\u03ba\u03b5\u03c0\u03c4\u03cc\u03bc\u03b5\u03bd\u03bf\u03b9 \u03bf\u03b9 \u03c4\u03b7\u03c2 \u0395\u039c\u0395 \u03b4\u03b5\u03bd \u03c4\u03bf \u03ad\u03ba\u03b1\u03bd\u03b1\u03bd \u03b4\u03b5\u03ba\u03c4\u03cc \u03c4\u03bf \u03b1\u03af\u03c4\u03b7\u03bc\u03b1 \u0397 \u03b5\u03c0\u03b9\u03c4\u03ae\u03c1\u03b7\u03c3\u03b7 \u03b5\u03c0\u03af\u03c3\u03b7\u03c2 \u03b8\u03b1 \u03c3\u03c5\u03bc\u03c6\u03c9\u03bd\u03ae\u03c3\u03c9 \u03cc\u03c4\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03ac\u03b8\u03bb\u03b9\u03b1 \u03ba\u03b9 \u03b8\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03c0\u03c1\u03bf\u03c3\u03b5\u03c7\u03b8\u03b5\u03af \u03b1\u03c5\u03c4\u03cc \u03c0\u03bf\u03bb\u03cd \u03b1\u03bd \u03b8\u03ad\u03bb\u03bf\u03c5\u03bc\u03b5 \u03bd\u03b1 \u03ba\u03c1\u03b1\u03c4\u03ae\u03c3\u03bf\u03c5\u03bc\u03b5 \u03c3\u03b5 \u03c5\u03c8\u03b7\u03bb\u03cc \u03b5\u03c0\u03af\u03c0\u03b5\u03b4\u03bf \u03c4\u03bf \u03ba\u03cd\u03c1\u03bf\u03c2 \u03c4\u03b7\u03c2 \u0395\u0398\u039d\u0399\u039a\u0397\u03a3 \u039c\u0391\u0398\u0397\u039c\u0391\u03a4\u0399\u039a\u0397\u03a3 \u039f\u039b\u03a5\u039c\u03a0\u0399\u0391\u0394\u0391\u03a3 \u03b5\u03b3\u03ce \u03c3\u03c5\u03b3\u03ba\u03b5\u03ba\u03c1\u03b9\u03bc\u03ad\u03bd\u03b1 \u03c3\u03c4\u03bf\u03bd \u0395\u03c5\u03ba\u03bb\u03b5\u03af\u03b4\u03b7 \u03bf \u03b4\u03b9\u03c0\u03bb\u03b1\u03bd\u03cc\u03c2 \u03bc\u03bf\u03c5 \u03bc\u03b5 \u03b5\u03af\u03c7\u03b5 \u03c4\u03c1\u03b5\u03bb\u03ac\u03bd\u03b5\u03b9 \u03c6\u03c9\u03bd\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03c2 \u03c4\u03bf\u03bd \u03bc\u03c0\u03c1\u03bf\u03c3\u03c4\u03b9\u03bd\u03cc \u03c4\u03bf\u03c5 (\u039a\u03ce\u03c3\u03c4\u03b1 \u039a\u03ce\u03c3\u03c4\u03b1-\u039a\u03ce\u03c3\u03c4\u03b1) \u03c3\u03ba\u03ad\u03c4\u03b7 \u03b1\u03b7\u03b4\u03af\u03b1 \u03b1\u03c5\u03c4\u03cc \u03c4\u03bf \u03c0\u03c1\u03ac\u03b3\u03bc\u03b1 \u03c4\u03bf \u03cc\u03c4\u03b9 \u03b4\u03b5\u03bd \u03c4\u03bf\u03bd \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03b5\u03b9 \u03cc\u03bc\u03c9\u03c2 \u03c6\u03af\u03bb\u03b5 \u03a3\u03c4\u03ac\u03c1\u03c4 \u03ba\u03b1\u03c4\u03b1\u03b4\u03b5\u03b9\u03ba\u03bd\u03cd\u03b5\u03b9 \u03ad\u03bd\u03b1 \u03ac\u03bb\u03bb\u03bf \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03bf \u03c4\u03b7\u03c2 \u03b1\u03b9\u03ce\u03bd\u03b9\u03b1\u03c2 \u03b5\u03bb\u03bb\u03b7\u03bd\u03b9\u03ba\u03ae\u03c2 \u03b9\u03b4\u03b9\u03bf\u03c3\u03c5\u03b3\u03ba\u03c1\u03b1\u03c3\u03af\u03b1\u03c2 (\u03c4\u03bf \u03cc, \u03c4\u03b9 \u03bc\u03b5 \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03b5\u03b9 \u03ba\u03b9 \u03c3\u03c4\u03b1 \u03c4\u03c3\u03b1\u03ba\u03af\u03b4\u03b9\u03b1 \u03cc\u03bb\u03b1 \u03c4\u03b1 \u03ac\u03bb\u03bb\u03b1) \u03b1\u03bd \u03b4\u03b5\u03bd \u03c3\u03b5 \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03b5\u03b9 \u03ba\u03cd\u03c1\u03b9\u03b5 (\u03b3\u03b5\u03bd\u03b9\u03ba\u03cc\u03c2 \u03c7\u03b1\u03c1\u03b1\u03ba\u03c4\u03b7\u03c1\u03b9\u03c3\u03bc\u03cc\u03c2 \u03ba\u03b9 \u03ba\u03bb\u03ae\u03c3\u03b7 \u03b4\u03b5\u03bd \u03b1\u03c0\u03b5\u03c5\u03b8\u03cd\u03bd\u03b5\u03c4\u03b1\u03b9 \u03c3\u03b5 \u03c3\u03c5\u03b3\u03ba\u03b5\u03ba\u03c1\u03b9\u03bc\u03ad\u03bd\u03bf \u03c0\u03c1\u03cc\u03c3\u03c9\u03c0\u03bf) \u039d\u0391 \u039c\u0397\u039d \u03a3\u03a5\u039c\u039c\u0395\u03a4\u0391\u03a3\u03a7\u0395\u0399\u03a3 \u03a3\u03a4\u0397\u039d \u0395\u0398\u039d\u0399\u039a\u0397 \u039c\u0391\u0398\u0397\u039c\u0391\u03a4\u0399\u039a\u0397 \u039f\u039b\u03a5\u039c\u03a0\u0399\u0391\u0394\u0391 \u03c4\u03ce\u03c1\u03b1 \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u03b3\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03b9 \u03c5\u03c0\u03b1\u03b9\u03bd\u03b9\u03b3\u03bc\u03bf\u03af \u03ba\u03b9 \u03b3\u03b9\u03b1 \u03c3\u03c5\u03bc\u03c6\u03ad\u03c1\u03bf\u03bd\u03c4\u03b1 \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u0395\u039c\u0395 \u03b8\u03b1 \u03c0\u03c9 \u03cc\u03c4\u03b9 \u03b1\u03c0\u03cc \u03c4\u03b9\u03c2 \u03ba\u03b1\u03c4\u03b1\u03b3\u03b3\u03b5\u03bb\u03af\u03b5\u03c2 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af \u03bd\u03b1 \u03c3\u03c5\u03bc\u03c0\u03b5\u03c1\u03ac\u03bd\u03b5\u03b9 \u03ba\u03b1\u03bd\u03b5\u03af\u03c2 \u03cc\u03c4\u03b9 \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd \u03b1\u03bb\u03bb\u03ac \u03b8\u03b1 \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03b1\u03c0\u03cc\u03bb\u03c5\u03c4\u03bf\u03c2 \u03b1\u03bd \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9 \u03b1\u03c0\u03c4\u03ac \u03c3\u03c4\u03bf\u03b9\u03c7\u03b5\u03af\u03b1 \u03c3\u03c4\u03b1 \u03c7\u03ad\u03c1\u03b9\u03b1 \u03c4\u03bf\u03c5 \u03b3\u03b9\u03b1\u03c4\u03af \u03b4\u03b5\u03bd \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03b9 \u03c3\u03c9\u03c3\u03c4\u03cc \u03b8\u03af\u03b3\u03b5\u03b9 \u03ad\u03bc\u03bc\u03b5\u03c3\u03b1 \u03ba\u03b9 \u03c4\u03bf\u03bd \u03b5\u03b1\u03c5\u03c4\u03cc \u03b4\u03b9\u03cc\u03c4\u03b9 \u03c4\u03bf\u03bd \u03b8\u03ad\u03c4\u03b5\u03b9 \u03c3\u03c4\u03b1 \u03bc\u03ac\u03c4\u03b9\u03b1 \u03c4\u03c9\u03bd \u03ac\u03bb\u03bb\u03c9\u03bd \u03c0\u03b9\u03b8\u03b1\u03bd\u03cc \u03b5\u03c5\u03bd\u03bf\u03bf\u03cd\u03bc\u03b5\u03bd\u03bf \u03c4\u03b7\u03c2 \u03ba\u03b1\u03c4\u03ac\u03c3\u03c4\u03b1\u03c3\u03b7\u03c2\r\n16.\t\u0393\u03b9\u03b1 \u03c4\u03b7\u03bd \u039c\u03b5\u03c3\u03bf\u03b3\u03b5\u03b9\u03ac\u03b4\u03b1 \u03ba\u03b9 \u03c4\u03bf \u03bb\u03ac\u03b8\u03bf\u03c2 \u03b8\u03ad\u03bc\u03b1\r\n17.\t\u0391\u03c5\u03c4\u03cc \u03c4\u03bf \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 \u03c4\u03bf \u03b1\u03bd\u03c4\u03b9\u03bc\u03b5\u03c4\u03ce\u03c0\u03b9\u03c3\u03b1\u03bd \u03cc\u03bb\u03b5\u03c2 \u03bf\u03b9 \u03bc\u03b5\u03c3\u03bf\u03b3\u03b5\u03b9\u03b1\u03ba\u03ad\u03c2 \u03c7\u03ce\u03c1\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03ad\u03b4\u03c9\u03c3\u03b1\u03bd \u03b1\u03c5\u03c4\u03cc\u03bd \u03c4\u03bf\u03bd \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03cc \u03ba\u03b9 \u03c4\u03b1 \u03b8\u03ad\u03bc\u03b1\u03c4\u03b1 \u03c4\u03b1 \u03ad\u03c7\u03bf\u03c5\u03bd \u03c0\u03c1\u03bf\u03c6\u03b1\u03bd\u03ce\u03c2 \u03bb\u03c5\u03bc\u03ad\u03bd\u03b1 \u03ac\u03c1\u03b1 \u03c0\u03b9\u03bf \u03c0\u03b9\u03b8\u03b1\u03bd\u03ae \u03b5\u03ba\u03b4\u03bf\u03c7\u03ae \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b1\u03c1\u03c7\u03ae \u03bb\u03ac\u03b8\u03bf\u03c2 \u03ba\u03b9 \u03b4\u03b5\u03bd \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03ba\u03b1\u03c4\u03b7\u03b3\u03bf\u03c1\u03bf\u03cd\u03bc\u03b5 \u03c3\u03b5 \u03c4\u03b5\u03bb\u03b9\u03ba\u03ae \u03b1\u03bd\u03ac\u03bb\u03c5\u03c3\u03b7 \u03c4\u03bf \u03b8\u03ad\u03bc\u03b1 \u03b4\u03b5\u03bd \u03ad\u03b4\u03c9\u03c3\u03b5 \u03b1\u03bd\u03b9\u03c3\u03cc\u03c4\u03b7\u03c4\u03b1 \u03bc\u03b5\u03c4\u03b1\u03be\u03cd \u03c4\u03c9\u03bd \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03b6\u03cc\u03bc\u03b5\u03bd\u03c9\u03bd \u03ba\u03b1\u03b8\u03cc\u03c4\u03b9 \u03cc\u03bb\u03bf\u03b9 \u03c0\u03ae\u03c1\u03b1\u03bd \u03c4\u03bf\u03c5\u03c2 \u03b2\u03b1\u03b8\u03bc\u03bf\u03cd\u03c2 \u03c4\u03b7\u03c2 3\u03b2\r\n18.\t\u03a4\u03ad\u03bb\u03bf\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd DULAC \u03ba\u03b9 \u03c4\u03bf\u03c5 \u03c3\u03c7\u03bf\u03bb\u03af\u03bf\u03c5 \u03c4\u03bf\u03c5 \r\n19.\t\u0391\u03c1\u03c7\u03b9\u03ba\u03ac \u03b8\u03b1 \u03c3\u03bf\u03c5 \u03ba\u03ac\u03bd\u03c9 \u03bb\u03af\u03b3\u03bf \u03ba\u03c9\u03bc\u03b9\u03ba\u03ae \u03ba\u03c1\u03b9\u03c4\u03b9\u03ba\u03ae \u0393\u03b9\u03b1\u03c4\u03af \u03c6\u03af\u03bb\u03b5 \u03bc\u03bf\u03c5 \u03a3\u03c4\u03ac\u03c1\u03c4 \u03c3\u03b5 \u03c0\u03b5\u03b9\u03c1\u03ac\u03b6\u03b5\u03b9 \u03c0\u03bf\u03c5 \u03be\u03ad\u03c1\u03b5\u03b9 \u03bf DULAC \u03c4\u03cc\u03c3\u03b1 \u03c0\u03bf\u03bb\u03bb\u03ac \u03b8\u03b1 \u03c4\u03bf\u03bd \u03c6\u03b9\u03bc\u03ce\u03c3\u03b5\u03b9\u03c2 \u03ce\u03c3\u03c4\u03b5 \u03bd\u03b1 \u03bc\u03b7\u03bd \u03bc\u03b9\u03bb\u03ae\u03c3\u03b5\u03b9 ? \u03ae \u03b8\u03b1 \u03c4\u03bf\u03bd \u03c3\u03ba\u03bf\u03c4\u03ce\u03c3\u03b5\u03b9\u03c2 \u03b5\u03c0\u03b5\u03b9\u03b4\u03ae \u00ab\u03bc\u03af\u03bb\u03b7\u03c3\u03b5\u00bb? \u03a4\u03b9 \u03c4\u03bf \u03c4\u03cc\u03c3\u03bf \u03c0\u03b1\u03c1\u03ac\u03bd\u03bf\u03bc\u03bf \u03ac\u03c4\u03b9\u03bc\u03bf \u03ae \u03c0\u03bf\u03bd\u03b7\u03c1\u03cc \u03b2\u03c1\u03af\u03c3\u03ba\u03b5\u03b9\u03c2 \u03c3\u03b5 \u03b1\u03c5\u03c4\u03cc ? \u03c0\u03ac\u03bd\u03c4\u03c9\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03bf \u03c3\u03c7\u03cc\u03bb\u03b9\u03bf \u03c4\u03bf\u03c5 DULAC \u03b8\u03b1 \u03c4\u03bf \u03c3\u03c7\u03bf\u03bb\u03b9\u03ac\u03c3\u03c9 \u03cc\u03c4\u03b9 \u03c4\u03bf \u03b1 \u03c8\u03ac\u03c7\u03bd\u03b5\u03b9 \u03bf \u039a \u0396\u03ce\u03c4\u03bf\u03c2 \u03b3\u03b9\u03b1 \u03c5\u03c0\u03b1\u03c1\u03c7\u03b7\u03b3\u03cc \u03b4\u03b5\u03bd \u03bc\u03bf\u03c5 \u03bb\u03ad\u03b5\u03b9 \u03c4\u03af\u03c0\u03bf\u03c4\u03b1 \u03c4\u03bf\u03c5\u03bb\u03ac\u03c7\u03b9\u03c3\u03c4\u03bf \u03c4\u03af\u03c0\u03bf\u03c4\u03b1 \u03c0\u03bf\u03bd\u03b7\u03c1\u03cc \u03c4\u03bf \u03c3\u03c7\u03cc\u03bb\u03b9\u03bf \u03b3\u03b9\u03b1 \u03c4\u03bf\u03bd \u03ba\u03cd\u03c1\u03b9\u03bf \u039c\u03c0\u03cc\u03bb\u03b7 \u03ba\u03b9 \u03c4\u03bf\u03bd \u039a\u03cd\u03c1\u03b9\u03bf \u03a6\u03b5\u03bb\u03bb\u03bf\u03cd\u03c1\u03b7 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c0\u03bb\u03b7\u03c1\u03bf\u03c6\u03bf\u03c1\u03af\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03c4\u03b9\u03c2 \u03be\u03ad\u03c1\u03bf\u03c5\u03bd \u03ba\u03b9 \u03bf\u03b9 \u03c0\u03ad\u03c4\u03c1\u03b5\u03c2 \u03ba\u03b9 \u03cc\u03c0\u03bf\u03b9\u03bf\u03bd \u03ba\u03b9 \u03b1\u03bd \u03c1\u03c9\u03c4\u03ae\u03c3\u03b5\u03b9\u03c2 \u03c4\u03b9\u03c2 \u03be\u03ad\u03c1\u03b5\u03b9 \u03ac\u03c1\u03b1 \u03b4\u03b5\u03bd \u03c3\u03c5\u03bd\u03b9\u03c3\u03c4\u03ac \u03bf\u03cd\u03c4\u03b5 \u03ba\u03b9 \u03b1\u03c5\u03c4\u03cc \u03ba\u03ac\u03c0\u03bf\u03b9\u03bf\u03bd \u03c0\u03bf\u03bd\u03b7\u03c1\u03cc \u03c0\u03c1\u03bf\u03b2\u03bb\u03b7\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03cc \r\n\u0394\u03b5\u03bd \u03bd\u03bf\u03bc\u03af\u03b6\u03c9 \u03cc\u03c4\u03b9 \u03c3\u03b5 \u03ac\u03c6\u03b7\u03c3\u03b1 \u03c0\u03b1\u03c1\u03b1\u03c0\u03bf\u03bd\u03b5\u03bc\u03ad\u03bd\u03bf \u03c6\u03af\u03bb\u03b5 \u03bc\u03bf\u03c5 \u03a3\u03c4\u03ac\u03c1\u03c4 \u03ba\u03bb\u03b5\u03af\u03bd\u03bf\u03bd\u03c4\u03b1\u03c2 \u03b8\u03b1 \u03ae\u03b8\u03b5\u03bb\u03b1 \u03bd\u03b1 \u03c3\u03bf\u03c5 \u03b5\u03c5\u03c7\u03b7\u03b8\u03ce \u03ba\u03b1\u03bb\u03ae \u03c3\u03c5\u03bd\u03ad\u03c7\u03b5\u03b9\u03b1 \u03ba\u03b9 \u03ba\u03b1\u03bb\u03ae \u03bc\u03b5\u03bb\u03ad\u03c4\u03b7 \u03b4\u03bf\u03c5\u03bb\u03b5\u03af\u03b1 \u03ba\u03b9 \u03c4\u03b1 \u03af\u03b4\u03b9\u03b1 \u03b8\u03b1 \u03ae\u03b8\u03b5\u03bb\u03b1 \u03bd\u03b1 \u03b5\u03c5\u03c7\u03b7\u03b8\u03ce \u03c3\u03b5 \u03cc\u03bb\u03bf\u03c5\u03c2 \u03bc\u03b1\u03c2 \u03b5\u03bd\u03cc\u03c8\u03b5\u03b9 \u03c4\u03b7\u03c2 \u03bd\u03ad\u03b1\u03c2 \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae\u03c2 \u03c7\u03c1\u03bf\u03bd\u03b9\u03ac\u03c2 \u03b7 \u03bf\u03c0\u03bf\u03af\u03b1 \u03b5\u03bb\u03c0\u03af\u03b6\u03c9 \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03bc\u03b5 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b5\u03c2 \r\n\u039c\u03b5 \u03c3\u03c5\u03bd\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03c4\u03b9\u03ba\u03bf\u03cd\u03c2 \u03c7\u03b1\u03b9\u03c1\u03b5\u03c4\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2",
"Solution_2": "\u039f\u03a7\u0399 \u03a0\u0391\u039b\u0399 \u03a1\u0395 \u03a0\u0391\u0399\u0394\u0399\u0391 :mad: :mad: :cursing:",
"Solution_3": "[quote=\"ambros\"]\n16.\t\u0393\u03b9\u03b1 \u03c4\u03b7\u03bd \u039c\u03b5\u03c3\u03bf\u03b3\u03b5\u03b9\u03ac\u03b4\u03b1 \u03ba\u03b9 \u03c4\u03bf \u03bb\u03ac\u03b8\u03bf\u03c2 \u03b8\u03ad\u03bc\u03b1\n17.\t\u0391\u03c5\u03c4\u03cc \u03c4\u03bf \u03c0\u03c1\u03cc\u03b2\u03bb\u03b7\u03bc\u03b1 \u03c4\u03bf \u03b1\u03bd\u03c4\u03b9\u03bc\u03b5\u03c4\u03ce\u03c0\u03b9\u03c3\u03b1\u03bd \u03cc\u03bb\u03b5\u03c2 \u03bf\u03b9 \u03bc\u03b5\u03c3\u03bf\u03b3\u03b5\u03b9\u03b1\u03ba\u03ad\u03c2 \u03c7\u03ce\u03c1\u03b5\u03c2 \u03c0\u03bf\u03c5 \u03ad\u03b4\u03c9\u03c3\u03b1\u03bd \u03b1\u03c5\u03c4\u03cc\u03bd \u03c4\u03bf\u03bd \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03cc \u03ba\u03b9 \u03c4\u03b1 \u03b8\u03ad\u03bc\u03b1\u03c4\u03b1 \u03c4\u03b1 \u03ad\u03c7\u03bf\u03c5\u03bd \u03c0\u03c1\u03bf\u03c6\u03b1\u03bd\u03ce\u03c2 \u03bb\u03c5\u03bc\u03ad\u03bd\u03b1 \u03ac\u03c1\u03b1 \u03c0\u03b9\u03bf \u03c0\u03b9\u03b8\u03b1\u03bd\u03ae \u03b5\u03ba\u03b4\u03bf\u03c7\u03ae \u03b5\u03af\u03bd\u03b1\u03b9 \u03c4\u03bf \u03b1\u03c0\u03cc \u03c4\u03b7\u03bd \u03b1\u03c1\u03c7\u03ae \u03bb\u03ac\u03b8\u03bf\u03c2 \u03ba\u03b9 \u03b4\u03b5\u03bd \u03c0\u03c1\u03ad\u03c0\u03b5\u03b9 \u03bd\u03b1 \u03ba\u03b1\u03c4\u03b7\u03b3\u03bf\u03c1\u03bf\u03cd\u03bc\u03b5 \u03c3\u03b5 \u03c4\u03b5\u03bb\u03b9\u03ba\u03ae \u03b1\u03bd\u03ac\u03bb\u03c5\u03c3\u03b7 \u03c4\u03bf \u03b8\u03ad\u03bc\u03b1 \u03b4\u03b5\u03bd \u03ad\u03b4\u03c9\u03c3\u03b5 \u03b1\u03bd\u03b9\u03c3\u03cc\u03c4\u03b7\u03c4\u03b1 \u03bc\u03b5\u03c4\u03b1\u03be\u03cd \u03c4\u03c9\u03bd \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03b6\u03cc\u03bc\u03b5\u03bd\u03c9\u03bd \u03ba\u03b1\u03b8\u03cc\u03c4\u03b9 \u03cc\u03bb\u03bf\u03b9 \u03c0\u03ae\u03c1\u03b1\u03bd \u03c4\u03bf\u03c5\u03c2 \u03b2\u03b1\u03b8\u03bc\u03bf\u03cd\u03c2 \u03c4\u03b7\u03c2 3\u03b2\n[/quote]\r\n\r\n\u03a3\u03b1\u03c6\u03ce\u03c2 \u03b7 \u03b4\u03b9\u03b1\u03c4\u03cd\u03c0\u03c9\u03c3\u03b7 \u03c3\u03bf\u03c5 \u03b4\u03b5\u03bd \u03b1\u03c6\u03ae\u03bd\u03b5\u03b9 \u03b1\u03bc\u03c6\u03b9\u03b2\u03bf\u03bb\u03af\u03b1 \u03b1\u03bb\u03bb\u03ac [b] \u03cc\u03bb\u03bf\u03b9 [/b] \u03c0\u03ae\u03c1\u03b1\u03bd \u03c4\u03bf\u03c5\u03c2 \u03b2\u03b1\u03b8\u03bc\u03bf\u03cd\u03c2? \u0395\u03c0\u03b5\u03b9\u03b4\u03ae \u03c3\u03c5\u03bc\u03bc\u03b5\u03c4\u03b5\u03af\u03c7\u03b1 (\u03cc\u03c7\u03b9 \u03b2\u03ad\u03b2\u03b1\u03b9\u03b1 \u03c0\u03c9\u03c2 \u03b5\u03af\u03c7\u03b1 \u03b1\u03be\u03b9\u03ce\u03c3\u03b5\u03b9\u03c2)...",
"Solution_4": "\u0391\u03b3\u03b1\u03c0\u03b7\u03c4\u03ad \u03c6\u03af\u03bb\u03b5 \u03bc\u03bf\u03c5 \u03ad\u03c4\u03c3\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u039f\u039b\u039f\u0399 \u0391\u039d\u0395\u039e\u0391\u0399\u03a1\u0395\u03a4\u0391 \u03c4\u03bf\u03c5\u03c2 \u03c0\u03ae\u03c1\u03b1\u03bd \u03c4\u03bf\u03c5\u03c2 \u03c0\u03cc\u03bd\u03c4\u03bf\u03c5\u03c2 \u03ba\u03b9 \u03bc\u03b7\u03bd \u03bb\u03ad\u03c2 \u03bf\u03c4\u03b9 \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9\u03c2 \u03b1\u03be\u03b9\u03ce\u03c3\u03b5\u03b9\u03c2 \u03b3\u03b9\u03b1\u03c4\u03af \u03b1\u03ba\u03cc\u03bc\u03b1 \u03ba\u03b9 \u03ad\u03c4\u03c3\u03b9 \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u039a\u0391\u039d\u0395\u0399\u03a3 \u039c\u0391 \u039a\u0391\u039d\u0395\u0399\u03a3 \u0394\u0395\u039d \u0395\u03a7\u0395\u0399 \u0394\u0399\u039a\u0391\u0399\u03a9\u039c\u0391 \u039d\u0391 \u03a3\u039f\u03a5 \u03a3\u03a4\u0395\u03a1\u0395\u0399 \u039c\u039f\u039d\u0391\u0394\u0395\u03a3 \u0391\u03a0\u039f \u03a4\u039f\u039d \u039a\u039f\u03a0\u039f \u03a3\u039f\u03a5 \r\n\u039a\u03b1\u03bb\u03b7 \u03a3\u03c5\u03bd\u03ad\u03c7\u03b5\u03b9\u03b1 \u03bc\u03b5 \u03b5\u03c0\u03b9\u03c4\u03c5\u03c7\u03af\u03b5\u03c2 \r\n\u03c7\u03b1\u03b9\u03c1\u03b5\u03c4\u03ce",
"Solution_5": "\u03a6\u03af\u03bb\u03b5 start, \r\n\r\n \u03c4\u03bf \u03be\u03ad\u03c1\u03b5\u03b9\u03c2 \u03cc\u03c4\u03b9 \u03b3\u03b5\u03bd\u03b9\u03ba\u03ac \u03ba\u03b9\u03bd\u03bf\u03cd\u03bc\u03b1\u03b9 \u03c3\u03c4\u03bf \u03af\u03b4\u03b9\u03bf \u03bc\u03ae\u03ba\u03bf\u03c2 \u03ba\u03cd\u03bc\u03b1\u03c4\u03bf\u03c2 \u03c0\u03c1\u03bf\u03b2\u03bb\u03b7\u03bc\u03b1\u03c4\u03b9\u03c3\u03bc\u03bf\u03cd \u03bc\u03b5 \u03c3\u03ad\u03bd\u03b1.\u0394\u03c5\u03c3\u03c4\u03c5\u03c7\u03ce\u03c2 ,\u03b1\u03c5\u03c4\u03ae \u03c4\u03b7\u03bd \u03b9\u03c3\u03c4\u03bf\u03c3\u03b5\u03bb\u03af\u03b4\u03b1 \u03b4\u03b5\u03bd \u03c4\u03b7\u03bd \u03c0\u03b1\u03c1\u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03bf\u03cd\u03bd \u03c0\u03bf\u03bb\u03bb\u03bf\u03af , \u03b1\u03c0\u03cc \u03b1\u03c5\u03c4\u03bf\u03cd\u03c2 \u03c0\u03bf\u03c5 \u03b8\u03b1 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03c3\u03b1\u03bd \u03bd\u03b1 \u03b1\u03bb\u03bb\u03ac\u03be\u03bf\u03c5\u03bd \u03ba\u03ac\u03c4\u03b9.\u0391\u03c0\u03cc \u03c4\u03b7\u03bd \u03ac\u03bb\u03bb\u03b7 \u03bc\u03b5\u03c1\u03b9\u03ac \u03bf\u03b9 \u03bc\u03b1\u03b8\u03b7\u03c4\u03ad\u03c2 \u03bc\u03b1\u03c2 [b]\u03b5\u03c0\u03b7\u03c1\u03c1\u03b5\u03ac\u03b6\u03bf\u03bd\u03c4\u03b1\u03b9 \u03b1\u03c1\u03bd\u03b7\u03c4\u03b9\u03ba\u03ac \u03b1\u03c0\u03bf \u03c3\u03c7\u03cc\u03bb\u03b9\u03b1 \u03c0\u03bf\u03c5 \u03b1\u03c6\u03bf\u03c1\u03bf\u03cd\u03bd \u03c4\u03b7\u03bd \u03b5\u03c4\u03b1\u03b9\u03c1\u03b5\u03af\u03b1 [/b], \u03b4\u03b9\u03cc\u03c4\u03b9 \u03b4\u03b5\u03bd \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03bd\u03b1 \u03ba\u03b1\u03c4\u03b1\u03bb\u03ac\u03b2\u03bf\u03c5\u03bd \u03c4\u03b7 \u03bb\u03bf\u03b3\u03b9\u03ba\u03ae(\u03c0\u03b1\u03c1\u03b1\u03bb\u03bf\u03b3\u03b9\u03c3\u03bc\u03cc) \u03c4\u03c9\u03bd \u03b4\u03b9\u03bf\u03b9\u03ba\u03bf\u03cd\u03bd\u03c4\u03c9\u03bd.\u039a\u03b1\u03b9 \u03b5\u03b3\u03ce \u03c4\u03b5\u03bb\u03b5\u03c5\u03c4\u03b1\u03af\u03b1 \u03ac\u03ba\u03bf\u03c5\u03c3\u03b1 \u03c3\u03c5\u03b6\u03b7\u03c4\u03ae\u03c3\u03b5\u03b9\u03c2 \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u0395\u039c\u0395 \u03ba\u03b1\u03b9 \u03c4\u03bf\u03c5\u03c2 \u03b4\u03b9\u03b1\u03b3\u03c9\u03bd\u03b9\u03c3\u03bc\u03bf\u03cd\u03c2 , \u03c0\u03bf\u03c5 \u03bc\u03b5 \u03c4\u03c1\u03cc\u03bc\u03b1\u03be\u03b1\u03bd. \u0394\u03b5 \u03b8\u03b1 \u03c3\u03c7\u03bf\u03bb\u03b9\u03ac\u03c3\u03c9 \u03cc\u03bc\u03c9\u03c2 \u03c4\u03af\u03c0\u03bf\u03c4\u03b1.\u039a\u03b1\u03bb\u03bb\u03af\u03c4\u03b5\u03c1\u03b1 \u03bd\u03b1 \u03bc\u03b5\u03af\u03bd\u03bf\u03c5\u03bc\u03b5 \u03bb\u03af\u03b3\u03bf \u03c1\u03bf\u03bc\u03b1\u03bd\u03c4\u03b9\u03ba\u03bf\u03af \u03ba\u03b1\u03b9 \u03c4\u03c1\u03bf\u03b2\u03b1\u03b4\u03bf\u03cd\u03c1\u03bf\u03b9 \u03c4\u03c9\u03bd \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ce\u03bd \u03ba\u03b1\u03b9 \u03bd\u03b1 \u03bc\u03b7 \u03c6\u03b8\u03b1\u03c1\u03bf\u03cd\u03bc\u03b5 \u03b1\u03c0\u03cc \u03c4\u03b7 \u03b3\u03b5\u03bd\u03b9\u03ba\u03cc\u03c4\u03b5\u03c1\u03b7 \u03b4\u03b9\u03b1\u03c0\u03bb\u03bf\u03ba\u03ae \u03ae \u03b4\u03b9\u03b1\u03c6\u03b8\u03bf\u03c1\u03ac \u03c4\u03b7\u03c2 \u03ba\u03bf\u03b9\u03bd\u03c9\u03bd\u03af\u03b1\u03c2 \u03bc\u03b1\u03c2.[b]\u03a3\u03c5\u03bc\u03c6\u03c9\u03bd\u03ce \u03bc\u03b1\u03b6\u03af \u03c3\u03bf\u03c5 [/b]\u03b1\u03bb\u03bb\u03ac \u03b5\u03c5\u03b8\u03c5\u03b3\u03c1\u03b1\u03bc\u03bc\u03af\u03b6\u03bf\u03bc\u03b1\u03b9 \u03bc\u03b5 \u03c4\u03bf \u03a3\u03b9\u03bb\u03bf\u03c5\u03b1\u03bd\u03cc.\u0391\u03c2 \u03b2\u03c1\u03bf\u03cd\u03bc\u03b5 \u03ac\u03bb\u03bb\u03bf \u03c7\u03ce\u03c1\u03bf \u03bd\u03b1 \u03c4\u03b1 \u03bb\u03ad\u03bc\u03b5, \u03c4\u03bf \u03ad\u03c7\u03c9 \u03ba\u03b1\u03b9 \u03b5\u03b3\u03ce \u03b1\u03bd\u03ac\u03b3\u03ba\u03b7.\u03a3\u03c4\u03bf forum \u03b1\u03c2 \u03c0\u03c1\u03bf\u03c3\u03c0\u03bf\u03b9\u03b7\u03b8\u03bf\u03cd\u03bc\u03b5 \u03cc\u03c4\u03b9 \u03b4\u03b5\u03bd \u03be\u03ad\u03c1\u03bf\u03c5\u03bc\u03b5 \u03c4\u03af\u03c0\u03bf\u03c4\u03b1 \u03ba\u03b1\u03b9 \u03b1\u03c2 \u03b1\u03c0\u03bf\u03bb\u03b1\u03cd\u03c3\u03bf\u03c5\u03bc\u03b5 \u03c4\u03b1 \u03c0\u03c1\u03bf\u03b2\u03bb\u03ae\u03bc\u03b1\u03c4\u03b1 \u03ba\u03b1\u03b9 \u03c4\u03b7 \u03c6\u03b9\u03bb\u03af\u03b1.\r\n \u039a\u03b1\u03b9 [color=red]\u03c3\u03b5 \u03ac\u03bb\u03bb\u03bf \u03b2\u03ae\u03bc\u03b1 \u03b1\u03c2 \u03b5\u03af\u03bc\u03b1\u03c3\u03c4\u03b5 \u03c0\u03cd\u03c1\u03b9\u03bd\u03bf\u03b9[/color] \u03b5\u03bd\u03b1\u03bd\u03c4\u03af\u03bf\u03bd \u03b1\u03c5\u03c4\u03ce\u03bd \u03c0\u03bf\u03c5 \u03b4\u03b5\u03bd \u03ba\u03ac\u03bd\u03bf\u03c5\u03bd \u03c4\u03b1 \u03b1\u03c0\u03b1\u03c1\u03b1\u03af\u03c4\u03b7\u03c4\u03b1 \u03b3\u03b9\u03b1 \u03c4\u03b7\u03bd \u03c0\u03b1\u03b9\u03b4\u03b5\u03af\u03b1 , \u03c4\u03b7\u03bd \u03bc\u03b1\u03b8\u03b7\u03bc\u03b1\u03c4\u03b9\u03ba\u03ae \u03b5\u03c4\u03b1\u03b9\u03c1\u03b5\u03af\u03b1 \u03ba\u03b1\u03b9 \u03b1\u03c5\u03c4\u03ac \u03c4\u03b1 \u03c0\u03b1\u03b9\u03b4\u03b9\u03ac \u03c0\u03bf\u03c5 \u03b1\u03be\u03af\u03b6\u03bf\u03c5\u03bd \u03ba\u03b1\u03b9 \u03bc\u03c0\u03bf\u03c1\u03bf\u03cd\u03bd \u03c0\u03bf\u03bb\u03bb\u03ac.\r\n\u039d\u03b1 \u03b5\u03af\u03c3\u03b1\u03b9 \u03c3\u03af\u03b3\u03bf\u03c5\u03c1\u03bf\u03c2 \u03cc\u03c4\u03b9 \u03c5\u03c0\u03bf\u03c3\u03c4\u03b7\u03c1\u03af\u03b6\u03c9 \u03c4\u03b9\u03c2 \u03b8\u03ad\u03c3\u03b5\u03b9\u03c2 \u03c3\u03bf\u03c5 \u03ba\u03b1\u03b9 \u03b4\u03b5 \u03b4\u03b9\u03b1\u03c6\u03c9\u03bd\u03ce \u03bc\u03b1\u03b6\u03af \u03c3\u03bf\u03c5.\u03a3\u03c4\u03b5\u03af\u03bb\u03b5 \u03bc\u03bf\u03c5 \u03ad\u03bd\u03b1 PM \u03bd\u03b1 \u03c4\u03b1 \u03c0\u03bf\u03cd\u03bc\u03b5. \u0398\u03b1 \u03c7\u03b1\u03c1\u03ce.\r\n \u039c\u03c0\u03ac\u03bc\u03c0\u03b7\u03c2"
}
{
"Tag": [
"pigeonhole principle",
"number theory unsolved",
"number theory"
],
"Problem": "Prove that among $81$ natural numbers whose prime divisors are in the set $\\{2, 3, 5\\}$ there exist four numbers whose product is the fourth power of an integer.",
"Solution_1": "It's very similar to an old IMO problem.\r\n\r\nEach integer $ 2^a\\cdot 3^b\\cdot 5^c$ corresponds with a triple $ (a,b,c)$. There are 8 possible parity combinations, so there are two triples with the same parity combinations. So the product of the integers must be a perfect square, call this $ a_1^2$. From the other 79 integers, there exist two whose product is a perfect square, call this $ a_2^2$. Continuing this, we will get more than 8 perfect squares $ a_1^2,a_2^2,\\ldots,a_9^2$. With the same method, there exist two of them whose product is a perfect fourth power. Q.E.D.",
"Solution_2": "Each of these numbers can be represented in form $ 2^{\\alpha _i} \\cdot 3^{\\beta _i} \\cdot 5^{\\gamma _i}$. Now we will study all $ \\alpha _i, \\beta _i, \\gamma _i$ moduli $ 2$. So each of them can give one of the $ 2$ different rests, there can be $ 2 \\cdot 2 \\cdot 2 \\equal{} 8$ different \"oddities\" of these powers. According to pigeonhole principe, there will be such two numbers that have the same their power \"oddities\", because $ \\frac {81}{8} > 1$. Now we multiply them and denote the obtained number as $ a_1$ and remove it. And we can continue to obtain such $ a_i$ until the number of the rest numbers will become $ 7$. By that time we will have obtained $ \\frac {81 \\minus{} 9}{2} \\equal{} 36$ such pairs.\r\nNow we observe numbers $ \\sqrt {a_i}$ - each of them is a natural number because $ a_i$ is a perfect square (. And again we observe their power \"oddities\" moduli $ 2$. Using the pigeonhole principe again, we obtain that there are two numbers $ \\sqrt {a_m}$ and $ \\sqrt {a_n}$ such that their \"oddities\" moduli $ 2$ will be equal, because $ \\frac {36}{8} > 1$. Now we multiply $ a_m$ and $ a_n$ and we get the fourth power, because their are both perfect squares and their \"oddities\" moduli $ 2$ are the same, $ q.e.d.$",
"Solution_3": "to end with, there are $37$ possible squares with two elements, $15$ possible fourth powers with $4$ elements and $4$ different eighth powers with $8$ elements if we count exactly (at least).",
"Solution_4": "IMO 1985 P4\nhttp://www.artofproblemsolving.com/Forum/viewtopic.php?p=366595&sid=5cb6ec4ca43e6f9bd2d6c2d3eb2b87d8#p366595\n\n\nThe bound I feel is very weak as 25 (in place of 81) works"
}
{
"Tag": [
"linear algebra",
"matrix",
"geometry",
"geometric transformation",
"calculus",
"integration",
"vector"
],
"Problem": "I want to draw up a list of conditions for solution of simultaneous linear equations\r\n$Ax=b$ where $A\\in{\\mathbb R}^{m\\times n},~x\\in{\\mathbb R}^{n},~b\\in{\\mathbb R}^{m}$\r\nin all cases $mn$ The case $m=n$ is well knwon.\r\n\r\nIf $mn$ as well?",
"Solution_1": "Think of the columns of $A$ is vectors. A solution exists precisely when $b$ is in the linear span of the columns of $A$, which in turn is equivalent to ${\\rm rank }[A]={\\rm rank}[A,b]$ (because rank = dimension of the column space).\r\nThe uniqueness of solution means that $b$ can be represented as a linear combination of columns in only one way. This requires an additional assumption that the columns of $A$ are linearly independent, i.e., ${\\rm rank }[A]=n$.",
"Solution_2": "Of course the case $m=n$ is simple and well known I guess for $m=n.\r\n\r\nAs a sidenote, you'll notice that if k columns of rref (A) do not contain pivots then the solution space is an affine translation with k free variables.",
"Solution_4": "I am surprised no one answered this question citing a famous theorem which concerns itself with exactly this question: the Fredholm Alternative theorem .\r\n\r\nI just checked Wikipedia and it looks like they develop it for compact integral operators which is all valid and sufficiently abstract, but there is a version of the theorem for finite dimensional vector spaces and it goes like this:\r\n\r\nThe equation $Ax=b$ has a solution if and only if the inner product $ =0$ for every vector v satisfying $A^{*}v=0$. \r\n\r\nIn words, there is a solution iff the right-hand side (b) is orthogonal to the null space of the adjoint of the operator (A).\r\n\r\nFor uniqueness, we have the solution to $Ax=b$ is unique iff the solution of $Ax=0$ is $x=0$.\r\n\r\nI hope this is helpful."
}
{
"Tag": [
"algebra proposed",
"algebra"
],
"Problem": "Find the number of the solutions of the system or prove, that it has infinite number of solutions.\r\n\r\nx = y(mod z)\r\nx - z = y",
"Solution_1": "Just take $ x\\equal{}2z$ and $ y\\equal{}z$, then $ (x,y,z)\\equal{}(2t,t,t)$ is solution for any integer t so system has infinitely many solutions.",
"Solution_2": "I think should be find out all solution of this equation.\r\nBecause $ x\\equiv y (\\mod z)$ then\r\n$ x\\equal{}y\\plus{}kz$ where $ k\\in Z$\r\nSo \r\n$ (k\\minus{}1)z\\plus{}y\\equal{}y$ so $ k\\equal{}1$\r\nSolution of this equation are $ (x,y,z)\\equal{}(a\\plus{}b,b,a)$",
"Solution_3": "The second equation implies the first, so if you want all solutions just take all points on the plane defined by the second equation."
}
{
"Tag": [
"algebra proposed",
"algebra"
],
"Problem": "You have a machine which, when you give it two real numbers a and b (with $b\\neq 0$), gives you out the number $\\displaystyle 1-\\frac{a}{b}$. Prove that using this machine, you can add, subtract, multiply and divide any two real numbers.\r\n\r\n Darij",
"Solution_1": "This is a very nice problem. I don't know why I like it so much :).\r\n\r\nLet $\\displaystyle (a,b)=1-\\frac ab$. Then $(a,1)=1-a$, so $((a,b),1)=\\frac ab$, so we have the division.\r\n\r\nWe also have $((a,b),a)=1-(\\frac 1a+\\frac 1b)$, so $(((a,b),a),1)=\\frac 1a+\\frac 1b$, and from the first paragraph we find that we can obtain $\\frac 1a,\\frac 1b$, so $(((\\frac 1a,\\frac 1b),\\frac 1a),1)=a+b$.\r\n\r\nWe can find $1-(\\frac 1a+\\frac 1b)$, so we can also find $\\frac 1-\\frac 1a-1=-\\frac 1a$, so we can also obtain $-a$, which gives $b-a$ after addition with $b$.\r\n\r\nMultiplication is, of course, obtained as the quotient of $a$ and $\\frac 1b$.",
"Solution_2": "[quote=\"grobber\"]We also have $((a,b),a)=1-(\\frac 1a+\\frac 1b)$,[/quote]\r\n\r\nHmm, at this step I get $((a,b),a)=1-(\\frac 1a-\\frac 1b)$, so I think you have a little mistake in your calculation. But actually the solution can be corrected easily.\r\n\r\nHere is my solution:\r\n\r\nI will also denote $\\displaystyle \\left(a,b\\right) = 1-\\frac{a}{b}$. Then, I can construct $0 = \\left(a,a\\right)$ (provided that $a \\neq 0$; else we can take b instead of a, and if both a and b are 0, then we can actually add, subtract and multiply without the help of our machine :D ). Then, $1 = \\left(0,a\\right)$ (or $1 = \\left(0,b\\right)$). Hence, we can force our machine to give out the numbers 0 and 1. Therefore, e can divide, since $\\displaystyle \\frac{a}{b} = \\left(\\left(a,b\\right),1\\right)$. Hence, since we have the number 1, we can find for every x the number $\\displaystyle \\frac{1}{x}$. Therefore, we can also multiply: $\\displaystyle a \\cdot b = \\frac{a}{\\left(\\frac{1}{b}\\right)}$. Also, we can subtract: $a-b = \\left(b,a\\right) \\cdot a$. Since we have the number 0, we thus can also find for every x the number $-x = 0-x$. Therefore, we can also add: $a+b = a-\\left(-b\\right)$. And thus the problem is solved.\r\n\r\n Darij",
"Solution_3": "You're right, sorry for that."
}
{
"Tag": [
"combinatorics proposed",
"combinatorics"
],
"Problem": "I have a problem of my own. I would like all of you discuss it and give your result. Thank you very much!\r\n\r\nLet ABC be a triangle. On the side AB, take out m points; on the side BC, take out n points; on the side CA, take out k points. These points are distinct and coincisive with the vertex of the triangle ABC. How many triangles of which vertex are those arbitrary points including A, B and C are there?",
"Solution_1": "[hide]${m+n+k+3\\choose 3}-{m+2\\choose 3}-{n+2\\choose 3}-{k+2\\choose 3}$[/hide]",
"Solution_2": "Dear Bojan Basic,\r\nThank you very much for your answer. Your result is the same as mine."
}
{
"Tag": [
"limit",
"function",
"algebra unsolved",
"algebra"
],
"Problem": "$\\lim \\frac{tan x}{x}$ \r\nwhen $x-----\\to \\ infinty$",
"Solution_1": "I don't think there exists a limit, since the function is unbounded even when x is very large."
}
{
"Tag": [
"inequalities"
],
"Problem": "Prove that \r\n$\\sqrt{\\left|\\vec{a}+\\vec{b}\\right|\\left|\\vec{v}+\\vec{w}\\right|}\\geq \\sqrt{\\left|\\vec{a}\\right|\\left|\\vec{v}\\right|}+\\sqrt{\\left|\\vec{b}\\right|\\left|\\vec{w}\\right|}$",
"Solution_1": "What if $\\vec b=-\\vec a\\neq\\vec 0$ :?:",
"Solution_2": "If we let $\\vec{a}\\neq-\\vec{b}$ it holds ?\r\nIf not could anyone find a form of this inequality to hold .(maybe reserved).",
"Solution_3": "It doesn't seem it should hold. You can always choose $\\vec a,\\vec b$ and/or $\\vec v,\\vec w$ so that the $\\measuredangle(\\vec a,\\vec b)$ and/or $\\measuredangle(\\vec v,\\vec w)$ are arbitrarily close to $\\pi$. Then it's obvious that $|\\vec a+\\vec b|$ and/or $|\\vec v+\\vec w|$ are arbitrarily close to $|a-b|,|v-w|$ respectively and $\\sqrt{|a-b|\\,|v-w|}\\geqslant\\sqrt{av}+\\sqrt{bw}$ doesn't hold for positive $a,b,v,w$, which is easily shown.",
"Solution_4": "Thank you .Reserved signed holds ? Or could we find a similar one to hold ?"
}
{
"Tag": [
"trigonometry",
"geometry",
"perpendicular bisector",
"similar triangles"
],
"Problem": "If the altitude AE and the median BF intersect at the point H of the triangle ABC, such that the perpendicular bisector of BC intersects BF and BC at the points O and D, then prove that OD = AH/2.",
"Solution_1": "Let FX be the perpendicular from F to BC.\r\n\r\nWe have FX= H/2=h (where AE=H )\r\nBX= a-b/2 cosC...a,b,c and A,B,C carry the usual meaning of sides and angles\r\n\r\nIn similar triangles BOD and BFX we have OD/FX=BD/BX....OD=ah/(2a-bcosC)\r\n\r\nIn similar triangles BHE and BFX we have HE/FX=BE/BX......HE=2chcosB/(2a-bcosC)...\r\nAH= 2h-HE=2ah/(2a-bcosC)....(we have a=b cosC+c cosB).\r\n\r\nHence the result.",
"Solution_2": "Here is my proof with no trigonometry (picture included!):\r\n\r\nConsider the midpoint G of AB. Draw the auxiliary line DG. Since DG connects the midpoints of sides AB and BC, ABC and GBD are similar triangles. Now drop a perpendicular from G such that GP is parallel to OD and AE, where P is the midpoint of BH because of similar triangles. GBP and ABH are similar so GP = AH/2. All we need to prove now is that GP = OD. Notice that 99.5) = Prob(\\frac{X-96}{7.065}> \\frac{3.5}{7.065}) \\approx Prob(Z > 0.50) = ... \\]\r\n(Look it up in a table, $Z$ is standard normal).\r\n\r\nHope this helps.",
"Solution_3": "It sure does. Thanks a lot!",
"Solution_4": "Hey Guys, thank you for helping a poor fellow.\r\n\r\nI have another problem, which buggs me enormously. It goes like this:\r\n\r\n10 persons work at a company. 6 of them use a computer. The boss wants to pick out 4 of them, randomly, and offer them a health care examination. \r\n\r\nWhat is the probability that at most half of the persons chosen work with a computer."
}
{
"Tag": [
"ARML",
"search",
"AMC 8",
"AMC"
],
"Problem": "i thought we weren't suppose to talk about this until after the 25th....\r\ni feel stupid i think i got a 21 22 or 23 but my mistakes could've been avoided. what's the arml?",
"Solution_1": "it doesnt matter if you say what you got or how you did, just as long as you dont discuss the problems.",
"Solution_2": "max: im not sure anyone can tell you what they're looking for, since different places have different ways of picking their teams. The best thing would just be to move on and just do better on the next contests that you do. They wouldnt overlook you just because you did not do that well on one exam, so long as you show that you can do well on others.",
"Solution_3": "[quote=\"cats...\"]The best thing would just be to move on and just do better on the next contests that you do. They wouldnt overlook you just because you did not do that well on one exam, so long as you show that you can do well on others.[/quote]\r\n\r\nThat makes sense. For programs that emphasize writing proofs, a test like the AMC 8 isn't going to tell the whole story. For programs of multiple-choice problem-solving, the level of the AMC 8 is different from the level of other programs, so it isn't always a perfect predictor of performance on tests of secondary-level math. \r\n\r\nJust move on and try something else is generally a good philosophy of life. Don't give up too soon--find another way to prove yourself. \r\n\r\nI don't know specifically about ARML contests, but I hope to have my son participating in them in a few years.",
"Solution_4": "I don't think any ARML team uses the AMC 8 to pick their team members . . . but where do you live? What team are you competing for?\r\n\r\nFor the person who asked about ARML -- it's the American Regions Math League, a nationwide math contest for (mostly) high school students. http://www.arml.com. If you do a search of this site, you can find much more information about it.",
"Solution_5": "there is no oklahoma team. sigh",
"Solution_6": "Probably fairly competitive, that. You should check to see if they even look at AMC 8 scores for middle schoolers trying to make the team. Anyhow, don't be discouraged if you can't beat out all the high schoolers from the second-largest state in the country . . ."
}
{
"Tag": [
"number theory"
],
"Problem": "Hello, I have discovered that I have a lack of inginuity in Number Theory. Would you please recommend some books that will improve my problem solving skills involving Number Theory? Thanks in advance.\r\n\r\nMasoud Zargar",
"Solution_1": "This is hardly an open question.\r\n\r\nAnyways, try Andrews' Number Theory (published by Dover). I'm not a huge fan of it, but it has a good selection of problems AND solutions. The kicker is that it's like 15 bucks.",
"Solution_2": "I moved it to the right section.",
"Solution_3": "[quote=\"blahblahblah\"]This is hardly an open question.\n\nAnyways, try Andrews' Number Theory (published by Dover). I'm not a huge fan of it, but it has a good selection of problems AND solutions. The kicker is that it's like 15 bucks.[/quote]\r\nOh, okay. Thanks. I'll try ans see if Chapters has it. :)\r\n\r\nMasoud Zargar",
"Solution_4": "Well, \r\n\r\n a) D.Burton , Elementary Number theory. This is a text book\r\n\r\n b) Adler and Coury. Many solved problems. One of the best.(expensive !)\r\n \r\n [u]Babis[/u]",
"Solution_5": "A friend of mine recently referred me to a great number theory book that has a really good mix of substance and focused problem solving at an accessible level. Elementary Number Theory with Applications by Thomas Koshy. According to my friend, all math \"majors\" at Duke's TIP program take a course based on this book.",
"Solution_6": "There is book called the \"Elementary Number Theory and Its Applications(Fourth Edition)\". It was written by Kenneth H.Rosen. How do you guys think about this book?",
"Solution_7": "I have \"Classical introduction to Modern Number Theory\" by Rosen which is an expanded and modified version of what you wrote (if I remember correctly preface :D ) and I like it very much. Every chapter has some exercises plus I love bibliography given at the end of every chapter and the way authors direct the reader to other texts.",
"Solution_8": "I haven't read too much of it yet, but I like \"The Higher Arithmetic\" by Davenport so far.",
"Solution_9": "Has anyone read a number theory book by Sierpinski?\r\nIf so, was the book good?",
"Solution_10": "I like Sierpinski very much, in fact :)",
"Solution_11": "I've taught a class using Rosen, and for teaching the class, I would choose Rosen again. We expect our students to take number theory before they take abstract algebra (groups-rings-fields), and this course is required for all of those going into high school teaching. For that audience, It may be a fairly difficult book for that audience (and we only try to do a little over half of the book in a semester), but it does work as a textbook.\r\n\r\nOne of the books stergiu mentioned was Burton. That's been used around here as well, but Rosen has more life to it, more applications, and better exercises.",
"Solution_12": "Wait, when you guys say that the book has excersizes, are you differentiating between problems and excersizes like Zeitz?",
"Solution_13": "Well, this is number theory so there is no big difference between an exercise and a problem :D",
"Solution_14": "[quote=\"Arne\"]I like Sierpinski very much, in fact :)[/quote]\r\n\r\nIf you have a Ti-84, 84+, 84 se, or 84+ se, you can make a program that creates the Sierpinski triangle. If you don't have the code, I'll give it to you.",
"Solution_15": "This is not about the Sierpinski triangle :p but rather about the book Sierpinski wrote. He was a great number theorist."
}
{
"Tag": [
"Support",
"function",
"probability",
"geometry",
"linear algebra",
"matrix",
"articles"
],
"Problem": "Someone in the thread about Intelligent Design (ID) was complaining that ID was not supported by evidence. I'd like to create this thread to discuss possible evidence for both theories.\r\n\r\nPlease give me some evidence for evolution that's conclusive enough to pass the tests that people are using on ID.",
"Solution_1": "I think the one most commonly cited example in biology textbooks for easily observable evolution (not a long-term fossil record sort of thing, which is what we're primarily interested in, but over a time period that we can actually observe) is that one about butterflies or moths or something in Britain (they may have been some other species, it doesn't matter). The general story is that they originally used to be white because they hid on the light-colored trees; as the Industrial Revolution progressed, the trees became black with smoke and soot, and accordingly, the animals that were born had progressively darker tints until they were pretty much back. Then, environmental regulations helped the trees become light in color again, and indeed, the species returned to its previous state.",
"Solution_2": "We deal with evolution every year when we make up flu vaccines.\r\n\r\nThe flu virus mutates very quickly. There are many strains as environmental conditions (including immune system conditions and vaccinations) force dramatic shifts in evolutionary stability of various genetic patterns.\r\n\r\nIn fact, this happens in many viruses.\r\n\r\n\r\nAnother piece of evidence is the precision of the model of genetic evolution when we overlay DNA patterns. About a decade ago I worked at the Institute for Biomedical Computing where genomes from various species were compared for deviations using a program called BLAST, which is ultimately a brute-force base pair matching program. The differences between the DNA of various species matches up beautifully with how far apart we would expect those species to be evolutionarily. In other words, if we assume that no matter what type of genetic selection occurs, that DNA mutates (mutations/insertions/deletions) at a particular rate based on its complexity, then, for instance, humans and chimps are expected to be genetically similar to a certain degree. And according to BLAST...we are indeed similar to that degree!\r\n\r\nThose who refuse to incorporate genetic evolution into their worldview are either not aware of the tremendous evidence or are intentionally turning a blind eye to science. It's not just a matter of evidence. We observe on a small scale something that when extrapolated to a large scale equals evolution. We don't deny relativity because we can't observe it from moment to moment -- so why deny evolution?",
"Solution_3": "Yeah, but that stuff only proves microevolution and adaptation. It does not prove that is how life formed.",
"Solution_4": "[quote=\"ToBe\"]Yeah, but that stuff only proves microevolution and adaptation. It does not prove that is how life formed.[/quote]\r\n\r\nI'm not just talking about microevolution when I point at that point accepted mutation (PAM) time models of evolution match up perfectly with mapped genomes and it has absolutely zero to do with adaptation.\r\n\r\nAnd genetic evolution is not a theory of how life formed, but rather, how it proceeds.\r\n\r\nQ: Are humans intelligent designers?",
"Solution_5": "I'd like to refer to the Creationism/Intelligent Design thread. The only evidence (correct me if I'm wrong) presented for intelligent design was the supposed complexity of life forms today. Life can be corresponded with computer hardware/software today. Computers today combined with their wonderous software are so complex, they could not have been developed without devine intervention. Does this seem reasonable?\r\n\r\n Another common complaint from the Creationism/Intelligent Design thread says that Intelligent Design is not a religious theory, if it can be considered a theory at all. Look around you. Do you see atheist ID supporters? Now can you claim ID to be a non-religious \"theory\"?",
"Solution_6": "I'd also like to point out, if some people dont know it, that Darwin himself was a believer in a creator, and several times in Origin of Species he says \"nature does not make transitions\" - nature only makes jumps. (and as a believer, the proof is that at the end he mentions something about all being possible from a breath of a creator.) \r\n\r\nI'd also like to ask a question: why is this even an argument? If one looks closely at Genesis, it mentiones NOWHERE anything about special individual creation for each animal. The only creature that it says in specific is man. For general evolution, aside from man (i'll get back to y'all later on man) there is no real dispute, no reason for it. Everyone just 'assumes' that all creatures according to Genesis were created specially. That is not true. ~ in truth, there doesnt even have to be a dispute about religion vs. science, because when both percieved from an unbiased view, and from the view of someone that actually has researched the material on both sides, they are actually interdependent~ - I'd reccomend the book 'The Science of God' by Gerald L. Schroeder.\r\n\r\n\r\n- yes i know this is on creationism, not quite intelligent design, but all the same.. *i realized this afterwards. maybe i should revive old threads.*",
"Solution_7": "[quote=\"MCrawford\"]I'm not just talking about microevolution when I point at that point accepted mutation (PAM) time models of evolution match up perfectly with mapped genomes and it has absolutely zero to do with adaptation.[/quote]\r\n\r\nI'm sorry, but I didn't quite follow that sentence. Could you please rephrase it? Thanks.",
"Solution_8": "[quote=\"mathnerd314\"]Someone in the thread about Intelligent Design (ID) was complaining that ID was not supported by evidence. [/quote]\nWho said that, exactly (quote the original text, please)? ID is consistent with anything and its opposite, so there is no possibility of evidence for or against it. \n\n[quote] I'd like to create this thread to discuss possible evidence for both theories. [/quote]\nThere are very few legitimate reasons to discuss both together. It appears more as a rhetorical tactic than shedding light on either theory. ID doesn't rise to the level of what is usually called a \"theory\", by the way; to do that it would have to make at least one meaningful statement about the natural world.\n\n[quote] Please give me some evidence for evolution that's conclusive enough to pass the tests that people are using on ID.[/quote]\r\n\r\nOne test that ID fails and evolution passes is \"does it say anything meaningful whatsoever\". \r\nAs it happens, the assertions of evolution are generally borne out by observation, and this is why it ends up being taught in the classroom; but all that is subsequent to the first hurdle, which ID doesn't pass, of actually saying something. That the quality of fit between theory and observation for evolution can actually be debated is a huge, knockout-level point in its favor as compared to ID. Even literal bible-thumping creationism does better than meaningless ID in that respect, although it ends up being discarded in light of observation. \r\n \r\nAnyway, the thread for evolution is the following:\r\nhttp://www.artofproblemsolving.com/Forum/viewtopic.php?t=23486",
"Solution_9": "Has anyone heard of the Burgess Shales? Some smithsonian guy went on a mission to canada and found fossils that proved that gradual evolution did not exist, as he found dif creatures jumping from small creatures to fully formed, fully capable ones. He didnt believe in the stuff, so he hid it in his lab for 80 years till it was discovered, made news like 10 years ago.",
"Solution_10": "er, yes. I think currently accepted theory does in fact reveal that Darwin's idea of very, very gradual mutations was in fact not exactly correct and that changes are slightly more abrupt (there's a term for this, but I can't remember it ... took AP Bio 2 years ago now...)",
"Solution_11": "punctuated equilibrium.",
"Solution_12": "What's the [b]evidence[/b] for punctuated equilibrium, please? As I understand it, there's more evidence for Darwinian evolution than punctuated equilibrium.",
"Solution_13": "cambrian explosion,",
"Solution_14": "[quote=\"mathnerd314\"]What's the evidence for punctuated equilibrium, please? As I understand it, there's more evidence for Darwinian evolution than punctuated equilibrium.[/quote]\r\nThese are not theories on the same level; they are consistent with each other, and seek to explain different facets of the history of life on earth.\r\n\r\nDarwinian evolution: Natural selection, forcing adaptation to local conditions, is the primary cause of change in species over time. Also, history is very long, and even this inefficient method produces great change over such timescales.\r\nThe power of selection as an agent of change is undeniable- see what we have done to our domestic animals and crops. This human selection is positive rather than negative, and therefore works much faster\r\n\r\nPunctuated equilibrium: Most change, especially speciation, happens relatively rapidly in small isolated populations. Successful and widespread species like those we see in the fossil record are extremely unlikely to change significantly except in abundance, even over millions of years.\r\nThe historical evidence here supports the second sentence- the first is an inference from the second and from statistics. This theory does not really deal in methods- it accepts natural selection.\r\n\r\nYou might have been looking for gradualism as a foil, and that is very hard to support with historical evidence, because the theory denies the \"equilibrium\" part of punctuated equilibrium. How would you explain the great profusion of species unique to geologically young small islands?\r\n\r\nSome modern notes, which have come to light with DNA analysis. In closely related species, the genes which matter are much more similar than the sections of DNA that don't code for anything. This suggests that natural selection normally acts as a barrier to change by striking down variations that don't work. On the other hand, random change is predictable, moving at an easily calculated rate which depends inversely on population size- we can even estimate how long two modern species have been distinct this way.\r\n\r\nAs for the Burgess Shale- find \"Wonderful Life\" by Stephen J. Gould. Many of the details are now obsolete, but it's a great book. A central theme, which the title brings up, is the unpredictability of life's history, at every level. Even if the processes are deterministic, their interactions are so complex that nothing is inevitable from a broad picture.",
"Solution_15": "Hey guys- in case you guys are still interested -\r\n\r\nhttp://www.pennlive.com/news/patriotnews/index.ssf?/base/news/1130667685324780.xml&coll=1\r\n\r\n[quote]The Discovery Institute, one of the nation's leading organized proponents of intelligent design, said it warned the Dover Area School District not to institute a policy on the concept because of the risk it would be found \"somehow unconstitutional.\"[/quote]\n\nhttp://aclupa.blogspot.com/\nClosing statements of the K v DASD trial.\n[quote]The defendants' attorney Patrick Gillen noted that the day was the fortieth day of the trial, and that night was the fortieth night. He asked if the judge had done that on purpose.\n\nThe judge said with a smile that it was \"not by design.\" The entire courtroom broke out in laughter and then applause.\n\nAnd with that, the biggest trial on evolution in two decades came to a close.[/quote]",
"Solution_16": "[quote=\"ToBe\"]I noticed that too. As I said, I didnt think that all of those questions were even valid (some were just...easy to answer too).[/quote]\r\n\r\nThen tell us which ones you think are valid, so we don't waste our time answering all the questions on the site?",
"Solution_17": "zeus_three is absolutely right. We talked about this in the first day of science this year, and why we do not teach ID in [b] public [/b] school. ID is pseudoscience, whether true or not. For something to be considered science, it must be empirically demonstratable. This is why I do not believe theories of multiple universes can be currently considered science either. There are no tests we can perform to prove ID correct.",
"Solution_18": "stop questioning questions that don't need answers",
"Solution_19": "[quote=\"zeus_three\"]\nReligious teachers [i]do not teach quantum theory [/i]and [u]special relativity [/u]in [b]Religious studies lessons[/b], so [i]why are they even contemplating teaching ID [/i]in [b]science lessons[/b]?! [b]The absurdity confounds me, and the logic is non-existent.[/b]\n[/quote]\r\n\r\n :D My religion teacher taught us a bit about special relativity, because of Einstein and his connection with the Big Bang. Its actually in the curriculum. edit: its in the provincial curriculum that is.\r\n\r\nIn other words, don't say things you obviously don't know anything about. If you're not in a religion class, how do you know? Or if you actually are, then you sure aren't learning anything from it.",
"Solution_20": "[quote=\"Sunny\"] My religion teacher taught us a bit about special relativity .... Its actually in the curriculum .... In other words, don't say things you obviously don't know anything about. [/quote]\r\n\r\nIt is possible that somewhere in the world in a class labelled Religion, people are learning how to solve physics problems. Nevertheless, the other poster's statement is clear enough, and clearly correct, that \"special relativity and quantum mechanics are not taught in religion classes\". If your religion class has problem sets on relativistic velocity addition or inflationary cosmology, as opposed to the historical/philosophical/sociological matters supposedly related to modern physics, that would be unusual, but it would not justify treating his self-evident observation as some sort of absurd hallucination. \r\n\r\nThere are different selection criteria for the curricula in science and non-science classes, and (see posting 219 above) ID doesn't qualify for the former.",
"Solution_21": "I have something to say about an argument I saw a while back, which said something like, it is highly unlikely that you will flip ten coins in a row and they will all come up heads. Is that any more likely than flipping ten coins in a row and them coming up heads, tails, heads, heads, tails, heads, tails, tails, tails, heads?\r\n\r\nHoly crap, I flipped ten coins and they came up tails, heads, heads, tails, heads, tails, heads, tails, tails, tails! That was a 1/1024 chance! The coin must be rigged!",
"Solution_22": "[quote=\"ToBe\"]Um, why is [something both irreducible and complex] the expected result [of evolution]? Things with no goal other than survival do not have to strive for optimization, only what is required.[/quote]\r\n\r\nThat's right, which is why (as I said several times before) evolution ruthlessly optimizes systems that are critical to survival and reproduction, and does not necessarily optimize other systems. The optimized systems will tend to be \"irreducible\", and most evolved systems of any kind will also be \"complex\", so as pointed out many times but answered zero times in this thread, \"both irreducible and complex\" is a signature of evolved systems. It is brain-dead for ID to use a signature of evolution as attempted proof that evolution did not occur.",
"Solution_23": "I think in general the tension in this debate arrises from a matter on which compromise must occur, but few are willing to accept.\r\n\r\nWhat i am referring to is the existence of probabalistic systems. It is preferrable that science would always be able to say with certainty, this happened in so many years, or that is impossible. But when probabilities are dealt with, certainty in these issues are impossible. However, this does not mean that probabilities are never of any use! Although i am no expert, i think that in physics the uncertainty principle makes schrodinger's wave equation the best possible location of an election, but does that make quantum mechanics useless?\r\n\r\nTo return to the earlier references in this thread dealing with the likelihood of royal straight flushes... It is [i]always[/i] possible that a royal straight flush is dealt fairly, and hence it is always possible that 20000 royal straight flushes, are dealt fairly, and again that 2^20000 royal straight flushes are dealt fairly. But does this mean that if 2^20000 royal straight flushes occurred, it might not be necessary to question the validity of the occurrence?\r\n\r\nMy point is this. Probabilities cannot be viewed as 100% distinct from scientific truths. I think any reasonable person can agree, that if the evolution from the earliest form of life to the human being were supposed to occur in a period of 100 years, that evolution would be impossible.\r\n\r\nI think that it would be perfectly scientific to analyze the probability of the occurrence of the evolutionary process, (for i think we can agree that natural selection is not of itself enough to produce the changes that exist, but that mutations subject to probability must play a part), and construct an expected value for length of the process thereby. Should this deviate significantly from the known length of time during which the process occurred, evolution would be impossible.\r\n\r\nI have heard several ID theorests argue this point, specifically William A. Dembski, who contrary to the generalizations of ID theorests as stupid, possesses doctorates in both mathematics and philosophy, in his book [i]The Design Inference: Eliminating Chance through Small Probabilities[/i]. Those who continue to argue the ID is completely unfounded and worthless,(not that everyone here is) would do well to take a look at this book and cease discussing what they do not know.",
"Solution_24": "[quote=\"lotrfan\"]... if 2^20000 royal straight flushes occurred, it might not be necessary to question the validity of the occurrence? [/quote]The question is how to credibly arrive at those astronomically low probabilities, not what to do given such a very small number. ID has to do the first before it can start the drumbeats about the second. So far the probability models in ID literature discredit ID, not evolution. \n\n[quote] Should this [expected length of time] deviate significantly from the known length of time during which the process occurred, evolution would be impossible. [/quote] All you have to do is to find a reasonably inclusive probability model where that can be demonstrated, then sit back and collect the Nobel Prizes.\n\n[quote] I have heard several ID theorests argue this point, specifically William A. Dembski, who contrary to the generalizations of ID theorests as stupid, possesses doctorates in both mathematics and philosophy, [/quote]\r\n\"By their fruits shall ye know them\". \r\n\r\nDembski's public displays of inanity are quoted and linked in this thread, so we don't need any academic pedigree to calibrate his competence. But it would be interesting to find out where his degrees are from and what his alma mater thinks of him.",
"Solution_25": "[quote]The question is how to credibly arrive at those astronomically low probabilities, not what to do given such a very small number.[/quote]\r\n\r\ni agree completely. The task of determining the expected length of time of the evolutionary process is monolithic, but it is still an unanswered question, a question which can either prove evolution beyond all shadow of doubt, or leave it in need of deep-seated reform. \r\n\r\nSuch a significant question should be of interest to scientists of whatever religious persuasion or of whatever camp, and i see no grounds for the consideration of an entity that attempts to explore this question as unscientific! Agreed, ID attempts to do more than just answer this question, and the manner in which it goes about this qeustion can discredit its scienticity(lol, a new word). My objection is more to the general consideration of all objections to evolution as unscientific. As fleeting_guest and i both agree, the expected length of the evolutionary process is as yet an unanswered question. Why is there no room for any other theories to sit alongside evolution in our textbooks? Even if no specific theory is given, people should know that there are still defensible grounds for rejecting evolution. If no one knows the expected length of the process, a critical consideration to the establishment of its feasability, no one can know that it is true.",
"Solution_26": "[quote=\"lotrfan\"]a significant question should be of interest to scientists of whatever religious persuasion or of whatever camp, [/quote]Mathematical models of evolution are a large area of research. The ordinary, non-ID approach is to just study the models and draw whatever conclusions are warranted. The ID approach is to assume certain conclusions and look for models that might support those. Since the ID starting point is almost certainly wrong, it is no surprise that the resulting models are irrelevant, nonsensical or un-quantifiable.\n\n[quote]grounds for the consideration of an entity that attempts to explore this question as unscientific! [/quote]ID acquires its abysmal reputation by ignoring 100 years of data in favor of half-baked wishful thinking, meaningless assertions and pseudoscientific posturing. Instead of fighting its battles with data and relevant mathematical models in the scientific literature, it fights using advertising, lawyers, politics, and in media under the sole control of ID supporters (ID journals and web sites, ID biology textbooks, etc). This, not the particular questions it raises, are what leads to the contempt for ID. The questions are problematic only insofar as they are too vague to be the basis of any research program. \n\n[quote] My objection is more to the general consideration of all objections to evolution as unscientific.[/quote]ID falsely equates criticism of its aggressive spam pseudo-objections, with dogmatic refusal of all objections. I doubt that there are any examples of \"consideration of all objections to evolution as unscientific\".\n\n[quote]As fleeting_guest and i both agree, the expected length of the evolutionary process is as yet an unanswered question. Why is there no room for any other theories to sit alongside evolution in our textbooks? [/quote]Please speak for yourself. I don't agree with your assertions in the least, and talk of \"expected length\" is not an argument against evolution. The reason there is no room for other theories is that there are, as of 2005, no viable non-evolutionary ideas to be taught, whereas evolution stands on a mountain of supporting evidence.",
"Solution_27": "[quote=\"zeb\"]I have something to say about an argument I saw a while back, which said something like, it is highly unlikely that you will flip ten coins in a row and they will all come up heads. Is that any more likely than flipping ten coins in a row and them coming up heads, tails, heads, heads, tails, heads, tails, tails, tails, heads?\n\nHoly crap, I flipped ten coins and they came up tails, heads, heads, tails, heads, tails, heads, tails, tails, tails! That was a 1/1024 chance! The coin must be rigged![/quote]\r\n\r\nI don't know what this has to do with intelligent design, but people usually look at THHTHTHTTT, and only see 4H and 6T. In other words, people don't usually pay attention to the order of the flips, just the number of heads and tails. Thus, 10H has a probability of 1/1024 and 5H/5T has a probability of 63/256, much greater.",
"Solution_28": "So... by [b]ignoring[/b] data, you can come to the conclusion of ID?",
"Solution_29": "[quote=\"dasherm\"][quote=\"zeb\"]I have something to say about an argument I saw a while back, which said something like, it is highly unlikely that you will flip ten coins in a row and they will all come up heads. Is that any more likely than flipping ten coins in a row and them coming up heads, tails, heads, heads, tails, heads, tails, tails, tails, heads?\n\nHoly crap, I flipped ten coins and they came up tails, heads, heads, tails, heads, tails, heads, tails, tails, tails! That was a 1/1024 chance! The coin must be rigged![/quote]\n\nI don't know what this has to do with intelligent design, but people usually look at THHTHTHTTT, and only see 4H and 6T. In other words, people don't usually pay attention to the order of the flips, just the number of heads and tails. Thus, 10H has a probability of 1/1024 and 5H/5T has a probability of 63/256, much greater.[/quote]\r\nIt has to do the following:\r\nmost ID-supporters here claimed that the chances (in evolution) where to small to be occured all together.\r\nAnd in evolution, the ordering of the mutations does play a big role, they cant occure in any order. So also THHTHTHTTT has just a chance of 1 to 1024.\r\n\r\nBut some people here also claim that we [u]must[/u] have developed this only way to get what we are now. But if other changes would have occured, we would just be different, and not, as some claim, not exist (possibly it would have needed some 100 millions years more or less for animals with 'human'-type intelligence to occure, but noone would notice)."
}
{
"Tag": [
"inequalities",
"search",
"email",
"inequalities unsolved"
],
"Problem": "1) Let a,b,c be positive real numbers such that $ abc\\equal{}1$. Prove that: \r\n$ \\frac{1}{(a\\plus{}b)b}\\plus{}\\frac{1}{(b\\plus{}c)c}\\plus{}\\frac{1}{(c\\plus{}a)a}\\geq\\frac{3}{2}$\r\n2) Prove that following ineq: $ (3a\\plus{}2b\\plus{}c)(\\frac{1}{a}\\plus{}\\frac{1}{b}\\plus{}\\frac{1}{c})\\leq\\frac{45}{2}$\r\nwhere $ a,b,c \\in [1;2]$\r\n3) Let x,y,z be non-nagetive real numbers. Find the min value of \r\n$ F\\equal{}(xy\\plus{}yz\\plus{}zx)(\\frac{1}{(x\\minus{}y)^2}\\plus{}\\frac{1}{(x\\minus{}y)^2}\\plus{}\\frac{1}{(x\\minus{}y)^2})$",
"Solution_1": "2. It's problem in Mathematics and Youth Magazine :mad: \r\nSee here http://www.mathlinks.ro/viewtopic.php?t=197349",
"Solution_2": "1) $ a\\equal{}2$ , $ b\\equal{}2$ , $ c\\equal{}\\frac{1}{4}$ :wink:",
"Solution_3": "[quote=\"mehdi cherif\"]1) $ a \\equal{} 2$ , $ b \\equal{} 2$ , $ c \\equal{} \\frac {1}{4}$ :wink:[/quote]\r\nIf $ a \\equal{} b \\equal{} 2,c \\equal{} \\frac {1}{4}$ \r\n$ \\rightarrow \\frac {1}{(a \\plus{} b)b} \\plus{} \\frac {1}{(b \\plus{} c)c} \\plus{} \\frac {1}{(c \\plus{} a)a}\\equal{} 2\\frac {1}{8}$",
"Solution_4": "[quote=\"mr.danh\"]\n3) Let x,y,z be non-nagetive real numbers. Find the min value of \n$ F \\equal{} (xy \\plus{} yz \\plus{} zx)(\\frac {1}{(x \\minus{} y)^2} \\plus{} \\frac {1}{(x \\minus{} y)^2} \\plus{} \\frac {1}{(x \\minus{} y)^2})$[/quote]\r\nI think $ F \\equal{} (xy \\plus{} yz \\plus{} zx)(\\frac {1}{(x \\minus{} y)^2} \\plus{} \\frac {1}{(y \\minus{} z)^2} \\plus{} \\frac {1}{(y \\minus{} z)^2})$\r\nIt is an inequality form VMO 2008,the min value is $ 4$\r\nhttp://www.mathlinks.ro/viewtopic.php?search_id=1330177774&t=185643",
"Solution_5": "I don't understand your idea about Pr 1. The ineq is true for every x,y,z satisfying xyz=1.",
"Solution_6": "[quote=\"mr.danh\"]1) Let a,b,c be positive real numbers such that $ abc \\equal{} 1$. Prove that: \n$ \\frac {1}{(a \\plus{} b)b} \\plus{} \\frac {1}{(b \\plus{} c)c} \\plus{} \\frac {1}{(c \\plus{} a)a}\\geq\\frac {3}{2}$\n2) Prove that following ineq: $ (3a \\plus{} 2b \\plus{} c)(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c})\\leq\\frac {45}{2}$\nwhere $ a,b,c \\in [1;2]$\n3) Let x,y,z be non-nagetive real numbers. Find the min value of \n$ F \\equal{} (xy \\plus{} yz \\plus{} zx)(\\frac {1}{(x \\minus{} y)^2} \\plus{} \\frac {1}{(x \\minus{} y)^2} \\plus{} \\frac {1}{(x \\minus{} y)^2})$[/quote]\r\nThe first is in my ebook at toanthpt.net. Here is my solution for it :)\r\n[b]Problem.[/b](VasC) Let $ a,b,c$ be positive real numbers. Prove that\r\n\\[ \\sum \\frac {1}{a(a \\plus{} b)} \\ge \\frac {3}{2\\sqrt [3]{a^2b^2c^2}}\r\n\\]\r\n[i]Solution[/i]",
"Solution_7": "Can you please post the exact address of your email book at at toanthpt.net, can_hang2007.\r\nThank you very much.",
"Solution_8": "plz share it us. :roll:",
"Solution_9": "[quote=\"Sunjee\"]plz share it us. :roll:[/quote]\r\nHere is it (but in Vietnamese) and maybe there is some mistakes in typing, please forgive me if you found them...",
"Solution_10": "Thank you very much, can_hang2007 for this BEAUTIFUL GIFT at all people very keen on inequalities :)"
}
{
"Tag": [
"algebra unsolved",
"algebra"
],
"Problem": "f is defined from $ R to R$ .find all $ F$ such that the folllowing holds,\r\n$ f(x^2\\minus{}y^2)\\equal{}(x\\minus{}y)( f(x) \\plus{} f(y) )$",
"Solution_1": "* $ y \\leftarrow 0 \\Rightarrow f(x^2) \\equal{} x [ f(x) \\plus{} f(0) ] \\forall x$\r\n* $ x \\leftarrow 0 \\Rightarrow f(\\minus{}y^2) \\equal{} (\\minus{}y) [ f(0) \\plus{} f(y) ] \\forall y \\Rightarrow f(\\minus{}x^2) \\equal{} (\\minus{}x) [ f(0) \\plus{} f(x) ]$ (change of variable).\r\nThen\r\n\\[ f(x^2) \\plus{} f(\\minus{}x^2) \\equal{} 0 \\forall x\\]\r\n\\[ \\Rightarrow f(x) \\plus{} f(\\minus{}x) \\equal{} 0 \\forall x\\]\r\n\r\n* $ y \\leftarrow (\\minus{}y) \\Rightarrow f(x^2\\minus{}y^2) \\equal{} (x\\plus{}y) [ f(x) \\plus{} f(\\minus{}y) ] \\equal{} (x\\plus{}y) [ f(x) \\minus{} f(y) ]$\r\nFrom this & the hypothesis, $ 2y f(x) \\equal{} 2x f(y) \\forall x,y$\r\n\r\nThen $ f(x) \\equal{} \\frac{2x f(1)}{2 \\times 1} \\equal{} x f(1) \\forall x$.\r\n\r\n* In conclusion, $ f(x) \\equiv ax$.",
"Solution_2": "can we take any a instead of $ f(1)$???",
"Solution_3": "No matter which $ a$ you choose, $ f(1) \\equal{} a$."
}
{
"Tag": [
"geometry",
"geometry solved"
],
"Problem": "Here is the $COWBOYS$ tale\r\nIn the land of $cowboys$ , at the doom day, $n$ $cowboys$ were dueling. we know that the distances between the $cowboys$ were all different from each other. we know that by the alarm of the city clock , $cowboys$ started shotting eachothers. each $cowboy$ shot the nearest one to himself. prove that:\r\n1. at least one $cowboy$ survived to rule the kingdome.\r\n2. dedicate that at least how many $cowboys$ lost thier lives?",
"Solution_1": "[quote=\"lomos_lupin\"]Here is the $COWBOYS$ tale\nIn the land of $cowboys$ , at the doom day, $n$ $cowboys$ were dueling. we know that the distances between the $cowboys$ were all different from each other. we know that by the alarm of the city clock , $cowboys$ started shotting eachothers. each $cowboy$ shot the nearest one to himself. prove that:\n1. at least one $cowboy$ survived to rule the kingdome.[/quote]\n\nSee http://www.mathlinks.ro/Forum/viewtopic.php?t=28690 for this problem.\n\n[quote=\"lomos_lupin\"]2. dedicate that at least how many $cowboys$ lost thier lives?[/quote]\r\n\r\nDo you think there is a closed form for this number in terms of n ? I don't know, but already the case n = 10 is considerably difficult. See http://www.kalva.demon.co.uk/short/soln/sh00g7.html for this case.\r\n\r\n Darij",
"Solution_2": "as i have posted it in $Geometry Unsolved Problems$ i didnt know the solution but i asked one of our trainig courses proffs (the problem was suggested by iran)\r\nas far as i found out , the problem would be enourmesly hard for greater $n$s. i am sorry for posting it here \r\nit seems it needs stronger tools"
}
{
"Tag": [
"function",
"integration",
"search",
"real analysis",
"real analysis unsolved"
],
"Problem": "If $ f$ is a nonnegative, continuous, concave function on the closed interval $ [0,1]$ such that $ f(0)=1$, then \\[ \\int_0^1 xf(x)dx \\leq \\frac 23 \\left[ %Error. \"diaplaymath\" is a bad command.\n\\int_0^1 f(x)dx \\right]^2.\\] \r\n\r\n[i]Z. Daroczy[/i]",
"Solution_1": "I think it is easier to prove the stronger version\r\n\r\n[hide=\"hint\"]$ \\frac32 \\int_{0}^1 xf(x)dx\\le \\int_0^1f(x)dx \\minus{}\\frac14$[/hide]",
"Solution_2": "Does it help to think of it as:\r\n\r\n$ \\bar{x} \\le \\frac{2}{3} \\int_0^1 f(x) \\,dx$\r\n\r\nwhere $ \\bar{x}$ is the center of mass for $ x$?",
"Solution_3": "Here is the proof for the stronger version.\r\nhttp://www.mathlinks.ro/viewtopic.php?search_id=476848550&t=202028",
"Solution_4": "Schweitzer problem are very interesting"
}
{
"Tag": [
"geometry",
"real analysis",
"advanced fields",
"advanced fields theorems"
],
"Problem": "1. Let there be 'some' points in a plane such that distance between any two them is < 1. What is the minimum area/shape which can inscribe (cover) all these points?\r\n\r\n2. You grab an area by putting infinite points on a plane. Only condition - any two points you put should be <1 apart. What is the maximum area you can grab?\r\n\r\nAny pointers? I think I know the answer but am struggling to justify it. Also if we are able to show (1) above (2) follows directly from that.\r\n\r\nThanks",
"Solution_1": "2. This is called the [color=blue]isodiametric problem[/color]. It has been shown by Bieberbach that the open disk of diameter $ 1$ has largest area.\r\n\r\n1. This is called the [color=red]Lebesgue universal cover problem[/color] and is currently unsolved. There are too many facts to say about it - if interested, Google it."
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "Prove that if $a+b+c|a^2+b^2+c^2$,then there existe infinite many natural numbers $n$,such that $a+b+c|a^n+b^n+c^n$.",
"Solution_1": "This problem is (if I remember well), from a Romanian contest, and has already been posted/solved.",
"Solution_2": "let $u_n=a^n+b^n+c^n$ then $u_{n+2}=(a+b+c)u_{n+1}-(ab+bc+ca)u_{n}+abcu_{n-1}$",
"Solution_3": "[quote=\"Hawk Tiger\"]let $u_n=a^n+b^n+c^n$ then $u_{n+2}=(a+b+c)u_{n+1}-(ab+bc+ca)u_{n}+abcu_{n-1}$[/quote]\r\nAnd what then?",
"Solution_4": "http://www.mathlinks.ro/Forum/viewtopic.php?t=47652"
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Let $a, b, c$ three non-negative real numbers. Prove the following inequality\r\n\r\n\\[ 4(a^{4}+b^{4}+c^{4})+\\frac{4}{3}(ab+bc+ca)^{2}\\geq (ab+bc+ca)(a^{2}+b^{2}+c^{2})+\\frac{1}{3}(a^{2}+b^{2}+c^{2})^{2}. \\]",
"Solution_1": "[quote=\"Cezar Lupu\"]Let $a, b, c$ three non-negative real numbers. Prove the following inequality\n\n\\[ 4(a^{4}+b^{4}+c^{4})+\\frac{4}{3}(ab+bc+ca)^{2}\\geq (ab+bc+ca)(a^{2}+b^{2}+c^{2})+\\frac{1}{3}(a^{2}+b^{2}+c^{2})^{2}. \\][/quote]\r\n\r\nBy Cauchy-Schwarz and AM-GM:\r\n\r\n$4(a^4+b^4+c^4) \\geq \\frac{4}{3}(a^2+b^2+c^2)^2 = \\frac{1}{3}(a^2+b^2+c^2)^2 + (a^2+b^2+c^2)(a^2+b^2+c^2) \\geq \\frac{1}{3}(a^2+b^2+c^2)^2 + (ab+bc+ca)(a^2+b^2+c^2)$\r\n\r\nAnd obviously:\r\n\r\n$\\frac{4}{3}(ab+bc+ca)^2 \\geq 0$"
}
{
"Tag": [
"geometry",
"circumcircle",
"geometry unsolved"
],
"Problem": "Let $ ABC$ is a triangle, let $ M, N, P$ are midpoints of $ AB, AC, BC$ respectively. Let $ (ABN), (ACM), (AMN)$ are circumcircles of $ \\triangle{ABN}, \\triangle{ACM}, \\triangle{AMN}$ respectively. Let $ BN$ meet $ CM$ at point $ G$, let $ (ABN)$ meet $ (ACM)$ at point $ D$, let $ AD$ meet $ (AMN)$ at point $ E$. Prove that $ GE // DP$.\r\n\r\n[color=darkblue][u][b]Note[/b][/u]: Hi, all my friend! I don't like use Argentina TST 2009 to prove this problem.[/color]",
"Solution_1": "http://www.mathlinks.ro/viewtopic.php?t=276858",
"Solution_2": "I have read problem at your link. If we use one of all solution those for prove $ \\frac {AP}{AQ} \\equal{} \\frac {3}{2}$ then $ \\Longrightarrow GE // DP$, I think it's not well and I don't like this way."
}
{
"Tag": [],
"Problem": "How many numbers are in the list $ \\minus{}33, \\minus{}28, \\minus{}23, \\ldots, 52, 57?$",
"Solution_1": "Okay, so difference between each term is $ 5$, so the number of terms is $ \\frac{57 \\minus{} (\\minus{}33)}{5} \\plus{} 1$, or $ \\boxed{19}$.",
"Solution_2": "Why not just add 33 and have 0,5,10,...90. The divide by five- 0,1,2,3,4,5,6,7,8,.....18, which is 18-0+1=19",
"Solution_3": "I like pianolover's solution...but i just subtracted 2 from all the numbers so its: -35, -30, -25,...50, 55...and divided by 5. There are 8 numbers from -35 to 0 plus the 11 positive ones = 19. Its less efficient than most but thats how my brain works :D",
"Solution_4": "I used ernie\u2019s strategy. It\u2019s quick and easy and you can use it for any question like this one.",
"Solution_5": "[quote=pirsquared]I like pianolover's solution...but i just subtracted 2 from all the numbers so its: -35, -30, -25,...50, 55...and divided by 5. There are 8 numbers from -35 to 0 plus the 11 positive ones = 19. Its less efficient than most but thats how my brain works :D[/quote]\n\nAlternatively I [i]added[/i] $3$ like the Alcumus Solution. :-D",
"Solution_6": "Add 34 to each number and then get \n1, 6, 11, . . . 91.\nThen subtract 1 from each number\n0, 5, 10, . . . 90\nThen divide everything by 5 to get\n0, 1, 2, . . . 18.\nThen add 1 to every number to get \n1, 2, 3, . . . 19\nSo the answer is 19",
"Solution_7": "[quote=technoblade]Add 34 to each number and then get \n1, 6, 11, . . . 91.\nThen subtract 1 from each number\n0, 5, 10, . . . 90\nThen divide everything by 5 to get\n0, 1, 2, . . . 18.\nThen add 1 to every number to get \n1, 2, 3, . . . 19\nSo the answer is 19[/quote]\nOG username, nice. o7 Techno.\n\nI have clinical depression. I answered $18$, then $17$.\n",
"Solution_8": "[quote=RTB][quote=technoblade]Add 34 to each number and then get \n1, 6, 11, . . . 91.\nThen subtract 1 from each number\n0, 5, 10, . . . 90\nThen divide everything by 5 to get\n0, 1, 2, . . . 18.\nThen add 1 to every number to get \n1, 2, 3, . . . 19\nSo the answer is 19[/quote]\nOG username, nice. o7 Techno.\n\nI have clinical depression. I answered $18$, then $17$.[/quote]\n\nThis was my first problem on Alcumus. I answered $18$, then $16$ :oops:."
}
{
"Tag": [
"algebra",
"polynomial",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Explain why $ \\mathbb{Q} \\subset \\mathbb{Q}(^3\\sqrt2)$ is not a Galois extension. Find the smallest extension field $ E/\\mathbb{Q}(^3\\sqrt2)$ so that $ E/\\mathbb{Q}$ is Galois. Determine the isomorphism class of $ \\text{Gal}(E/\\mathbb{Q})$.\r\n\r\nI know how to prove that $ \\mathbb{Q} \\subset \\mathbb{Q}(^3\\sqrt2)$ is not a Galois extension. But I don't know how to do the next two parts of this question. Thanks in advance.",
"Solution_1": "What's the splitting field of $ x^3 \\minus{} 2$?",
"Solution_2": "One important property of Galois extensions is that every irreducible polynomial which has a root in such an extension also splits in that extension. What's a good example of an irreducible polynomial with a root in this extension?"
}
{
"Tag": [
"LaTeX",
"calculus",
"derivative",
"real analysis",
"real analysis unsolved"
],
"Problem": "One more problem :\r\nLet g(x)=f(x+2)-2f(x+1)+f(x). Suppose that f(x) is twice differentiable on R. Apply the mean value theorem twice to prove that, for any given a, there is a point b in (a,a+2) such that g(a)=f''(b).",
"Solution_1": "for convenience, let a=-1, i.e. we consider f,g in (-1,1)\r\n\r\nlet h(x)=(f(x)-f(0))/x for non-zero x, and h(0)=f'(0)\r\n\r\nrecall Taylor's:\r\nf(a+y) = f(a) + yf'(a) + y^2f''(a+ky)/2 , 0 1$ for all $ x\\in \\mathbb{N}.$",
"Solution_1": "This doesn't feel like the \"right\" proof for a) (mainly on aesthetic grounds...more on this after the argument), but here goes anyways:\r\n\r\nWe will actually prove the following:\r\n$ \\sum_{k\\equal{}0}^n (\\minus{}1)^k 2^{2n\\plus{}1\\minus{}2k} \\binom{2n\\plus{}1\\minus{}k}{k}\\equal{}2(n\\plus{}1)$\r\n\r\nWe view the left hand side as follows: How many ways are there to tile a $ 2n\\plus{}1 \\times 1$ strip by red $ 1 \\times 1$ squares, blue $ 1 \\times 1$ squares, and uncolored $ 2 \\times 1$ dominoes? To do so, we first need to pick some number $ k$ of dominoes (at least 0, at most $ n$). There will be $ 2n\\plus{}1\\minus{}2k$ $ 1 \\times 1$ squares, so $ \\binom{2n\\plus{}1\\minus{}k}{k}$ ways to place the dominoes relative to the squares. We then color the squares in one of the $ 2^{2n\\plus{}1\\minus{}2k}$ ways. \r\n\r\nIn other words, the left side would exactly count these arrangements, except for that $ (\\minus{}1)^k$ term. So now what we aim to do is to \"pair off\" the arrangements with an even number of dominoes with those with an odd number. We do so as follows: Start from the left of the row of blocks and move right until we encounter either\r\n\r\na) a domino with a $ 1 \\times 1$ square immediately to its left. We replace it by two identical squares of the opposite color of that $ 1 \\times 1$ square. Or \r\n\r\nb) a pair of identically colored squares with a $ 1 \\times 1$ square of the opposite color immediately to their left. We replace them by a domino.\r\n\r\nThis operation is its own inverse, and changes the parity of the number of dominoes, so the sum cancels everywhere the operation is defined (as it cancels on each pair $ (x, f(x))$). However, there are some configurations where it is not defined. \r\n\r\nWhat are those configurations? No domino can appear to the right of a colored square (else we could apply a) ), so they consist of some number of dominoes, then colored squares thereafter. To avoid being able to apply b), the colored squares must consist of some number of identically colored squares, followed by squares alternating in color.\r\n\r\nIn other words, we can define our remaining configurations by picking some number $ k$ of dominoes to start with, picking the color of the next square (2 choices), and choosing how long the next stretch of that color is ($ 2n\\plus{}1\\minus{}2k$ choices). There are therefore $ 2(2n\\plus{}1\\minus{}2k)$ choices with $ k$ dominoes remaining, so the whole sum is $ 2( (2n\\plus{}1)\\minus{}(2n\\minus{}1)\\plus{}(2n\\minus{}3)\\minus{}...\\plus{}1)\\equal{}2(n\\plus{}1)$.\r\n\r\n----------------------------\r\nNow, at the beginning I said I thought this was the wrong argument. There are two reasons for this.\r\n\r\n1. It doesn't generalize to part $ b$. But I don't think any pure combinatorial proof would do that anyways, so I don't mind that too much.\r\n\r\n2. I dislike that I still had more counting and cancellation even after defining my involution. The \"right\" involution would have left exactly $ 2(n\\plus{}1)$ elements unpaired, all of which would have an even number of dominoes (0, perhaps?)\r\n\r\nTherefore my challenge is\r\n\r\n[b]Define a better pairing on the configurations I defined in my proof, in the sense of objection 2.[/b]",
"Solution_2": "[quote=\"kevinatcausa\"]\n\n1. It doesn't generalize to part $ b$. But I don't think any pure combinatorial proof would do that anyways, so I don't mind that too much.\n\n[/quote]\r\nAre you sure ? :lol:",
"Solution_3": "Not at all actually. I guess I was misled somewhat by the notation $ p(n)$...there's no real reason to think of it as a function of $ n$ at all, at least for my proof method. Here's a reworked proof, with what I consider to be the \"right\" bijection:\r\n\r\nAs before, we view the left hand side as counting colorings of a strip of length $ p(n)$ with red and blue $ 1 \\times 1$ blocks and colorless $ 2 \\times 1$ dominoes, with signs determined by the number of dominoes. Given a configuration $ A$, we define $ f(A)$ as follows.\r\n\r\nRead from left to right until the first time we encounter either a domino or a red block immediately to the left of a blue block. We replace the domino by Red/Blue, and Red/Blue by a domino. As $ f(f(A)) \\equal{} A$, the sum over all pairs $ (A,f(A))$ is 0.\r\n\r\nWhat's left is all the colorings containing no dominoes, and no red blocks to the left of blue blocks. In other words, there is some number $ j$ of blue blocks ($ 0 \\leq j \\leq p(n)$), followed by red blocks. There are $ p(n) \\plus{} 1$ choices for $ j$, so the sum is $ p(n) \\plus{} 1$."
}
{
"Tag": [
"AMC",
"AIME"
],
"Problem": "are H20 and H402 the same, explain.\r\n\r\n\r\nyeah , i know this is kind of elementary , but what do you expect from a 5th grader who can't even make it past AIME?",
"Solution_1": "H4O2 is probably referring to two hydrogen bonded molecules of water."
}
{
"Tag": [
"calculus",
"integration",
"trigonometry",
"function",
"calculus computations"
],
"Problem": "Evaluate the following integrals using Laplace Transform:\r\n1. $ \\int_{0}^{\\infty} e^{\\minus{}tx^2} dx$. \r\n2. $ \\int_{0}^{\\infty} \\cos tx^{2} dx$ . Hence evaluate, $ \\int_{0}^{\\infty} \\cos x^{2} dx$\r\n\r\nPl help in these problems. Thanks :)",
"Solution_1": "please can anyone explain me what is laplace transformation?",
"Solution_2": "Let f(t) be a function defined for t > 0. Then, the Laplace transform of f(t), denoted $ \\mathcal{L} [f(t)]$, is, by definition, given by\r\n\r\n$ \\mathcal{L} [f(t)] \\equal{} \\int_0^{ \\plus{} \\infty} f(t) e^{ \\minus{} st}\\,dt \\equal{} F(s)$,\r\n\r\nwhere F(s) is a way of saying that the transform is a function of the variable s. Of course, for the Laplace transform of f(t) to exist the integral in the definition must converge. Sufficient, but not necessary, conditions of convergence are piecewise continuity and exponential order. One the more important applications of the Laplace transform is in solving differential equations and systems of differential equations. Look [url=http://en.wikipedia.org/wiki/Laplace_transform]here[/url].",
"Solution_3": "hello\r\n1. $ \\int^{\\infty}_{0}e^{\\minus{}tx^2} dx$\r\nputting $ tx^2\\equal{}y , dx\\equal{}\\frac{dy}{2\\sqrt{ty}}$\r\n=$ \\int^{\\infty}_{0}e^{\\minus{}y}\\frac{dy}{2\\sqrt{ty}}$\r\n=$ \\frac{1}{2\\sqrt{t}}\\int^{\\infty}_{0}e^{\\minus{}y}y^{\\minus{}\\frac{1}{2}} dy$\r\nAs u know \r\n$ \\int^{\\infty}_{0}e^{\\minus{}x}x^{n\\minus{}1}dx \\equal{}$Gamma (n)\r\n=$ \\frac{1}{2\\sqrt{t}}$Gamma($ \\frac{1}{2}$)\r\n=$ \\frac{\\sqrt{\\pi}}{2\\sqrt{t}}$\r\nThank u"
}
{
"Tag": [],
"Problem": "Bueno, queremos ver como esta el nivel de ingles en la comunidad latina para difusion de articulos:\r\n\r\nAlgun Comentario?",
"Solution_1": "[color=darkblue]Bueno........ :? me defiendo en algo!!...pero m\u00e1s all\u00e1 todos estar\u00e1n(y los que no deberian hacerlo) estudiando idiomas basicos para la difusion de articulos(como el ingles), pero buena iniciativa de consultar como va la gente de ingles :) [/color]",
"Solution_2": "[quote=\"Pascual2005\"]Bueno, queremos ver como esta el nivel de ingles en la comunidad latina para difusion de articulos:\n\nAlgun Comentario?[/quote]\r\n\r\nYo estoy en mi 12\u00ba a\u00f1o de Ingl\u00e9s; no soy un experto ni mucho menos, pero lo b\u00e1sico lo s\u00e9, y puedo tanto escuchar, como leer, escribir y hablar. \r\n\r\nSobre el tema de si saber un idioma o no es importante (adem\u00e1s de la lengua principal, obviamente) creo que si, que es una herramienta tanto para enriquecerce culturalmente como para trabajar.",
"Solution_3": "[color=blue]Desde principios del a\u00f1o pasado estudio ingl\u00e9s. Puedo afirmar que aprend\u00ed basante, pues entiendo bastante bien cuando alguien me habla en dicho idioma y, a pesar de que no me resulta nada sencillo, puedo tambi\u00e9n responder. Lo he podido comprobar en la I.M.O. de este a\u00f1o.\nPor supuesto, siempre es m\u00e1s f\u00e1cil entender el lenguaje escrito.\n\nA pesar de ello, tengo ciertas complicaciones con el ingl\u00e9s matem\u00e1tico. (De hecho, reci\u00e9n pude averiaguar lo que significa [i]odd[/i]\u2026 :blush:)[/color]",
"Solution_4": "Est\u00e1 bien ...",
"Solution_5": "Se refiere a un numero impar",
"Solution_6": "Igual me cuesta un poco. Pero con el tiempo, mejorar\u00e1. \r\n\r\nSaludos"
}
{
"Tag": [],
"Problem": "Natalie and Asya can finish a job in 1.5 hours. Megan and Asya can finish a job in 2 hours. Natalie and Megan can finish a job in one hour. How long would it take them to finish if all three worked on the job?\r\n\r\nMy answer\r\n[hide]$ \\frac {12} {13}$[/hide]\n\nTheir answer\n[hide]$ \\frac {13} {12}$[/hide]\n\nThe reason I disagree\n[hide]If one pair can finish it in one hour, why would the addition of a worker increase the amount of time before the job was done?[/hide]",
"Solution_1": "Here is why $ \\frac{13}{12}$ is right:\r\n\r\nLet $ a, m, n$ be Asya, Megan, and Natalie respectfully. We then have:\r\n\\[ (\\frac1n+\\frac1a)\\frac32 = 1\r\n\\]\r\n\\[ (\\frac1m + \\frac1a )2 = 1\r\n\\]\r\n\\[ ( \\frac1n + \\frac1m )1 = 1\r\n\\]\r\n\r\nDivide each equation by the number that is multiplying the parenthesis. You get:\r\n\\[ \\frac1n + \\frac1a = \\frac23\r\n\\]\r\n\\[ \\frac1m + \\frac1a = \\frac12\r\n\\]\r\n\\[ \\frac1m + \\frac1n = 1\r\n\\]\r\n\r\nAdd all those and divide by $ 2$. You get the final answer of $ \\boxed{\\frac{13}{12}}$\r\n\r\nI see why you think it should be less, but really, I don't know.",
"Solution_2": "the reason that it would be greater than an hour is because the pesron who is NOT involved in that one hour duo has less productivity, so the other two have to work for less time with someone who has less, but still SOME productivity, and still does some work. when you think about it, it makes sense-the duo who could do it in an hour isn't getting enough work OR time to do it.",
"Solution_3": "Actually 12/13 is the right answer. The work that 007math posted wasn't finished. When you add the 3 equations together and divide by 2, you are finding 1/a + 1/m + 1/n = 13/12, but that is the amount of the job that the three working together complete in one hour. If they finish 13/12 of the work in 1 hour, they finished 1 job in 12/13 hour. One lesson here is, don't ignore common sense! :)"
}
{
"Tag": [
"geometry",
"perimeter",
"Pythagorean Theorem"
],
"Problem": "Terry bicycled due north for 7 miles and due east for 24 miles. He then returned straight back to his starting point. How many miles did he bicycle on the trip?",
"Solution_1": "Drawing a diagram gives us a right triangle with legs 7 and 24, because distance is defined as the length of the perpendicular line from one place to the next. You could manually do the Pythagorean Theorem to find that the answer is 25, or just recognize it as a triple.",
"Solution_2": "When it says how many miles does he bicycle on his trip, is that referring to the whole entire bicycling? If it is that, then we would have $ 7\\plus{}24\\plus{}25\\equal{}\\boxed{56}$ instead...",
"Solution_3": "The answer is not 25, as they asked for the total number of miles he went. The answer, as jxl28 said, is the perimeter of the triangle, or $ 56$."
}
{
"Tag": [
"probability"
],
"Problem": "If 3*x^2 + 9*k*x + 3*x +1 = 0, where k is a real number such that k < 2, what is the probability that the equation has real roots ?",
"Solution_1": "What do you mean by probability? There are infinitely many such k that the equation does have real roots, and infinitely many such that the equation does not have real roots",
"Solution_2": "The problem is on The Arbelos, volume 5, chapter 3.",
"Solution_3": "[quote=\"lo zio\"]If 3*x^2 + 9*k*x + 3*x +1 = 0, where k is a real number such that k < 2, what is the probability that the equation has real roots ?[/quote]Taking the discriminant, we get $\\sqrt{b^{2}-4ac}=\\sqrt{(9k+3)^{2}-12}$. The equation does not have real solutions in the interval $(-\\frac19 \\sqrt{12}-\\frac13, \\frac19 \\sqrt{12}-\\frac13)$. Anything less than the lower bound suffices, making the probability 1. (If k had a lower bound the probability would be less than 1.)",
"Solution_4": "There is a mistake in the post because the probability actually is not 1. There are infinitely many k such that the equation does not have real roots ( for example k = 0).",
"Solution_5": "Since the interval $]\\infty , 2[$ the desired $k$ comes from has infinite length, you have to define what probability even means (how you choose reals by random).\r\nIn a natural and intuitive way (which can also be done exact by e.g. limits), the probability of numbers that have non-real roots is indeed $0$.\r\nThis can happen, independent from the infinity of that set.\r\n\r\nOr is the true condition something like $|k| <2$ \u00bf",
"Solution_6": "I was going to ask as well if it was intended to be absolute value. The probability given from what my post says is the infinitely bounded interval length divided by itself plus the finitely bounded interval length, thus it is 1. This is due to the continuous nature of the variable.",
"Solution_7": "[quote=\"Hokkage\"]This is due to the continuous nature of the variable.[/quote]\r\nTo be even more precise, the cardinality of any interval $(a,b)$, with $a 0$, the variation of $ M_t(w)$ over $ [0;T]$ is finite. (i.e. $ \\sup_{t_1 1$, the bounded region determined by $ C$ contains two zeros of $ f$, namely $ z_1 \\equal{} e^{i\\alpha / 2}$ and $ z_2 \\equal{} \\minus{}e^{\\minus{}i\\alpha / 2}$. Thus by residue theorem, \r\n\r\n$ \\int_{C} f(z) \\, dz \\equal{} 2\\pi i \\left[ \\underset{z\\equal{}z_1}{\\text{Res}} f(z) \\plus{} \\underset{z\\equal{}z_2}{\\text{Res}} f(z)\\right] \\equal{} 2 \\pi i e^{\\pi q i} \\frac{\\sin q (\\beta \\minus{} \\pi)}{\\sin \\beta}$.\r\n\r\nAlso, since $ \\textstyle \\int_{\\Gamma} f(z) \\, dz$ vanishes as $ R \\to \\infty$, we have\r\n\r\n$ \\int_{C} f(z) \\, dz \\equal{} \\int_{\\minus{}\\infty}^{0} f(x) \\, dx \\plus{} \\int_{0}^{\\infty} f(x) \\, dx \\equal{} (1 \\minus{} e^{2q \\pi i}) \\int_{0}^{\\infty} f(x) \\, dx.$\r\n\r\nCombining these two equations gives the desired equality."
}
{
"Tag": [],
"Problem": "In a triangle $ABC$ the median $AM$ is equall to $BC$ . On $AC$ we take point $D$ with $2AD = CD$ . If $N$ is midpoint of $BD$ , prove that \r\n $CN = CD$.\r\n\r\n [u]Babis[/u]",
"Solution_1": "Well , I see that you have difficulties. The problem is proposed in Gazeta matematica 2005 - Romany. I have made a solution. I'll send it after some days. Try it again! It is not difficult but not easy!\r\n \r\nBabis",
"Solution_2": "[quote=\"stergiu\"]Well , I see that you have difficulties. The problem is proposed in Gazeta matematica 2005 - Romany. I have made a solution. I'll send it after some days. Try it again! It is not difficult but not easy!\n \nBabis[/quote]\r\nThank for sharing a problem. I think there is a rule here that people would not solve a recent problem from a journal. It has to be at least one year old. It's not fair that someone come here, gets the solution, and then go send the solution to journals to publish an unjust work and credit to someone who did not do the problem. It's just common sense.",
"Solution_3": "From an other point of vew I agree with you.But I think that in this forum nobody wants to present a solution as own , if it not belong to him. Any way , SORRY!\r\n I do not send the solution yet.\r\n\r\n Babis",
"Solution_4": "[i]Lets clear this topic with the solution. It took me 15 minutes to find the solution , so I share it with you. [/i]- [b]No objections please ![/b]\r\n We have \r\n\r\n $MN = \\frac{1}2 DC = AD$ , so $ADMN$ is a parallelogram. If $AM$ meets $ND$ at $P$ , then $P$ is the mid point of $ND$ and $AM$ .\r\n Since $PM = \\frac{1}2 AM = \\frac{BC}2$ , we take that $CP \\perp BP$.\r\n But then , in triangle $\\triangle CDN$, $CP$ is both altitude and median , so $CD = CN$ .\r\n\r\n[u]Babis[/u]",
"Solution_5": "a beautiful solution for a nice problem. Thanks for the problem."
}
{
"Tag": [
"symmetry"
],
"Problem": "Anyway, I take a Combinatorics elective at my school, and our instructor posed this questiona nd the class didnt finish in time before the class came to an end. He told me to post it here and challenge you all.\r\n\r\nWe have the set {1,2,3....,17,18,19,20}\r\n\r\nwe want to know how many subsets exist such that the sum of the numbers in the subset is divisible by five.\r\n\r\nfor example, {1,4} is such a subset, but {1,5,7} is not (random examples).\r\n\r\nhave fun. \r\n\r\ni also apologize if this came out of a book, for my instructor never told me of a book to put a referance to. If anyone knows, they can put the book.",
"Solution_1": "[hide]There are 4 subsets with one element.\nThere are 8 two element subsets containing 1 or 2 but since {2,3} is the same as {3,2}, there's one less for 3-7. But for 8-12, there's two less since {8,2} and {8,7} have already been done. For 13-17, there's three less since {13,2} {13,7} {13,12} has already been done. For 18-20, there is none since all of its possiblities have been done in previous numbers. 2*4+5*3+5*2+5*1+3*0=8+15+10+5=38.\nNotice this has a symmetry just like Pascal's triangle. For example there's 1 20 element subset, since 1+2+...+20 is a multiple of five. The only way to have a nineteen element subset that works is to subtract a single element set that works from the 20 element set, giving 4 possibilities. So in other words, we only have to find the values for 1-10. We sum the values for 1-9 element sets and multiply that by 2 since that will be the same for 11-19 element sets. Then, we add the value for a ten element set and 1 (20 element set) and we're done. I think the formula for subsets in x-element set for this problem is (20 choose x)/5 where the result is always rounded up to the nearest integer. Now, we have 2(4+38+228+969+3101+7752+15504+25194+33592)+36952+1=209717 subsets. Am I right? If I am, can anyone tell me why my formula works, if it did work.[/hide] \r\nI know, I had to use a calculator to do this, but I didn't want to calculate 4+38+228+969+4*19*3*17*16/5+... larger numbers that would take me even more time to solve.",
"Solution_2": "[quote]I think the formula for subsets in x-element set for this problem is (20 choose x)/5 where the result is always rounded up to the nearest integer.[/quote]\r\nI've never heard that theorem before...\r\n\r\nAlso, we do not multiply the number of n-element subsets from 1-10, the converse of a 10 element subset is another 10-element subset. We need to multiply subsets from 0-9 by 2 and then add the number of subsets of size 10."
}
{
"Tag": [
"real analysis",
"topology",
"real analysis unsolved"
],
"Problem": "$E \\subset \\mathbb{R}$ a measurable subset with null lebesgue measure. Does there always exist an homeomorphism $h: \\mathbb{R}\\to \\mathbb{R}$ such that $E$ does not intersect $h(E)$ ?",
"Solution_1": "Not if $\\mathbb R\\setminus E$ is of first category.\r\n\r\nWhat if $E$ is assumed to be of first category instead of measure zero? :maybe:"
}
{
"Tag": [
"Asymptote",
"geometry",
"circumcircle",
"function"
],
"Problem": "I know how to construct perpendicular bisectors now with straightedge and compass, and was hoping to try something similar with Asymptote (I know, it has something for the circumcenter), but how to you make a dot on the intersection points of two circles, lines, etc. And also, how do you connect them?",
"Solution_1": "Asymptote introduced a new function that gives an array of intersection points in version 1.21 where you can have intersectionpoint[color=red]s[/color](path p, path q). Note the s on the end of intersectionpoint[color=red]s[/color].\r\nSo make sure you have a recent version (the latest is 1.28) and then\r\n[code]size(8cm,0); \nimport math; \nimport graph; \nreal r,s; \npair a,b, common; \npath circ1, circ2; \nr=1; s=1; \na=(0,0); \nb=(1,0); \ncirc1=circle(a,r); \ncirc2=circle(b,s); \ndraw(circ1,linewidth(1bp)); \ndraw(circ2,1bp+green); \npair [] x=intersectionpoints(circ1, circ2); \ndot(x[0],3bp+blue);\ndot(x[1],3bp+blue);\ndraw(x[0] -- x[1],1bp+red);[/code]will give\r\n[asy]size(8cm,0); import math; import graph; real r,s; pair a,b, common; path circ1, circ2; r=1; s=1; a=(0,0); b=(1,0); circ1=circle(a,r); circ2=circle(b,s); draw(circ1,linewidth(1bp)); draw(circ2,1bp+green); pair [] x=intersectionpoints(circ1, circ2); dot(x[0],3bp+blue); dot(x[1],3bp+blue); draw(x[0] -- x[1],1bp+red);[/asy]"
}
{
"Tag": [
"logarithms",
"algebra",
"polynomial"
],
"Problem": "Does there exist DISTINCT reals $a,b,c,d$, such that $a+b=c+d$ and $ab=cd$?",
"Solution_1": "[hide=\"Solution\"] $ab = cd \\implies \\log a+\\log b = \\log c+\\log d$. But by Jensen's this implies $\\log$ is linear, contradiction. Hence we cannot have such reals. [/hide]",
"Solution_2": "Alternatively (and more simply),\r\n[hide]Consider the two polynomials (x - a)(x - b) and (x - c)(x - d).[/hide]\n\nOr, instead, [hide]set $c = a+r, d = b-r$ for some real number $r$.[/hide]\r\n\r\nSo, that's 3 distinct ways; do we have any more?",
"Solution_3": "[hide]$a+b=c+d,ab=cd,a=c+d-b,b(c+d-b)=cd,bc+bd-b^{2}=cd,b(c-b)=d(c-b),b=d$\nhence no solutions.[/hide]\r\nI tried to generalise but I couldn't....is it possible?",
"Solution_4": "[quote=\"srulikbd\"]I tried to generalise but I couldn't....is it possible?[/quote]\r\n\r\nIt's not possible with just the sum and the product. But, as my first solution suggests, it is possible provided you give more equations, such as the symmetric sums of the roots taken $n$ at a time. There are many examples of failure for 3 numbers with just the sum and the product. {0, r, -r} has the same sum and product for any r, for example.",
"Solution_5": "[hide]substitute a, we get a=cd/b \n\nwe know that:\n a+b=c+d \ncd+b\u00b2=cb+db\nb\u00b2-db=cb-cd\nb(b-d)=c(b-d)\nb=c=impossible[/hide]",
"Solution_6": "[quote=\"t0rajir0u\"][hide=\"Solution\"] $ab = cd \\implies \\log a+\\log b = \\log c+\\log d$. But by Jensen's this implies $\\log$ is linear, contradiction. Hence we cannot have such reals. [/hide][/quote]\r\n\r\nCould you elaborate? I don't follow.",
"Solution_7": "Ah, I should probably not have used that argument. It's extremely unrigorous (and Jensen's wasn't what I really meant), but uses the concavity of $\\log$: the idea is to assume WLOG that $(a, b)$ majorizes $(c, d)$, which implies that the LHS must be strictly less than the RHS.",
"Solution_8": "[hide]\nWLOG, let $a R be periodic with period 2pi, and Riemann-integrable on [-pi,pi]. Suppose that, for some real number alpha > 0, the function f is identically equal to zero in the interval (-alpha,alpha). Prove that for each beta in (0,alpha), the Fourier series of f converges uniformly on [-beta,beta].",
"Solution_1": "$f_{n}(x)=(f*D_{n})(x) = \\frac{1}{2 \\pi}\\int_{a<|t|<\\pi}f(t) \\frac{\\sin((n+\\frac{1}{2})(x-t))}{\\sin(\\frac{1}{2}(x-t))}dt$ is the partial sum of the Fourier expansion of $f$. Suppose for a moment that $f$ is $C^{\\infty}$ periodic, and take $-a<-ba-b$ should conclude this particular case. For the general problem, remember that smooth functions are dense in $L^{1}$."
}
{
"Tag": [
"geometry",
"quadratics",
"perimeter",
"ratio",
"trigonometry",
"algebra",
"quadratic formula"
],
"Problem": "Organized List of Geometry Solutions:\r\nProblems are located at http://155.225.48.46/statemat/stuf2004/Geoprobs.htm\r\n\r\n[u]Solutions[/u]\r\n[hide=\"Number 2 by Joe\"]\nFirst, we use the fact that the altitude of a right right triangle is the geometric mean of the two segments that it divides the hypotenuse into: \\[ \\frac p{\\frac 14(p+q)}=\\frac{\\frac 14(p+q)}q\\Rightarrow p^2-14pq+q^2=0. \\] Now, divide through by $q^2$: \\[ \\left(\\frac pq\\right)^2-14\\left(\\frac pq\\right)+1=0. \\] Using the quadratic formula, \\[ \\frac pq=\\frac{14\\pm\\sqrt{196-4(1)(1)}}2=7\\pm4\\sqrt{3}. \\] Since the problem tells you it is in the form $a-b\\sqrt{3}$ you could be done right now. But assume you didn't know that. How would you decide between $7+4\\sqrt{3}$ and $7-\\sqrt{3}.$ Simple. We were given that $p 0$, we need at most $ 3k \\plus{} 3$ weighings. Proceed by induction on $ n\\ge2$, where $ n$ is the exponent of $ 3$ in the number of coins. As I said, I cannot prove the base case, $ n \\equal{} 2$. But here's the induction step:\r\n\r\nLet $ n > 2$.\r\n\r\n[i]Case 1. $ n$ is odd. [/i]\r\nLet $ n \\equal{} 2k \\plus{} 1$. Divide all coins in three groups: $ A$ , $ B$ and $ C$. Put $ A$ and $ B$ on balance $ B_1$ and then on balance $ B_2$. If we obtain the same result, since one of the balances works fine, this results is correct, hence we deduce in what heap - $ A$, $ B$ or $ C$ is the fake coin. According to the induction, we cand find the fake coin in $ 3k \\plus{} 1$ weighings now, giving a total of $ 2 \\plus{} 3k \\plus{} 1 \\equal{} 3k \\plus{} 3$ weighings.\r\nIf $ B_1$ and $ B_2$ give different results, then one of them is broken, implying $ B_3$ works fine. Using only $ B_3$, we deduce in $ 2k \\plus{} 1$ steps the fake coin, making $ 2k \\plus{} 3 < 3k \\plus{} 3$ weighings.\r\n\r\n[i]Case 1. $ n$ is even[/i]. \r\nLet $ n \\equal{} 2k$, $ k\\ge2$. Take an arbitray balance, say $ B_1$. Let's presume it works just fine. Then using $ k$ weighings with it we obtain a heap $ H$ of $ 3^k$ coins, among which is the fake coin (we suppose so). Before the step we obtained $ H$, we had $ 3^{k \\plus{} 1}$ coins, and we divided it into 3 heaps, say $ H,I,J$. We compared two of them, say $ A$ and $ B$, where $ A,B\\in\\{H,I,J\\}$ and concluded that $ H$ was easier.\r\nNow compare again $ I$ and $ H$ on another balance, say $ B_2$. If we do NOT obtain $ I > H$, then it means that either $ B_1$ or $ B_2$ is broken; or, that the third balance $ B_3$ works ok. Using $ B_3$ and all original $ 3^{2k}$ coins we can establish in $ 2k$ more weighings the fake coin, having done exactly $ 3k \\plus{} 1$ weighings. If we do obtain $ I > H$ then it means that in $ H$ is indeed the fake coin! That is because either $ B_1$ or $ B_2$ is correct, so either all first $ k$ weighings have been correct (and then the fake coin is indeed in $ H$), or the $ k \\plus{} 1$th weighing is correct, implying $ I > H$, thus $ H$ indeed contains the fake coin! At this moment we apply the induction hypothesis either for the $ 3^k$ coins of $ H$. If $ k$ is odd we will use $ \\displaystyle\\frac {3(k \\plus{} 1)}2$ more weighings, giving a total of $ \\displaystyle w \\equal{} k \\plus{} 1 \\plus{} \\frac {3(k \\plus{} 1)}2$ weighings. Using the fact that $ k$ is odd and $ \\ge3$ one can easily establish the inequality $ w\\le 3k \\plus{} 1$. If $ k$ is even then we use $ \\displaystyle\\frac {3k}2 \\plus{} 1$ weighings, giving a total of $ w \\equal{} k \\plus{} 1 \\plus{} \\frac {3k}2 \\plus{} 1\\le 3k \\plus{} 1$ weighings, what we needed.\r\n\r\n\r\nNote that even if for $ 9$ coins we need at least $ 5$ weighings, then [b]the proof still works[/b] for $ k\\ge4$ (that is for $ n > \\equal{} 8$), [hide=\"because\"] assuming we need at least $ 5$ weighings to find the fake coin among $ 9$ coins (this is certainly possible), the above proof still works for larger $ k$, namely $ k\\ge 4$ (or $ n\\ge8$), because the involved inequalities are strict for larger values of $ k$).\nFor example, if we denote by $ w_i$, $ i\\ge2$ the minimal number of weighings necessary to find a fake coin among $ 3^i$ then if we take $ w_2 \\equal{} 5$, the above algorithm gives $ w_3\\le 2 \\plus{} w_2 \\equal{} 7$, $ w_4\\le2 \\plus{} 1 \\plus{} w_2 \\equal{} 8$, $ w_5\\le 2 \\plus{} w_4 \\equal{} 10$, $ w_6\\le 3 \\plus{} 1 \\plus{} w_3\\le 11$, $ w_7\\le 2 \\plus{} w_6\\le13$, $ \\boxed{w_8\\le4 \\plus{} 1 \\plus{} w_4 \\equal{} 13 \\equal{} 3\\cdot4 \\plus{} 1}$, $ w_9\\le 2 \\plus{} w_8 \\equal{} 15$, $ \\boxed{w_{10}\\le5 \\plus{} 1 \\plus{} 10 \\equal{} 16 \\equal{} 3\\cdot5 \\plus{} 1}$, $ w_{11}\\le 2 \\plus{} w_{10} \\equal{} 18$, $ \\boxed{w_{12}\\le\\text{max}\\{3\\cdot 6 \\plus{} 1,6 \\plus{} 1 \\plus{} w_6\\} \\equal{} 19}$, $ w_{13}\\le 2 \\plus{} w_12 \\equal{} 21$, and so on. [/hide]",
"Solution_5": "[quote]Very nice problem! Though I wasn't able to prove it for $ 4$ coins, and $ 9$ weighings and this is all I need for the solution to be complete. Maybe someone can help with this case. \n[/quote]\r\nMe too :blush:",
"Solution_6": "Base case $ 9$ coins and $ 4$ weighings.\r\n\r\nPick two coins $ C_1$ and $ C_2$ and with two weighings on $ B_1$ and $ B_2$ identify one good balance $ B_g$.\r\n(same result then $ B_g\\equal{}B_1$, different result then $ B_g\\equal{}B_3$)\r\n\r\nPut three coins on each plate of $ B_g$ then pick the lighter group of $ 3$ coins\r\nOf the lighter group put one coin on each plate of $ B_g$ then pick the lighter coin.\r\n\r\nDid I misunderstood the problem?",
"Solution_7": "[quote=\"mszew\"]Base case $ 9$ coins and $ 4$ weighings.\n\nPick two coins $ C_1$ and $ C_2$ and with two weighings on $ B_1$ and $ B_2$ identify one good balance $ B_g$.\n[/quote]\r\n\r\nHow do you identify the good balance? If the results on the two balances are different, then you identify $ B_3$ as a good balance. But if results on $ B_1$ and $ B_2$ are the same then you cannot know for sure which balance is good. The broken balance need not give wrong results.",
"Solution_8": "... but it was fun to decipher what was written in Russian!\r\n\r\nI'll just solve the case $ k\\equal{}1$, since it follows almost directly by induction after that (as remarked before).\r\n\r\nArrange the $ 9$ coins in a $ 3\\times 3$ table:\r\n\\[ \\begin{matrix}\r\n1&2&3\\\\\r\n4&5&6\\\\\r\n7&8&9\\\\\r\n\\end{matrix}\\]\r\n\r\nThe first weighting is to compare rows $ 1$ and $ 2$ on balance $ 1$ (that is, $ \\{1,2,3\\}\\times \\{4,5,6\\}$). The second one is to compare columns $ 1$ and $ 2$ on balance $ 2$ (that is, $ \\{1,4,7\\}\\times \\{2,5,8\\}$). The third weighting depends on the results of the previous two weightings.\r\n\r\nSuppose, for instance, that balance $ 1$ indicates $ \\{1,2,3\\}< \\{4,5,6\\}$ and balance $ 2$ indicates $ \\{1,4,7\\} \\equal{} \\{2,5,8\\}$. Let's see what information we can get.\r\n\r\nIf the first balance is broken, the only information we get is from balance $ 2$, that is, the lightest coin is in column $ 3$. If the second one is broken, then the lighest coin is in row $ 1$. If balance $ 3$ is broken, the lightest coin is $ 3$.\r\n\r\nUnfortunately, we don't know which balance is broken, so the only thing we know is that the candidates for lightest coin are $ 1,2,3,6,9$ (row $ 1$ and column $ 3$). Then the third weighting is $ \\{1,2\\}\\times \\{6,9\\}$ on balance $ 3$. Let's see what the third balance indicates in each of the three cases: in the first case, either $ \\{1,2\\} \\equal{} \\{6,9\\}$ (in this case, the lightest coin is $ 3$) or $ \\{1,2\\} > \\{6,9\\}$; in the first case, either $ \\{1,2\\} \\equal{} \\{6,9\\}$ (again, $ 3$ is the lighest coin) or $ \\{1,2\\} < \\{6,9\\}$; in the third case, ot can go either way, but we already know that $ 3$ is the lightest coin.\r\n\r\nNow let's face it backwards: if balance $ 3$ indicates $ \\{1,2\\} > \\{6,9\\}$ either balance $ 1$ or $ 3$ is broken, anyway, balance $ 2$ is fine, and, additionally, the lightest coin is in column $ 3$. If balance $ 3$ indicates $ \\{1,2\\} < \\{6,9\\}$ then balance $ 1$ is fine and the lightest coin is in row $ 1$. If balance $ 3$ indicates $ \\{1,2\\} \\equal{} \\{6,9\\}$ then the lightest coin is $ 3$.\r\n\r\nNotice that now either we already know which coin is the lightest or we know a fine balance and a set of three coins, one of which is the lightest one, which can be determined by one more weighting.\r\n\r\nThe other cases are handled similarly.\r\n\r\nBy the way, it also says in the booklet that it is possible to determine the lightest among $ 3^{n(n\\plus{}1)}$ coins with at most $ (n\\plus{}1)^2$ weightings (there are still three balances, one of which is broken). Anyone?",
"Solution_9": "russian balance problems will be the death of me\n\nIt suffices to show that given nine coins, we can either find the lightest in $3$ moves or find the lightest in $4$ moves [i]and[/i] identify a non-broken scale. Then we are done by induction (by breaking the $3^{2k}$ coins into nine \"mega-coins\"), since once a non-broken scale is identified we can clearly find the lightest coin among $3$ in $1$ move using it.\n\nNumber the coins $1$ through $9$. First use two different scales to weigh $1,2,3$ against $4,5,6$, and then weigh $1,4,7$ against $2,5,8$. For symmetry reasons WLOG let both scales say that the group containing the coin $1$ is lighter. Then use the third scale to weigh $2,3$ against $4,7$. Finally,\n[list]\n[*] If the result is that $2,3$ and $4,7$ are balanced then the $1$ coin must be the lightest, since any pair of correct results alone implies that $1$ is the lightest coin, so we are done.\n[*] Otherwise WLOG the result is that $2,3$ is lighter (so one scale provides an incorrect result). Then the only possible candidates for lightest coin are $1,2,3$, since anything else implies that at least two scales lied. It therefore cannot be the case that the first scale (used to weigh $1,2,3$ against $4,5,6$) provided an incorrect result, implying that it is guaranteed to be non-broken. Then use it to weigh $2$ against $3$; the (correct) result will indicate the lightest coin among $1,2,3$.\n[/list]\nThis finishes the problem. $\\blacksquare$"
}
{
"Tag": [
"vector",
"linear algebra",
"matrix",
"algebra",
"polynomial",
"calculus",
"integration"
],
"Problem": "Define a cyclic diffrence operator $\\Delta $ on an integer $n-$tuple $a=(a_1,a_2,\\dots,a_n), \\Delta (a)=(a_2-a_1,\\dots,a_n-a_{n-1},a_1-a_n)$.Determine all solutions $\\Delta ^k(a)=a$.",
"Solution_1": "Let's consider $a$ to be a column vector, and $\\Delta$ to be a matrix acting on such vectors. It's easy to find the expression of $\\Delta$ and to see that it's characteristic polynomial $\\det(xI_n-\\Delta)$ equals $(x+1)^n+(-1)^n$.\r\n\r\nIn order for the equation $\\Delta^kx=x$ to have a non-trivial solution (I mean $x$ different from the null vector), $\\Delta^k$ must have $1$ as one of its eigenvalues, so $\\Delta$ must have an eigenvalue $x_0$ of absolute value $1$. We have $|x_0|=|1+x_0|=1$, so $x_0\\in\\{\\varepsilon,\\bar\\varepsilon\\}$, the two cubic roots different from $1$. This means that $1+x_0\\in\\{\\omega,\\bar\\omega\\}$, where $\\omega=e^{\\frac{\\pi i}3}$.\r\nThe equation becomes $\\omega^n+(-1)^n=0$, and we can easily check that this does not hold for any $n$, which means that there are no non-trivial solutions for our equation.\r\n\r\nI'm afraid I'm not entirely sure about this. I didn't use the fact that the vectors have to have integral entries, for example :?.",
"Solution_2": "grobber, you're overlooking the possiblities in the $n=6$ case.\r\n\r\nLet $x=(1,1,0,-1,-1,0).$ Then $\\Delta^3x=x.$",
"Solution_3": "Yes, I've found my mistake: the characteristic polynomial is $(x+1)^n-1$, so some cases did slip out. Thanks for pointing that out.\r\n\r\nWhen $n$ is even, we do have a chance whenever it's also a multiple of $6$ (because we must have $\\omega^n-1=0$). In such cases, $\\varepsilon$ is, indeed, an eigenvalue of $\\Delta$, and in order to have $1$ as an eigenvalue of $\\Delta^k$, the necessary and sufficient condition is to have $3|k$.\r\n\r\nI guess it's not solved, then, it's just reduced to the cases $6|n,3|k$ :(. I knew it couldn't have been that easy.. :)",
"Solution_4": "I'm looking at it from a discrete Fourier transform point of view. I haven't settled the details, but I'm starting to think that if you want integer-valued solutions, the example I posted in #3 is essentially the only example. (Oh, you can rotate it around, and multiply by integer scalars but that's it.)",
"Solution_5": "Actually, the work is almost entirely done :). Let's assume, as said above, that $6|n,3|k$.\r\n\r\n$\\Delta$ is diagonalizable, since its characteristic polynomial clearly has simple roots. This means that $\\Delta^k$ is also diagonalizable, so the algebraic multiplicities of its eigenvalues are equal to the geometric multiplicities. The only $k$'th roots of unity which are also eigenvalues of $\\Delta$ are $\\varepsilon,\\bar\\varepsilon$, and these have multiplicity $1$. This means that $1$, as an eigenvalue of $\\Delta^k$, has multiplicity $2$, which means that its eigenspace has dimension $1$. Since we can find two linearly independent eigenvectors, $(1,1,0,-1-1,0,\\ldots)$ and $(0,-1,-1,0,1,1,\\ldots)$, it means that the complex vectors fixed by $\\Delta^k$ are precisely the linear combinations of these two. \r\n\r\nSince we want vectors with integral entires, it means that the coefficients of the linear combinations must be integers, and we have found our integral solutions.\r\n\r\nI hope it's Ok this time.",
"Solution_6": "Let $T$ be the operator of translation $a_i \\to a_{i+1}$ on the space of cyclic n-tuples.\r\nThen the problem is (literally) asking for simultaneous solution of \r\n\r\n$T^n = 1$ \r\n$(T-1)^k = 1$\r\n\r\nEigenvalues of $T$ and $(T-1)$ are roots of unity that differ by 1,\r\nso the only possibilty is $n$ a multiple of $6$, $k$ a multiple of 3, and $T^2 - T + 1 = 0$.\r\nThis gives the same solutions as above.\r\n\r\n[b]Question[/b]: classify the solutions in characteristic $p$ (or, equivalently, find the ideal in $\\mathbb{Z}[T]$ generated by $T^n-1$ and $(T-1)^k-1$)."
}
{
"Tag": [
"geometry",
"geometry unsolved"
],
"Problem": "Find all plane triangles whose sides have integer length and whose incircles have unit radius.",
"Solution_1": "[quote=\"orl\"]Find all plane triangles whose sides have integer length and whose incircles have unit radius.[/quote]\n use this : $S=\\sqrt {p(p-a)(p-b)(p-c){}}$ and $S=r.p$"
}
{
"Tag": [],
"Problem": "A positive 3-digit integer is divided by a positive 2-digit number, yielding a integer quotent and zero remainder. What is the smallest possible positive integer quotient? (95 State Sprint#7)",
"Solution_1": "[hide]2. 198/99=2[/hide]",
"Solution_2": "[hide]The answer here is 2. You know the numbers are different, so it cannot be one. The integers values are insignificant, as many combinations will give you the correct answer. \n\nFor example:\n \n100/50 = 2\n102/51 = 2\n104/52 = 2\n\nect...[/hide]",
"Solution_3": "[hide]well, since it's looking for the least number, it can't be one cuz one is 3 digit and the other is 2 digit..... so it much be 2[/hide]",
"Solution_4": "[quote=\"solafidefarms\"]A positive 3-digit integer is divided by a positive 2-digit number, yielding a integer quotent and zero remainder. What is the smallest possible positive integer quotient? (95 State Sprint#7)[/quote]\r\n\r\nThere was another problem from a school or chapter level that said :\r\n\r\nA 3-digit integer is divided by a 2-digit number, yielding a integer quotent and zero remainder. What is the smallest possible integer quotient?\r\n\r\nIf memory serves the given answer ended up being incorrect. Answers?"
}
{
"Tag": [
"geometry",
"geometric transformation",
"national olympiad"
],
"Problem": "Can I translate in to English",
"Solution_1": "It'd be great if you translate these problems :)",
"Solution_2": "If you still remember what Mr. Zuming Feng said last year, then do not translate the problems until July.",
"Solution_3": "Shobber, why? Avoiding translation make nothing but distingushs Chinese-speaking people from not-Chinese-speaking, is it a goal?",
"Solution_4": "Read this and you know why. :) \r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?p=476480#476480",
"Solution_5": "Is there a problem-only version? Thanks.",
"Solution_6": "[quote=\"shobber\"]If you still remember what Mr. Zuming Feng said last year, then do not translate the problems until July.[/quote]\r\nNow, can you post all those problems?",
"Solution_7": "can someone translate this problems now?"
}
{
"Tag": [
"probability",
"symmetry"
],
"Problem": "How many 2 digit numbers are there where the tens digit is greater than the ones digit?\r\n\r\nhow would you do this problem?",
"Solution_1": "[hide]If the one's digit is 0, there are 9 possible ten's digits. If the one's digit is 1, there are 8 possible. etc.\n\n9+8+7...+2+1=45\n[/hide]",
"Solution_2": "wow i am impressed.\r\n\r\nI would just list them starting with 10 up that work lol, never thought of it that way.",
"Solution_3": "In general, the number of strictly increasing $n$ digit numbers is $\\dbinom{9}{n}$ and the number of strictly decreasing $n$ digit numbers is $\\dbinom{10}{n}$, with the exception that there are 9 strictly decreasing 1 digit numbers. I sort of derived this, so it's not any formal rule that I know of, but I'm pretty sure it's true.",
"Solution_4": "[hide=\"Alternately\"] The problem is also easily solved with symmetry.\n\n$\\text{Number of 2-digit numbers}= \\text{Number of 2-digit numbers where the tens digit is greater than the ones digit}$ $+\\text{Number of 2-digit numbers where the tens digit is less than the ones digit}$ $+\\text{Number of 2-digit numbers where the tens digit is equal to the ones digit}$\n\n\n(Where we're looking at the numbers $00, 01, 02, ... 99$ as having \"two digits.\" ) \n\nNote that if we reverse a 2-digit number from the first category we get a 2-digit number in the second category, so they are of equal size. Let that size be $x$. Then\n\n$100 = 2x+10$\n$x = 45$\n\nWe know we have not counted any numbers with $0$ as their first digit because they are all in the second or third groups. \n\n\n\nThis kind of reasoning is very instructive for more difficult probability problems where a straightforward summation would be difficult but where a symmetry is easy to take advantage of. \n[/hide]",
"Solution_5": "[quote=\"t0rajir0u\"][hide=\"Alternately\"] The problem is also easily solved with symmetry.\n\n$\\text{Number of 2-digit numbers}= \\text{Number of 2-digit numbers where the tens digit is greater than the ones digit}$ $+\\text{Number of 2-digit numbers where the tens digit is less than the ones digit}$ $+\\text{Number of 2-digit numbers where the tens digit is equal to the ones digit}$\n\n\n(Where we're looking at the numbers $00, 01, 02, ... 99$ as having \"two digits.\" ) \n\nNote that if we reverse a 2-digit number from the first category we get a 2-digit number in the second category, so they are of equal size. Let that size be $x$. Then\n\n$100 = 2x+10$\n$x = 45$\n\nWe know we have not counted any numbers with $0$ as their first digit because they are all in the second or third groups. \n\n\n\nThis kind of reasoning is very instructive for more difficult probability problems where a straightforward summation would be difficult but where a symmetry is easy to take advantage of. \n[/hide][/quote]\r\n\r\nOoo, very nice solution t0rajir0u! :clap2:",
"Solution_6": "[hide]Well for every choice of two digits, there is exactly one working order. So it is just $\\binom{n}{2}$[/hide]",
"Solution_7": "[hide=\"solution\"]We can attack this by casework:\n\n10\n2{1,0}\n3{2,1,0}\n4{3,2,1,0}\n.\n.\n.\n.\n9{8,7,6,5,4,3,2,1,0}\n\nWe can sum those up:\n\n1+2+3+4+5+6+7+8+9=45[/hide]"
}
{
"Tag": [
"inequalities"
],
"Problem": "Let $a,\\,b,\\,c$ are positive numbers satisfying the condition $abc=1,$ prove that: \\[(a+b)(b+c)(c+a)\\ge 2(1+a+b+c)\\]",
"Solution_1": "[hide]$(a+b)(b+c)(c+a)=2abc+\\sum_{cyc}(a^{2}b+a^{2}c)\\ge 2+2\\sum_{cyc}a^{\\frac{4}{3}}b^{\\frac{1}{3}}c^{\\frac{1}{3}}=2(1+a+b+c)$[/hide]",
"Solution_2": "[hide=\"identity usage\"]\nNotice \\[(a+b)(b+c)(c+a)=(a+b+c)(ab+bc+ca)-abc = (a+b+c)(ab+bc+ca)-1 \\]\nNow, \\[(a+b+c)(ab+bc+ca)-1\\ge 2+2(a+b+c)\\iff (a+b+c)(ab+bc+ca-2) \\ge 3 \\]\nBut we have that $\\sum a \\ge 3$ and $\\sum ab \\ge 3$ (by Am-Gm) thus the inequality \\[(a+b+c)(ab+bc+ca-2) \\ge (3)\\cdot(1)=3 \\]\nfollows and is equivalent to the problem.\n[/hide]",
"Solution_3": "[quote=\"SplashD\"]$2abc+\\sum_{cyc}(a^{2}b+a^{2}c)\\ge 2+2\\sum_{cyc}a^{\\frac{4}{3}}b^{\\frac{1}{3}}c^{\\frac{1}{3}}$[/quote]\r\n\r\nHow does that work? I don't think $a^{2}b+a^{2}c \\ge 2a^{\\frac{4}{3}}b^{\\frac{1}{3}}c^{\\frac{1}{3}}$ is true. For example, try $a = \\frac{1}{10}$, $b = 10$, $c = 1$.\r\n\r\nOn the other hand, we can use\r\n\r\n$\\frac{5}{12}a^{2}b+\\frac{5}{12}a^{2}c+\\frac{1}{12}b^{2}c+\\frac{1}{12}c^{2}b \\ge \\sqrt[12]{a^{20}b^{8}c^{8}}= a$,\r\n\r\nbut that might be unnecessarily complicated if I'm just misunderstanding you. It follows from weighted AM-GM.",
"Solution_4": "Well SplashD was just using muirhead... to say that $\\sum_{sym}a^{2}b\\geq \\sum_{sym}a^{\\frac43}b^{\\frac13}c^{\\frac13}$ is just to say that $[2,1,0]\\succ[4/3,1/3,1/3]$ which is clear.\r\nI guess you went ahead and solved for the specific Am-Gm...\r\nbut SplashD wasn't implying any specific inequality by the way he grouped the terms... at least i don't think",
"Solution_5": "Yeah, that makes more sense. But the way he wrote it suggested otherwise, beside the fact that Muirhead only applies for symmetric sums.",
"Solution_6": "Sorry, I made a mistake. I was trying to apply AM-GM\r\n\r\n[hide=\"fixed\"]$(a+b)(b+c)(c+a)=2abc+\\sum_{cyc}(\\frac{2a^{2}b+c^{2}a}{3}+\\frac{2a^{2}c+b^{2}a}{3})\\ge 2+2\\sum_{cyc}a^{\\frac{5}{3}}b^{\\frac{2}{3}}c^{\\frac{2}{3}}=2(1+a+b+c)$[/hide]",
"Solution_7": "here is a related (weaker) result is $(a+b)(b+c)(a+c) \\geq 8$.\r\n\r\nYou can prove this result by expanding to get the sum ${ 2+\\sum_{\\sigma}\\frac{a+b}{c}=2+\\sum_{\\sigma}\\frac{a}{c}+\\frac{b}{c}= 2+\\sum_{\\sigma}\\frac{a}{b}+\\frac{b}{a}\\geq 2+3(2)=8}$.\r\n\r\nThis is weaker since $1+a+b+c \\geq 4$ whenever $abc=1$.\r\n\r\nNow here's a way to prove the original result directly. Expand to get ${ 2+\\frac{a+b}{c}+\\frac{b+c}{a}+\\frac{c+a}{b}}$. Without loss of generality, suppose $a \\geq b \\geq c$. Then $b+c,c+a,a+b$ is a nondecreasing sequence, as is $\\frac{1}{a}, \\frac{1}{b}, \\frac{1}{c}$. Thus by Chebyshev,\r\n\r\n$2+\\frac{a+b}{c}+\\frac{b+c}{a}+\\frac{c+a}{b}\\geq 2+\\frac{1}{3}(2(a+b+c))(\\frac{1}{a}+\\frac{1}{b}+\\frac{1}{c})={1}{3}(2(a+b+c))(ab+bc+ca)$. Now applying AM-GM to $ab+bc+ca$ yields that the original expresion is greater than or equal to $2+2(a+b+c)$, which is what we wanted to show.",
"Solution_8": "[quote=\"drunner2007\"]\nThis is stronger since $1+a+b+c \\leq 4$ whenever $abc=1$.\n[/quote]\r\n\r\nNo, $1+a+b+c \\geq 4$ whenever $abc = 1$. Your result isn't stronger.",
"Solution_9": "[quote=\"Arne\"][quote=\"drunner2007\"]\nThis is stronger since $1+a+b+c \\leq 4$ whenever $abc=1$.\n[/quote]\n\nNo, $1+a+b+c \\geq 4$ whenever $abc = 1$. Your result isn't stronger.[/quote]\r\n\r\nhaha yeah i realized that about an hour later. i should have said weaker result."
}
{
"Tag": [],
"Problem": "For a finite sequence $ A \\equal{} (a_1, a_2, \\ldots, a_n)$ of numbers, the [i]Cesaro sum[/i] of $ A$ is defined to be\r\n\\[ \\frac {S_1 \\plus{} S_2 \\plus{} \\cdots \\plus{} S_n}{n},\r\n\\]\r\nwhere $ S_k \\equal{} a_1 \\plus{} a_2 \\plus{} \\cdots \\plus{} a_k$. If the Cesaro sum of the $ 99$-term sequence $ (a_1,a_2,\\ldots ,a_{99})$ is $ 1000$, what is the Cesaro sum of the $ 100$-term sequence $ (1,a_1,a_2,\\ldots , a_{99})$?",
"Solution_1": "If I'm not misunderstanding the problem\r\nThe answer is $ 991$",
"Solution_2": "[quote=\"IW@IT\"]If I'm not misunderstanding the problem\nThe answer is $ 991$[/quote]\r\n\r\nCould you please explain what you did? Thanks!",
"Solution_3": "EDIT: This is not correct, I have misunderstood your problem....\r\n\r\n\r\nSince\r\n\r\n1+99=100\r\n2+98=100\r\n3+97=100\r\n4+96=100\r\n5+95=100\r\n6+94=100\r\n7+93=100\r\n8+92=100\r\n.\r\n.\r\n.\r\n49+51=100\r\n\r\nand only left number you have is 50 which is unpaired...\r\n\r\nSk=4950\r\n\r\nyou can do it this way too:\r\n\r\nSk=[n*(n+1)]/2\r\n\r\nusing this:\r\n\r\nSk=[99*(99+1)]/2\r\n\r\nSk=99*100/2\r\n\r\nSk=4950\r\n\r\n\r\n\r\nEDIT: I have misunderstood your problem... I will maybe later psot the correct answer...",
"Solution_4": "[hide=\"Solution\"]\n$ S'_1 \\equal{} 1$\n\n$ S'_2 \\equal{} 1 \\plus{} a_1 \\equal{} 1 \\plus{} S_1$\n\n$ S'_3 \\equal{} 1 \\plus{} a_1 \\plus{} a_2 \\equal{} 1 \\plus{} S_2$\n\n$ \\vdots$\n\n$ S'_{100} \\equal{} 1 \\plus{} a_1 \\plus{} a_2 \\plus{} \\cdots \\plus{} a_{99} \\equal{} 1 \\plus{} S_{99}$\n\n$ \\Downarrow$\n\n$ S'_1 \\plus{} S'_2 \\plus{} \\cdots \\plus{} S'_{100} \\equal{} S_1 \\plus{} S_2 \\plus{} \\cdots \\plus{} S_{99} \\plus{} 100$\n\n$ \\frac {S'_1 \\plus{} S'_2 \\plus{} \\cdots \\plus{} S'_{100}}{100} \\equal{} \\frac {1000 \\times 99 \\plus{} 100}{100}$\n\n$ \\equal{} \\boxed{1090}$\n[/hide]",
"Solution_5": "@[b]miyomiyo[/b]\r\nHow did you know $ S_1 \\equal{} 1$\r\n\r\nEDIT: sorry I misread the question",
"Solution_6": "Because 1 is the first term of the sequence, so $ S'_1 \\equal{} 1$.\r\nI suppose you asked for $ S'_1$, because we don't know $ S_1$.",
"Solution_7": "Sorry, I didn't explain my notation. I'm calling $ S'_k$ the $ k^{th}$ sum of the new sequence, while $ S_k$ is the $ k^{th}$ sum of the old sequence.",
"Solution_8": "[quote=\"miyomiyo\"][hide=\"Solution\"]\n$ S'_1 \\equal{} 1$\n\n$ S'_2 \\equal{} 1 \\plus{} a_1 \\equal{} 1 \\plus{} S_1$\n\n$ S'_3 \\equal{} 1 \\plus{} a_1 \\plus{} a_2 \\equal{} 1 \\plus{} S_2$\n\n$ \\vdots$\n\n$ S'_{100} \\equal{} 1 \\plus{} a_1 \\plus{} a_2 \\plus{} \\cdots \\plus{} a_{99} \\equal{} 1 \\plus{} S_{99}$\n\n$ \\Downarrow$\n\n$ S'_1 \\plus{} S'_2 \\plus{} \\cdots \\plus{} S'_{100} \\equal{} S_1 \\plus{} S_2 \\plus{} \\cdots \\plus{} S_{99} \\plus{} 100$\n\n$ \\frac {S'_1 \\plus{} S'_2 \\plus{} \\cdots \\plus{} S'_{100}}{100} \\equal{} \\frac {1000 \\times 99 \\plus{} 100}{100}$\n\n$ \\equal{} \\boxed{1090}$\n[/hide][/quote]\r\nMy approach is exactly the same, but it seems to me that $ \\frac {1000 \\times 99 \\plus{} 100}{100} \\equal{} (99000 \\plus{} 100)/100 \\equal{} 991$.",
"Solution_9": "Haha, I didn't divide the $ 100$. Your answer is correct."
}
{
"Tag": [],
"Problem": "Well they are choosing the U.S. team...slowly. Theres at least 2 maybe 3 Texans on the team this year. A Gymnast from Grand Prarie and a runner from Texas Tech I think. Anyone know if they have chosen the equestrian team yet? If so whose on it. For the U.S. that is.\r\n\r\nFor all you peoples from other coutries Do you know whose representing your country in the games?",
"Solution_1": "Well the only Canadian that I know on our team is one gymnast who was in my class in 7th grade (she was in 8, but we had a split level class). However, she moved away before highschool since her hometown got a proper gymnastics center. The rest I only know by name/reputation...",
"Solution_2": "Well winter Olympiad happened half a year ago!",
"Solution_3": "[quote=mathleticguyyy]Well winter Olympiad happened half a year ago![/quote]\n\nPLEASE STOP ",
"Solution_4": "[quote=mathleticguyyy]Well winter Olympiad happened half a year ago![/quote]\n\n[quote=mikimoto12][quote=mathleticguyyy]Well winter Olympiad happened half a year ago![/quote]\n\nPLEASE STOP[/quote]\n\nTHIS THREAD IS 14 YEARS OLD, wait I'm actually older than it wow",
"Solution_5": "It's kinda sad: this thread os older than me....."
}
{
"Tag": [
"ratio",
"geometric series"
],
"Problem": "The last digit of the sum $ 2^0\\plus{}2^1\\plus{}2^2\\plus{}...\\plus{}2^{2009}$ is:\r\n\r\n$ (A)$ $ 1$\r\n$ (B)$ $ 3$\r\n$ (C)$ $ 5$\r\n$ (D)$ $ 7$\r\n$ (E)$ $ 9$",
"Solution_1": "The given expression is equal to $ \\frac {2^{0}(2^{2010} \\minus{} 1)}{2 \\minus{} 1}$ the last digit of $ 2^{2009}$ is 2 so the answer will be 3 that is $ B$\r\nEDITED",
"Solution_2": "[hide=\"Solution\"]\nThis is a geometric series with first term 1 and common ratio 2. Using the formula for a geometric series, \\begin{align*}2^0+2^1+2^2+\\dots+2^{2009}&=\\frac{2^0\\left(2^{2009+1}-1\\right)}{2-1}\\\\&=2^{2010}-1\\end{align*} is the value of the sum. The last digit of powers of two cycle in fours, so the last digit of $ 2^{2010}$ is the same as $ 2^2=4.$ Finally, subtract 1 from this to get the last digit of $ 3,$ which is $ \\boxed{\\textbf{(B)}}.$[/hide]"
}
{
"Tag": [
"quadratics",
"algebra",
"quadratic formula"
],
"Problem": "\\[ {\\rm{Using the quadratic formula}},{\\rm{ solve the equation}}: a^2 b^2 x^2 \\minus{}\\minus{}{\\rm{ }}\\left( {4b^4 {\\rm{ }}\\minus{}\\minus{}{\\rm{ }}3a^4 } \\right)x{\\rm{ }}\\minus{}\\minus{}{\\rm{ }}12a^2 b^2 {\\rm{ }} \\equal{} {\\rm{ }}0\r\n\\]",
"Solution_1": "[quote=\"sanbittu\"]\n\\[ {\\rm{Using the quadratic formula}},{\\rm{ solve the equation}}: a^2 b^2 x^2 \\minus{} \\minus{} {\\rm{ }}\\left( {4b^4 {\\rm{ }} \\minus{} \\minus{} {\\rm{ }}3a^4 } \\right)x{\\rm{ }} \\minus{} \\minus{} {\\rm{ }}12a^2 b^2 {\\rm{ }} \\equal{} {\\rm{ }}0\n\\]\n[/quote]\r\n\r\nNot quite sure what is going on due to the $ \\minus{} \\minus{}$ which is completely meaningless.\r\n\r\nStill...\r\n\r\nLet\r\n\r\n$ 0 \\equal{} (a^2b^2)x^2 \\plus{} (4b^4 \\plus{} 3a^4)x \\plus{} (12a^2b^2)$\r\nUsing the quadratic formula, taking $ a \\equal{} (a^2b^2), b \\equal{} (4b^4 \\plus{} 3a^4), c \\equal{} (12a^2b^2)$\r\n\r\n$ x \\equal{} \\frac { \\minus{} (4b^4 \\plus{} 3a^4)\\pm \\sqrt {(4b^4 \\plus{} 3a^4)^2 \\minus{} 4(a^2b^2)(12a^2b^2)} }{2(a^2b^2)}$\r\n$ x \\equal{} \\frac { \\minus{} (4b^4 \\plus{} 3a^4)\\pm (4b^4 \\minus{} 3a^4)}{2(a^2b^2)}$\r\n$ x \\equal{} \\frac { \\minus{} 3a^2}{b^2} \\mbox{ or } x \\equal{} \\frac { \\minus{} 4b^4}{a^2}$"
}
{
"Tag": [
"geometry",
"geometric transformation",
"reflection"
],
"Problem": "We have a $2 \\times 8$ billiards table, with a single ball at is center. The ball is sent in a straight line, and after 29 meters it stops at a corner of the table. How many times did the ball bounce on the borders of the table?\r\n\r\nNote: The ball bounces obeying the law of reflection, that is, the angle of incidence between the path of the ball and the border of the table is equal to the angle of reflection.",
"Solution_1": "[hide=\"Solution\"]\nLet the number of times the ball hits a short edge (i.e. it travels $8$ feet) be $a$, and the number of times it hits a long edge (i.e. it travels $2$ feet) be $b$. \n\nThen, by reflections, we are solving the equation\n\\[29^{2}= 16 \\cdot (2a+1)^{2}+(2b+1)^{2}\\] in non-negative integers. We test the few possible values of $a$, and find the only solution to be $(a,b)=(2,10)$. So, not counting the final pocketing of the ball, we have a total of $\\boxed{12 \\text{ bounces off of a wall}}$.[/hide]"
}
{
"Tag": [
"function",
"algebra unsolved",
"algebra",
"functional equation"
],
"Problem": "I have this problem i suspect i know how to solve it but i cant because i dont have a nice series of identities concerning cubes.\r\nfind all functions from the nonegative integers to theirselves such that:\r\n\r\n$f(a^3+b^3+c^3)=f(a)^3+f(b)^3+f(c)^3$ for all $a,b,c$ nonnegative integers.",
"Solution_1": "I seem to remember Pierre solving this one on the board, but i'm not sure at all.",
"Solution_2": "We used it for our correspondance program. You may find a solution (un french, but it doesn't matter here) at :\r\n\r\nhttp://www.animath.fr/tutorat/dossier_03044sol.pdf\r\n\r\nPierre.",
"Solution_3": "I'm not sure if this will help,\r\n$(1^3+2^3+...+n^3)=(1+2+...+n)^2$.",
"Solution_4": "Well...in my opinion, it won't...but who knows...\r\n\r\nPierre.",
"Solution_5": "I have solved this problem before.\r\nThe main lemma is : for each integer number n>=4 we have x,y,z,t,w integer numbers s.t their [b]absolute[/b] values less than n and:\r\n n^3=x^3+y^3+z^3+t^3+w^3\r\n\r\nTo prove this ,we may use this identity: \r\n (2n+1)^3=(2n-1)^3 + (n+4)^3 - (n-4)^3 - 5^3 - 1^3",
"Solution_6": "thats what Pierre did in the link!",
"Solution_7": "oh... :?",
"Solution_8": "Really?Actually,i didn't read pierre's solution. :) \r\nBut the problem i solved was to find f(x): Z -> Z , not f(x): Z+ -> Z+\r\nFor your one i think we should use the following identity:\r\n [b]a^3 + (5a)^3 + (2n+a)^3 =(2n-a)^3 + (n+4a)^3 + (4a-n)^3[/b]\r\n there [b]a,n[/b] are nonnegative integers and [b]a2p^k\r\nlet n be composite with at least 2 prime divisors and let n=ab where gcd(a,b)=1, a>b\r\nthen sigma(n)+phi(n)=sigma(a)sigma(b) +phi(a)phi(b)\r\n>=1/2( sigma(a)+phi(a)) (sigma(b)+phi(b) >=2ab=2n by induction.",
"Solution_2": "Here's what I did here (it might look A LOT more complicated, but once you write it down you'll see it isn't):\r\n\r\nFirst of all, it's easy to see that the sum of all divisors of \r\nn= \\pi {1<=i<=k}pi^ai is (1+p1+..+p1^a1)*(1+p2+..+p2^a2)*..*(1+pk+..+pk^ak)= \\pi {1<=i<=k}(pi^(ai+1)-1)/(pi-1). The inequality is, after subtracting phi(n) from both sides and multiplying both sides with \r\n \\pi {1<=i<=k} (pi-1), \r\n\r\n \\pi {1<=i<=k} (pi^(ai+1)-1) >= \\pi {1<=i<=k} (pi^(ai+1)-pi^(ai-1)), but this is obviously true because pi^(ai+1)-1 >= pi^(ai+1)-pi^(ai-1) for all i (because pi^(ai-1)<=1 for all i)."
}
{
"Tag": [
"function",
"integration",
"calculus",
"calculus computations"
],
"Problem": "Find all infinitely differentiable functions $ f: \\mathbb R\\to\\mathbb R$ such that\r\n\\[ f(x)\\plus{}f'(x)\\equal{}2x\\]\r\nfor all $ x\\in\\mathbb R$.",
"Solution_1": "[hide]Mutliplication by $ e^x$ gives finally $ f(x)\\equal{}2(x\\minus{}1)\\plus{}\\frac{c}{e^x}$[/hide]",
"Solution_2": "$ f(x)\\equal{}c.e^{\\minus{}x}\\plus{}2x\\minus{}2$",
"Solution_3": "$ \\{e^xf(x)\\}'\\equal{}2xe^x\\Longrightarrow e^xf(x)\\equal{}2(x\\minus{}1)e^x\\plus{}C$. $ \\therefore f(x)\\equal{}2(x\\minus{}1)\\plus{}Ce^{\\minus{}x}$.",
"Solution_4": "hello, solving the homogeneous part of the equation we get\r\n$ \\frac{dy}{dx}=-y$\r\n$ \\frac{dy}{y}=-dx$\r\n$ y_h=Ce^{-x}$\r\nFor the inhomogenius part we make the ansatz $ y=a\\cdot x +b$, solving this we get the solution\r\n$ y=C\\cdot e^{-x}-2+2x$\r\nSonnhard.",
"Solution_5": "[hide=\"solution\"]\nas this is a first order linear differential equation .\n$ I.F. \\equal{} e^{x}$\n$ ye^{x} \\equal{} \\int2xe^{x}$\n$ y \\equal{} 2(x \\minus{} 1) \\plus{} Ce^{ \\minus{} x}$[/hide]\r\nthanks",
"Solution_6": "[hide=\"Solution:\"]$ \\{f(x) \\plus{} f'(x)\\}e^x \\equal{} 2xe^x \\\\\n \\\\\n\\frac {d}{dx}\\{e^xf(x)\\} \\equal{} 2xe^x \\\\\n \\\\\ne^xf(x) \\equal{} \\int 2xe^xdx \\\\\n \\\\\ne^xf(x) \\equal{} 2xe^x \\minus{} 2e^x \\plus{} 2c \\\\\n \\\\\n\\;\\boxed{f(x) \\equal{} 2\\left(x \\minus{} 1 \\plus{} \\frac{c}{e^x}\\right)}$[/hide]",
"Solution_7": "y+y'=2x\r\ne^xy'+e^xy=2xe^x\r\nd(e^xy)/dx=2xe^x\r\n\r\nWhat so intersting in this simple question of yours?\r\n :oops:"
}
{
"Tag": [
"quadratics",
"number theory unsolved",
"number theory"
],
"Problem": "The digits of a number $n$ are 0,1,2,3,4,5,6,8,9 in some order. Prove that $n$ is not a square.",
"Solution_1": "$38 \\equiv 2 \\bmod 9$ is not a quadratic residue."
}
{
"Tag": [
"algebra",
"polynomial"
],
"Problem": "Prove that there exists an infinity of irrational numbers $x,y$ such that the number $x+y=xy$ is a nonnegative integer.",
"Solution_1": "We consider the polynomial $P(z)=z^2-mz+m$, where $m \\in \\mathbb Z$. We have $\\Delta = \\sqrt{m(m-4)}$. Take any $m \\equiv 0 \\left( \\bmod ~ 8 \\right)$, $m > 0$ $\\displaystyle \\left( \\Delta = 4 \\sqrt{2k(2k-1)} \\, \\, \\textrm{for some } \\, \\, k \\right)$. Then the roots of $P$ are the numbers we are looking for.\r\n\r\nRight?",
"Solution_2": "Of course :)",
"Solution_3": "$xy-x-y+1\\in \\mathbb{N}$, $(x-1)(y-1)\\in \\mathbb{N}$, $x=a+\\sqrt{b}+1$, $y=a-\\sqrt{b}+1$, $a,b \\in \\mathbb{N}$ :D",
"Solution_4": "[quote=\"JohnConstantine\"]$xy-x-y+1\\in \\mathbb{N}$, $(x-1)(y-1)\\in \\mathbb{N}$, $x=a+\\sqrt{b}+1$, $y=a-\\sqrt{b}+1$, $a,b \\in \\mathbb{N}$ :D[/quote]\r\n\r\nDid you forget about the condition $xy = x + y$?",
"Solution_5": "[quote=\"Valentin Vornicu\"]Prove that there exists an infinity of irrational numbers $x,y$ such that the number $x+y=xy$ is a nonnegative integer.[/quote]\r\n\r\n $\\large{x+y\\,\\in\\,N\\,\\,\\Longleftrightarrow{\\,\\,x\\,=\\,a_1+\\sqrt{b}\\,\\,\\,\\textrm{and}\\,\\,\\,y\\,=\\,a_2-\\sqrt{b}\\,\\,\\,|\\,a_i\\,\\in{\\mathbb{Z}}\\,\\,\\textrm{and there is no}\\,\\,\\exists{c}}\\,\\in{\\mathbb{Z}\\,\\,|\\,b\\,=\\,c^2}}$\r\n\r\nAnd :\r\n\r\n$\\large{(a_1+\\sqrt{b})(a_2-\\sqrt{b})\\,=\\,a_1a_2-b+\\sqrt{b}(a_2-a_1)\\,\\in\\,\\,\\mathbb{N}\\,\\,\\Longleftrightarrow{\\,a_1\\,=\\,a_2}}$\r\n\r\nWe have :\r\n\r\n$\\large{x\\,=\\,a+\\sqrt{b}\\,,\\,y\\,=\\,a-\\sqrt{b}}$\r\n\r\nThen :\r\n\r\n$\\large{2a\\,=\\,a^2-b\\,\\Longrightarrow{b\\,=\\,a^2+2a}}$\r\n\r\nWe FINALY have :\r\n\r\n$\\large{x\\,=\\,a+\\sqrt{a^2+2a}\\,=\\,a+\\sqrt{(a+1)^2-1}\\,\\,,\\,\\,y\\,=\\,a-\\sqrt{(a+1)^2-1}}$\r\n\r\nPS.: See that $\\large{(a+1)^1-1}$ CANNOT be a square perfect for $\\large{a>0}$\r\n\r\n$\\large{\\mathbb{C.Q.D.}}$"
}
{
"Tag": [
"induction",
"LaTeX"
],
"Problem": "pl help...........\r\n\r\n\r\n1)2^1001 mod 1001\r\n\r\n\r\n\r\n\r\n\r\n\r\n2)Another one.............................\r\n\r\n\r\nIf one is given a balance, and n weights of 1,3,3^2,...,3^(n-1)! \r\nlb. respectively, show that it is possible, by placing some \r\nweights in one pan and some in the other, to weigh out any \r\nweight of N lb., with n an integer >= 1 and <= 1/2 (3^n- 1)",
"Solution_1": "hello, to your first question, we have $ 2^{10} \\equiv 23 \\mod 1001$, from here we get $ 2^{100}\\equiv23^{10}\\equiv562\\mod1001$ and further $ 2^{1000}\\equiv562^{10}\\equiv 562\\mod 1001$ so we get as the result $ 2^{1001}\\equiv123\\mod 1001$.\r\nSonnhard.",
"Solution_2": "i wont post the complete solutions...\r\nwill just give u hints... \r\n\r\n1.\r\n2^6 == 1 (mod 7)\r\n2^12 == 1 (mod 13)\r\n2^10 == 1 (mod 11)\r\n\r\n( == stands for.. is congruent to)\r\n\r\n2.\r\nwe see that for 1, 3\r\nwe can form 1, 2, 3, 4;\r\ni.e. it is true for n = 2; (and n = 1)\r\n\r\nrest is simple induction\r\nassuming it true for n\r\n=> we can form >= 1 and <= 1/2 (3^n- 1)\r\n\r\nnow we have a new weight 3^n\r\nwe could already form >= 1 and <= (1/2) (3^n- 1) [without using 3^n]\r\n\r\nwe can form [3^n - (1/2) (3^n- 1)] upto 3^n \r\nby keeping 3^n on one side, and putting k on the other one \r\nk => 0 and <= (1/2) (3^n- 1)\r\n\r\nand [3^n + 1] upto (1/2)(3^(n+1) -1 ) \r\nby keeping 3^n on one side, and putting k on the same side \r\nk => 1 and <= (1/2) (3^n- 1)\r\n\r\nkeeping k on one side... means that we know the arrangement required to weigh k\r\nit does not mean that all smaller weights be kept on the same side\r\n\r\nsorry for not using latex\r\n(i dont kno how to)\r\n\r\n\r\n\r\nand Sonnhard..\r\nwe are not allowed calculators in competitions\r\nso it would not be so nice to calculate 562^10\r\n:)",
"Solution_3": "hello, but you can write $ 562^{10}\\equal{}(562^2)^5\\equiv529^5\\equiv562\\mod 1001$.\r\nSonnhard.",
"Solution_4": "how can the first ques be solved by using crt-chinese remainder theorem......???",
"Solution_5": "[quote=\"Arnav...!!!\"]how can the first ques be solved by using crt-chinese remainder theorem......???[/quote]\r\n\r\nI Hope this Helps !!"
}
{
"Tag": [
"Putnam",
"logarithms",
"integration",
"probability",
"analytic geometry",
"college contests"
],
"Problem": "For each positive integer $ n,$ let $ f(n)$ be the number of ways to make $ n!$ cents using an unordered collection of coins, each worth $ k!$ cents for some $ k,\\ 1\\le k\\le n.$ Prove that for some constant $ C,$ independent of $ n,$\r\n\\[ n^{n^2/2\\minus{}Cn}e^{\\minus{}n^2/4}\\le f(n)\\le n^{n^2/2\\plus{}Cn}e^{\\minus{}n^2/4}.\\]",
"Solution_1": "[hide]\nFirst, we find an approximation for $ \\prod_{i\\equal{}1}^n i^i$ which will figure in both parts of the inequality.\nLet $ g(x) \\equal{} \\frac{x^2}2 \\ln x \\minus{} \\frac{x^2}4.$ Then $ g'(x) \\equal{} \\frac{x}2 \\plus{} x \\ln x \\minus{} \\frac{x}2 \\equal{} x \\ln x.$ Because $ x \\ln x$ increases on $ [1,\\infty),$ \n\\[ \\sum_{i\\equal{}1}^n i \\ln i \\le \\sum_{i\\equal{}1}^n \\int_i^{i\\plus{}1} x\\ln x dx \\equal{}\\int_1^{n\\plus{}1} x\\ln x dx \\equal{} g(n \\plus{} 1) \\plus{} \\frac14\n\\]\nand\n\\[ \\sum_{i\\equal{}1}^n i \\ln i \\ge \\sum_{i\\equal{}2}^n \\int_{i\\minus{}1}^i x\\ln x dx \\ge \\int_1^n x\\ln x dx\\ge g(n).\n\\]\nThis implies\n\\[ e^{g(n)} \\le \\prod_{i \\equal{} 1}^n i^i \\le e^{g(n\\plus{}1) \\plus{} \\frac14}.\n\\]\n\nObserve that the quantity $ f(n)$ we need to bound is the number of sequences $ (k_i)_{i\\equal{}1}^n$ of nonnegative integers satisfying $ \\sum_{i\\equal{}1}^n k_i i! \\equal{} n!.$ \n\nWe know instantly that $ k_i \\le \\frac{n!}{i!}.$ Therefore, $ k_i$ can take at most $ \\frac{n!}{i!} \\plus{} 1 \\le 2\\frac{n!}{i!}$ different values. Thus,\n\\[ f(n) \\le \\prod_{i\\equal{}1}^n 2\\frac{n!}{i!} \\equal{} 2^n \\prod_{i\\equal{}1}^n i^{i \\minus{} 1} \\le 2^n \\prod_{i \\equal{} 1}^n i^i \\le 2^n e^{g(n\\plus{}1)\\plus{} \\frac14}.\n\\]\nWe have\n\\[ e^{g(n\\plus{}1) \\minus{} g(n)} \\le \\frac{(n\\plus{}1)^{\\frac{(n\\plus{}1)^2}2}}{n^{\\frac{n^2}2}} \\equal{} \\left(1 \\plus{} \\frac1n\\right)^{\\frac{n^2}2} (n\\plus{}1)^{n \\plus{} \\frac12} \\le e^{\\frac{n}2} (2n)^{2n} \\le n^{C_1n}\n\\]\nfor some constant $ C_1.$ Therefore, $ f(n) \\le e^{g(n)} n^{C n} \\equal{} n^{\\frac{n^2}2 \\plus{} C n} e^{\\minus{}\\frac{n^2}4}$ for some constant $ C.$ \n\nLet us set $ k_n \\equal{} 0$ and pick $ k_i \\in \\left[1, \\frac{(n\\minus{}1)!}{i!}\\right]$ for $ i \\equal{} 2,3,\\ldots, n\\minus{}1.$ Then \n\\[ \\sum_{i\\equal{}2}^n k_i i! \\le (n \\minus{} 1)(n \\minus{} 1)! < n!.\n\\]\nTherefore, for every such choice we can always find $ k_1$ so that $ \\sum_{i\\equal{}1}^n k_i i! \\equal{} n!.$ This yields\n\\[ f(n) \\ge \\prod_{i \\equal{} 2}^{n\\minus{}1} \\frac{(n\\minus{}1)!}{i!} \\equal{} \\prod_{i \\equal{} 3}^{n \\minus{} 1} i^{i \\minus{}2} \\equal{} (n!)^{\\minus{}2} n^{\\minus{}n} \\prod_{i \\equal{} 1}^n i^i \\ge e^{g(n)} n^{\\minus{}3n} \\equal{} n^{\\frac{n^2}2 \\minus{} 3n} e^{\\minus{}\\frac{n^2}4}.\n\\]\n[/hide]",
"Solution_2": "Since we have a solution to the standards of the problem already up, here's a much tighter estimate:\r\n\r\nFor reasonably large $ n$, $ f(n)\\approx \\frac1{(n\\minus{}2)!}\\frac{(n!)^{n\\minus{}2}}{2!3!\\cdots (n\\minus{}1)!}$.\r\nIn terms of exponentials and powers of $ n$, this is asymptotically close to $ A\\cdot n^{\\frac{(n\\minus{}1)(n\\minus{}2)}{2}\\plus{}\\frac1{12}}e^{\\minus{}\\frac{n^2}{4}\\plus{}2n}$ for some constant $ A$ as $ n\\to\\infty$.\r\n\r\nThere are $ n\\minus{}1$ ways to make $ n!$ using only one kind of coin (other than $ 1$ cent). We subtract those, and ignore them due to insignificant size.\r\nFor $ k\\equal{}2,3,\\dots,n\\minus{}1$, we choose between $ 0$ and $ \\frac{n!}{k!}\\minus{}1$ coins of denomination $ k!$ uniformly at random. If the sum of the values of the coins chosen this way is at most $ n!$, there is exactly one way to choose 1-cent coins to complete our pile to $ n!$; otherwise, there is no way. The total number of ways is thus $ \\frac{(n!)^{n\\minus{}2}}{2!3!\\cdots (n\\minus{}1)!}\\cdot p\\plus{}n\\minus{}1$, where $ p$ is the probability that the sum is at most $ n!$.\r\n\r\nTo estimate $ p$, we approximate the discrete uniform distributions of coin number by continuous uniform distributions; $ p$ is approximately the probability that a sum of $ n\\minus{}2$ independent uniform distributions on $ [0,1]$ is at most $ 1$. This probability is the volume in $ \\mathbb{R}^{n\\minus{}2}$ of the simplex bounded by the coordinate hyperplanes and the hyperplane $ x_1\\plus{}x_2\\plus{}\\dots\\plus{}x_{n\\minus{}2}$, which is $ \\frac1{(n\\plus{}2)!}$",
"Solution_3": "lol, this question isn't even true for $ n\\equal{}1$",
"Solution_4": "You're right- powers of 1 don't give a lot of room to maneuver.\r\nNote that my estimate only makes sense for $ n\\ge 2$.",
"Solution_5": "I have a question. In the solutions next to B6 it says (suggested by Oleg Goldberg). Does this mean that he suggested the problem or suggested the particular solution? I only ask because all the other problems have \"solution by\" next to them or something of the sort while this one doesn't.",
"Solution_6": "Seeing that Oleg Golberg was a contestant, it was probably the solution that he proposed.",
"Solution_7": "Note that this problem is very similar to [url=http://www.mathlinks.ro/viewtopic.php?p=356713&sid=6eca5a8d6067b2cf1e0fde92894916ff#356713]IMO 1997.6[/url]. The only difference is that there the coins are size $ 2^n$ instead of $ n!$. I believe the same proofs (with some modifications) still work though."
}
{
"Tag": [
"function",
"Asymptote"
],
"Problem": "Is there a prebuilt funtion that allows you to put ticks in without manually drawing the ticks? Thanks.",
"Solution_1": "[quote=\"dragon96\"]Is there a prebuilt funtion that allows you to put ticks in without manually drawing the ticks? Thanks.[/quote]\r\nFor graphs you can take a look to [url]http://piprim.tuxfamily.org/asymptote/graph/index.html[/url].\r\nIf you want markers on paths only, [url=http://piprim.tuxfamily.org/asymptote/odoc/index.html#fig0126]here an example[/url] (figure 0126)."
}
{
"Tag": [
"linear algebra",
"linear algebra theorems"
],
"Problem": "Doe's anyone know what formula represents the total number of gifts received in the song \" The 12 days of Christmas\" I know the total number of gifts is 364 but I would like to know how to derive it from a formula. :?",
"Solution_1": "\\[ \\sum_{k=1}^N\\sum_{i=1}^{k}i=\\frac{N(N+1)(N+2)}{6}, \\]\r\nso \r\n\\[ \\sum_{k=1}^{12}\\sum_{i=1}^{k}i=\\frac{12\\times 13\\times 14}{6}=364. \\]\r\n\r\nThis can be proved using two sum formulas:\r\n\\[ \\sum_{i=1}^{n}i=\\frac{n(n+1)}{2} \\]\r\nand\r\n\\[ \\sum_{i=1}^{n}i^2=\\frac{n(n+1)(2n+1)}{6}. \\]",
"Solution_2": "Thankyou so much for your help. We were able to derive part of the formula but not all of it."
}
{
"Tag": [
"algebra unsolved",
"algebra"
],
"Problem": "Find all positive real solutions to the following system of inequations\r\n\r\n$ (x_1^2\\minus{}x_3x_5).(x_2^2\\minus{}x_3x_5)\\le0$\r\n\r\n$ (x_2^2\\minus{}x_4x_1).(x_3^2\\minus{}x_4x_1)\\le0$\r\n\r\n$ (x_3^2\\minus{}x_5x_2).(x_4^2\\minus{}x_5x_2)\\le0$\r\n\r\n$ (x_4^2\\minus{}x_1x_3).(x_5^2\\minus{}x_1x_3)\\le0$\r\n\r\n$ (x_5^2\\minus{}x_2x_4).(x_1^2\\minus{}x_2x_4)\\le0$",
"Solution_1": "Sorry, I don't know how to type some TEX. You can see my solution in the attached file."
}
{
"Tag": [
"modular arithmetic",
"number theory unsolved",
"number theory"
],
"Problem": "Find all solutions if any so that\r\n\r\n$ q,r,p$ odd primes\r\n\r\n$ qr \\equiv 2 \\mod p^2$\r\n$ (q\\minus{}1)(r\\minus{}1) \\equiv 2 \\mod p$\r\n\r\nholds simultaneously",
"Solution_1": "There are way too many solutions.\r\nFor example, this is just one infinite series of solutions:\r\n\\[ \\begin{cases}\r\np\\equal{}11\\\\\r\nq\\equiv 106\\pmod{121}\\\\\r\nr\\equiv 16\\pmod{121}\r\n\\end{cases}\\]\r\n\r\nSuch series exist for every prime $ p$ such that $ p\\equiv 1,\\;9,\\;11,\\;15,\\;23,$ or $ 25\\pmod{28}$."
}
{
"Tag": [
"function",
"search",
"algebra unsolved",
"algebra"
],
"Problem": "Let $ D$ a line in the complex plane . Find all functions $ f: \\mathbb{C} \\to \\mathbb{C}$ with the propriety that $ \\forall \\text{ distinct } z_1,z_2 \\in \\mathbb{C}$ , the point $ z \\equal{} \\frac{f(z_1)\\minus{}f(z_2)}{z_1\\minus{}z_2} \\in \\mathbb{D}$ .",
"Solution_1": "$ \\mathbb{D}\\equal{}a\\mathbb{R}\\plus{}b$ for some $ a,b\\in\\mathbb{C},a\\not\\equal{}0$, and we find $ g: \\mathbb{C}\\to \\mathbb{R}$ with\r\n$ {f(x)\\minus{}f(0)\\over x}\\equal{}ag(x)\\plus{}b$ for all $ x\\in\\mathbb{C}\\setminus\\{0\\}$.\r\nThen for $ c: \\equal{}f(0)$ we have $ f(x)\\equal{}axg(x)\\plus{}bx\\plus{}c$ for all $ x\\in\\mathbb{C}$,\r\nwhich gives $ {f(x)\\minus{}f(y)\\over x\\minus{}y}\\equal{}a{xg(x)\\minus{}yg(y)\\over x\\minus{}y}\\plus{}b\\in \\mathbb{D}\\equal{}a\\mathbb{R}\\plus{}b$\r\nand therefore $ {xg(x)\\minus{}yg(y)\\over x\\minus{}y}\\in\\mathbb{R}$ for all $ x,y\\in\\mathbb{C},x\\not\\equal{}y$.\r\n\r\nBut then for $ x\\in\\mathbb{C}\\setminus\\mathbb{R}$ and $ y\\in\\mathbb{R}\\setminus\\{0\\}$ we have\r\n$ {xg(x)\\minus{}yg(y)\\over x\\minus{}y}\\equal{}g(x)\\plus{}(g(x)\\minus{}g(y)){y\\over x\\minus{}y}\\in\\mathbb{R}$\r\nwith $ g(x)\\in\\mathbb{R}$ and $ g(x)\\minus{}g(y)\\in\\mathbb{R}$ and $ {y\\over x\\minus{}y}\\not\\in\\mathbb{R}$.\r\nSo $ g(x)\\equal{}g(y)$ for all $ x\\in\\mathbb{C}\\setminus\\mathbb{R}$ and $ y\\in\\mathbb{R}\\setminus\\{0\\}$,\r\nwhich gives easily $ g(x)\\equiv g(1)$ constant for all $ x\\in\\mathbb{C}\\setminus\\{0\\}$\r\nand then $ f(x)\\equal{}(ag(1)\\plus{}b)x\\plus{}c$ for all $ x\\in\\mathbb{C}$.\r\n\r\nSo $ f$ is linear and $ {f(x)\\minus{}f(y)\\over x\\minus{}y}$ is constant.",
"Solution_2": "Thank you for your solution [b] olorin [/b] . Please , can you tell me how is this representation (of a line in a complex plane) called , so I can search more about that ? ( I didn't know this reprezentation :blush: ) .\r\n\r\n[quote=\"olorin\"]$ \\mathbb{D} \\equal{} a\\mathbb{R} \\plus{} b$ for some $ a,b\\in\\mathbb{C},a\\not \\equal{} 0$ [/quote]\r\n\r\nEdit : Oh silly me ... :blush: I didn't notice that it derives from the line equation :ewpu:",
"Solution_3": "How else would you define a line?\r\nIt's the formal definition [url=http://en.wikipedia.org/wiki/Line]here[/url].\r\nI simply identify $ \\mathbb{C}\\equal{}\\mathbb{R}^2$."
}
{
"Tag": [],
"Problem": "Has anyone tried to solve the sample problems posted on the AHSIMC website (http://www.ahsimc.com)? How does everyone feel about discussing them and sharing solutions?",
"Solution_1": "I've tried to solve them, and I'd welcome a discussion"
}
{
"Tag": [
"geometry"
],
"Problem": "Find static distributions of n charges (n=4 and n=5) such that each charge is in equilibrium. \r\n (geometry, lengths, and magnitudes)",
"Solution_1": "do u mean static charges and stable equilibrium then it is not possible see [url=http://en.wikipedia.org/wiki/Earnshaw's_theorem]here[/url]",
"Solution_2": "Probably my question is not clear enough. \"static\" just wants to mention that the system is not moving. And \"equilibrium\" means the net Column's force on each charge is 0.",
"Solution_3": "not possible",
"Solution_4": "would like to give a thought on it if i have time if u meant unstable equilibrium for stable equilibrium that's impossible see earnshaw's theorem",
"Solution_5": "take q on three vertices of equilateral triangle.\r\n\r\n\r\nand -q/sqrt(3) at the centre.\r\n\r\nfor n=4 we are done...\r\n\r\nbut its unstable equilibrium...\r\n\r\nfor n=5,\r\ntake q on 4 sides of a square,\r\n-q/sqrt(2) at the centre...\r\n\r\nbut unstable equilibrium..",
"Solution_6": "i did a small mistake for n=5,\r\n\r\ncorrection:\r\nthe central charge should be\r\n\r\n-q(1/sqrt(2) + 1/4)\r\n\r\nI hope this is the solution you were looking for."
}
{
"Tag": [
"calculus",
"integration",
"algebra unsolved",
"algebra"
],
"Problem": "Find the greatest positive integer $n$ such that the systems of equations \r\n$(x+1)^2+y_1^2 = (x+2)^2+y_2^2 = ... = (x+n)^2+y_n^2$\r\nhas an integral solution $(x,y_1,y_2,...,y_n)$",
"Solution_1": "This is actually a Vietnamese problem, I believe. Very nice problem by the way.\r\n\r\nWorking modulo some small number should help to get some restrictions on n.",
"Solution_2": "It`s hard. Can somebody help",
"Solution_3": "Look mod 3.",
"Solution_4": "for n=3,it's obvious that (x,y1,y2,y3)=(2,4,3,0) is suitable.\r\nif n>=4,let S=(x+1)*(x+1)+y1*y1=.......\r\nwe know that m*m=0,4,1(mod8),so:\r\n(1)x=0(mod4),if y1 is odd,then S=2(mod8),but(x+2)*(x+2)=4(mod8),so y2*y2=6(mod8),which comes to a contrariety.if y1=0(mod4),then S=1(mod8),but(x+2)*(x+2)=4(mod8),so y2*y2=5(mod8),which comes to a contrariety.if y1=2(mod4),then S=5(mod8),but(x+4)*(x+4)=0(mod8),so y4*y4=5(mod8),which comes to a contrariety.\r\n(2)x=2(mod4),if y1 is odd,then S=2(mod8),but(x+2)*(x+2)=0(mod8),so y2*y2=2(mod8),which comes to a contrariety.if y1=0(mod4),then S=1(mod8),but(x+4)*(x+4)=4(mod8),so y4*y4=5(mod8),which comes to a contrariety.if y1=2(mod4),then S=5(mod8),but(x+2)*(x+2)=0(mod8),so y2*y2=5(mod8),which comes to a contrariety.\r\n(3)x=1(mod4),if y1 is odd,then S=5(mod8),but(x+3)*(x+3)=0(mod8),so y3*y3=5(mod8),which comes to a contrariety.if y1=0(mod4),then S=4(mod8),but(x+4)*(x+4)=1(mod8),so y4*y4=3(mod8),which comes to a contrariety.if y1=2(mod4),then S=0(mod8),but(x+2)*(x+2)=1(mod8),so y2*y2=7(mod8),which comes to a contrariety.",
"Solution_5": "(4)x=3(mod4),if y1 is odd,then S=1(mod8),but(x+3)*(x+3)=4(mod8),so y3*y3=7(mod8),which comes to a contrariety.if y1=0(mod4),then S=0(mod8),but(x+2)*(x+2)=1(mod8),so y2*y2=7(mod8),which comes to a contrariety.if y1=2(mod4),then S=4(mod8),but(x+2)*(x+2)=1(mod8),so y2*y2=3(mod8),which comes to a contrariety.\r\nso,n=3 is the greatest :arrow:",
"Solution_6": "Hey,please use Latex."
}
{
"Tag": [
"geometry"
],
"Problem": "x, y, & z are consecutive integers. prove that x+z is divisible by y.",
"Solution_1": "Getting started is not a place for proofs, even though this one indeed is really easy.\r\n[hide]\n$\\frac{x+z}2=y$\n[/hide]",
"Solution_2": "[hide]\n$x+z=2y$\n$\\frac{x+z}{y}=n$\n$n=2$\n${n}\\subseteq{\\mathbb{Z}}$[/hide]\r\n\r\nnvm my proof sucks.",
"Solution_3": "[quote=\"deej21\"]Getting started is not a place for proofs, even though this one indeed is really easy.\n[hide]\n$\\frac{x+z}2=y$\n[/hide][/quote]This proof problem is very elementary and so it is OK. About proofs, this is as far as it can go in Getting Started as set by MCrawford (admin) in the guidelines.",
"Solution_4": "I knew it was too easy :) . I'll post something harder next time.",
"Solution_5": "[quote=\"Inspired By Nature\"]I knew it was too easy :) . I'll post something harder next time.[/quote]No, I don't think you understood what I was saying. If you want to post harder next time, you should post them in Intermediate Topics forum. Only \"extremely elementary\" proofs are allowed in Getting Started, like the one above.",
"Solution_6": "$x+z=2x+2=2(x+1)=2y$\r\n\r\nif you mean that\r\n$x0 prove if the following inequality hold true\r\n\r\n a/(b+c)+b/(c+a)+c/(a+b)>=(a+b+c)/2 :3sqrt:(abc).\r\n If this is not true prove the contrary. I hope you do understand what i have written.",
"Solution_1": "Do you mean:\r\nIf $a,b,c>0$,\r\n\\[\\frac{a}{b+c}+\\frac{b}{c+a}+\\frac{c}{a+b} \\ge \\frac{(a+b+c)}{2\\sqrt[3]{abc}}.\\]",
"Solution_2": "Expanding the symsums gives something that I don't believe is true.\r\n\r\nAnd indeed, my guess is correct; take $a=1.1,b=1,c=0.9$",
"Solution_3": "yes that's it.But can anyone give me a proof?",
"Solution_4": "It's not true, I provided a counterexample above!"
}
{
"Tag": [
"Inequality",
"polynomial",
"algebra",
"n-variable inequality"
],
"Problem": "Prove the following inequality:\r\n\\[\\prod^k_{i=1} x_i \\cdot \\sum^k_{i=1} x^{n-1}_i \\leq \\sum^k_{i=1}\r\nx^{n+k-1}_i,\\] where $x_i > 0,$ $k \\in \\mathbb{N}, n \\in\r\n\\mathbb{N}.$",
"Solution_1": "Please post your solutions. This is just a solution template to write up your solutions in a nice way and formatted in LaTeX. But maybe your solution is so well written that this is not required finally. For more information and instructions regarding the ISL/ILL problems please look here: [url=http://www.mathlinks.ro/Forum/viewtopic.php?t=15623]introduction for the IMO ShortList/LongList project[/url] and regarding[url=http://www.mathlinks.ro/Forum/viewtopic.php?t=15623]solutions[/url] :)",
"Solution_2": "[quote=\"orl\"]Prove the following inequality:\n\\[\\prod^k_{i=1} x_i \\cdot \\sum^k_{i=1} x^{n-1}_i \\leq \\sum^k_{i=1}\nx^{n+k-1}_i,\\] where $x_i > 0,$ $k \\in \\mathbb{N}, n \\in\n\\mathbb{N}.$[/quote]\r\n\r\nThis is a direct result of Muirhead's inequality; $(n+k-1,0,0,...0)$ majorizes $(n-1,1,1,...1)$. Alternatively, it follows from the AM-GM by the symmetric sum of $nx_1^{n+k-1}+x_2^{n+k-1}+...+x_k^{n+k-1} \\geq (n+k-1)x_1^{n}x_2...x_k$.",
"Solution_3": "I think you mean [tex] (n+k-1,0,0,...,0)\\succ(n,1,1,...,1) [/tex]",
"Solution_4": "Or we can use Chebishev Inequality:\r\n\r\nfor sequences $ \\left\\{a_{n}\\right\\}$ and $ \\left\\{b_{n}\\right\\}$ that both are increasing or both decreasing, and real numbers $ \\mu_{i}\\;\\;(1\\leq i\\leq n)$ we have:\r\n\r\n$ \\boxed{\\sum^{n}_{i \\equal{} 1}\\mu_{i}a_{i}\\sum^{n}_{i \\equal{} 1}\\mu_{i}b_{i}\\leq\\sum^{n}_{i \\equal{} 1}a_{i}b_{i}\\sum^{n}_{i \\equal{} 1}\\mu_{i}}$\r\n\r\nnow we have:\r\n\r\n$ \\sum^{k}_{i \\equal{} 1}(1\\cdot x^{k}_{i})\\sum^{k}_{i \\equal{} 1}(1\\cdot x^{n\\minus{}1}_{i})\\leq\\sum^{k}_{i \\equal{} 1}(x^{k}_{i}\\cdot x^{n\\minus{}1}_{i})\\sum^{k}_{i \\equal{} 1}1$\r\n\r\n$ \\Longrightarrow\\frac{1}{k}\\sum^{k}_{i \\equal{} 1}x^{k}_{i}\\cdot\\sum^{k}_{i \\equal{} 1}x^{n\\minus{}1}_{i}\\leq\\sum^{k}_{i \\equal{} 1}x^{n\\plus{}k\\minus{}1}_{i}$\r\n\r\nby AM-GM : $ \\prod^{k}_{i \\equal{} 1}x_{i}\\leq\\frac{1}{k}\\sum^{k}_{i \\equal{} 1}x^{k}_{i}$\r\n\r\nand the desired inequality is obtained!"
}
{
"Tag": [
"limit",
"real analysis",
"real analysis unsolved"
],
"Problem": "Hi, what's your solution to this problem: Show that, by applying Taylor Theorem, \r\n\r\n\\[ \\lim_{k\\rightarrow \\infty}k(e\\minus{}\\sum_{n\\minus{}0}^{k\\minus{}1}\\frac{1}{n!})\\equal{}0.\\]\r\n\r\nI'd like to hear about your solution. \r\n\r\nThank you.",
"Solution_1": "you just need to write the definition of $ e$ and note that $ \\sum_{j \\equal{} k}^{\\infty}\\frac{1}{j!} \\leq \\frac {const \\equal{} e}{k!}$ and $ \\lim\\frac {k}{k!} \\equal{} 0$",
"Solution_2": "Thank you."
}
{
"Tag": [
"geometry",
"ratio",
"incenter",
"trigonometry",
"geometry solved"
],
"Problem": "Let ABC be the triangle . (K) is the incircle of ABC.\r\n let B1,C1 respectively be the midpoints of AC,AB. C1K cuts AC at B2\r\nB1K cuts AB at C2. Suppose S(AB2C2)=S(ABC). find ^ACB",
"Solution_1": "I'm a bit confused. By ^ACB do you mean angle A of the triangle ABC? I'm asking because in Romania we have the habit of writing angle A as AC2/AB=AC/AB2. Now play with these ratios a bit. I can't work it all out right now, but theorems like Menelaus and the transversal theorem work here. \r\n\r\nWe can calculate the ratio AC/AB2 by applying Menelaus in the triangle ABC with transversal line C1B2, and then we can find AC2/AB by applying the same thm in the triangle ABC with transversal line B1C2. Just see what you get.",
"Solution_3": "[quote=\"A1lqdSchool\"]Let ABC be the triangle . (K) is the incircle of ABC.\n let B1,C1 respectively be the midpoints of AC,AB. C1K cuts AC at B2\nB1K cuts AB at C2. Suppose S(AB2C2)=S(ABC). find ^ACB[/quote]\r\n\r\nAs Grobber correctly remarked, the angle which has to be found should be < CAB, not < ACB.\r\n\r\nI have not much time now for writing solutions for MathLinks, but as for the above problem, I was asked some days ago by a Hyacinthian to solve it, so I had written down a solution, and now I am just copying this solution (the notations are slightly changed: in my solution, the incenter of triangle ABC is called I rather than K):\r\n\r\n[b]Extended Problem Statement:[/b] In the following, let [XYZ] denote the area of an arbitrary triangle XYZ.\r\n\r\nLet I be the incenter of a triangle ABC. Let $B_1$ and $C_1$ be the midpoints of the sides CA and AB, respectively. Let $B_2$ be the point of intersection of the lines $IC_1$ and CA, and let $C_2$ be the point of intersection of the lines $IB_1$ and AB. Prove that:\r\n\r\n[b](a)[/b] We have $[AB_2C_2]=[ABC]$ if and only if A = 60\u00b0.\r\n[b](b)[/b] We have $[AB_2C_2]=\\frac{[ABC]}{2}$ if and only if A = 90\u00b0.\r\n[b](c)[/b] We have $[AB_2C_2]=\\frac{[ABC]}{3}$ if and only if A = 120\u00b0.\r\n\r\n[b]Solution:[/b] All three claims [b](a)[/b], [b](b)[/b], [b](c)[/b] directly follow from the following theorem:\r\n\r\n[b]Theorem 1.[/b] For every triangle ABC, we have\r\n\r\n$\\frac{[AB_2C_2]}{[ABC]}=\\frac{1}{4\\sin^2\\frac{A}{2}}$.\r\n\r\n[i]Proof.[/i] Let the B-excircle of triangle ABC touch its side CA at the point Y, and let the C-excircle touch its side AB at the point Z. Then, we have a pretty known lemma:\r\n\r\n[b]Lemma 2.[/b] We have $B_1I \\parallel BY$.\r\n\r\nSee below a proof of this lemma.\r\n\r\nOf course, $B_1I \\parallel BY$ can be rewritten in the form $C_2B_1 \\parallel BY$, and after Thales this yields\r\n\r\n$\\frac{AB_1}{AC_2}=\\frac{AY}{AB}$.\r\n\r\nSince the point Y is the point of tangency of the B-excircle of triangle ABC with the side CA, we have AY = s - c, where $s=\\frac{a+b+c}{2}$ is the semiperimeter of triangle ABC. Also, AB = c. Thus,\r\n\r\n$\\frac{AY}{AB}=\\frac{s-c}{c}$.\r\n\r\nHence,\r\n\r\n$\\frac{AB_1}{AC_2}=\\frac{s-c}{c}$.\r\n\r\nSince the point $B_1$ is the midpoint of the segment CA, we have $AC=2\\cdot AB_1$, and thus\r\n\r\n$\\frac{AC}{AC_2}=2\\cdot\\frac{s-c}{c}$.\r\n\r\nSimilarly,\r\n\r\n$\\frac{AB}{AB_2}=2\\cdot\\frac{s-b}{b}$.\r\n\r\nBut after a well-known area formula for triangles,\r\n\r\n$[ABC]=\\frac{AC\\cdot AB\\cdot \\sin\\measuredangle CAB}{2}$\r\n\r\nand\r\n\r\n$[AB_2C_2]=\\frac{AC_2\\cdot AB_2\\cdot \\sin\\measuredangle C_2AB_2}{2}$.\r\n\r\nSince $\\measuredangle CAB = \\measuredangle C_2AB_2$, we thus get\r\n\r\n$\\frac{[ABC]}{[AB_2C_2]}=\\frac{AC\\cdot AB}{AC_2\\cdot AB_2}=\\frac{AC}{AC_2}\\cdot\\frac{AB}{AB_2}$\r\n$=(2\\cdot\\frac{s-c}{c})\\cdot(2\\cdot\\frac{s-b}{b}) = 4\\cdot\\frac{(s-b)(s-c)}{bc}$.\r\n\r\nBut a well-known trigonometric formula (half-angle law) tells us that\r\n\r\n$\\frac{(s-b)(s-c)}{bc}=\\sin^2\\frac{A}{2}$;\r\n\r\nthus,\r\n\r\n$\\frac{[ABC]}{[AB_2C_2]}=4\\sin^2\\frac{A}{2}$,\r\n\r\nand\r\n\r\n$\\frac{[AB_2C_2]}{[ABC]}=\\frac{1}{4\\sin^2\\frac{A}{2}}$,\r\n\r\ncompleting the proof of Theorem 1.\r\n\r\nRemains to prove Lemma 2. This is actually a fundamental fact, but it also can be easily derived from the theorem that the Nagel point N, the centroid G and the incenter I of triangle ABC are collinear, and that NG : GI = 2 (with directed segments). In fact, since the point $B_1$ is the midpoint of the segment CA, the line $BB_1$ is a median of triangle ABC, and hence, it passes through the centroid G of the triangle ABC and is divided there in the ratio $BG:GB_1=2$. Combining this with NG : GI = 2, we get $BG:GB_1=NG:GI$, what, using Thales, leads to $B_1I \\parallel BN$. But, being the Nagel point of triangle ABC, the point N must lie on the line BY; hence, the line BN is nothing different but the line BY, and we get $B_1I \\parallel BY$, what proves Lemma 2.\r\n\r\nThus, the problem is solved.\r\n\r\nFrancisco Javier Garc\u00eda Capit\u00e1n kindly sent me a link containing another solution of the initial problem: http://www.animath.fr/UE/audin/exercice.htm problem 4.\r\n\r\n Darij"
}
{
"Tag": [
"probability",
"combinatorics proposed",
"combinatorics"
],
"Problem": "This is more of a combinatorics oriented than statistics oriented, so I'm posting it here. ;)\r\n\r\nIf you choose 3 points on a circle. What is the probability that they lie on a semicircle (.",
"Solution_1": "what do u mean not necessarily of the same circle there exista an unique circle passing through three non collinear points...but isn't this more a probability problem so u cud have posted it therr",
"Solution_2": "As I said, this not exactly a probability kind of question. I personally felt it belonged in the olympiad section. \r\n\r\nOkay, okay. leave that. If the 3 points are such that the distance between the extremes is greater than $ \\PiR$, then the major arc can never form the diameter of any circle.\r\n\r\nBut then, the minor arc can always be the diameter of any circle. I just gave the sum as in a book. Its a bit confusing actually. I can give the answer if you want."
}
{
"Tag": [
"number theory"
],
"Problem": "Prove that if p is a prime of the form 4n+1 where n is a positive integer, then [(2n)!]^2\u2261-1(mod p)",
"Solution_1": "Hi,\n\n\n\nMy attempt :\n\n[hide]\n\n(I have used '=' (mod p) instead of 'equivalent to' (mod p), since I didn't know how to write the latter without LaTeX... Sorry!!!)\n\n\n\n2n + 1 = -2n (mod p); 2n + 2 = -(2n-1) (mod p); ................. ; 4n = 2n + 2n = -1 (mod p)\n\n\n\nThus, (2n+1)(2n+2)......(4n) = (-1)^(2n) (2n) (2n-1) .... 1 (mod p)\n\nimplies (2n+1)(2n+2)......(4n) = (2n)! (mod p)\n\nimplies (2n!) (2n+1)(2n+2)......(4n) = ( (2n)! )^2 (mod p)\n\nimplies (4n)! = ( (2n)! )^2 (mod p)\n\n\n\nAnd LHS = (p-1)! which is = -1(mod p) by Wilson's Theorem, \n\n[/hide]\n\n***********************************************************************************\n\nSambit"
}
{
"Tag": [
"trigonometry",
"calculus",
"calculus computations"
],
"Problem": "if $\\alpha\\geq\\beta\\geq\\gamma\\geq\\frac{\\pi}{15}$ and $\\alpha+\\beta+\\gamma=\\frac{\\pi}{2}$ , find the max and min of\r\n\r\n$S=\\cos\\alpha\\sin\\beta\\cos\\gamma$",
"Solution_1": "i havent tried yet but it seems that you could just apply Lagrange multipliers, although there might be a \"better\" solution."
}
{
"Tag": [
"geometry",
"3D geometry"
],
"Problem": "Please reply soon . If possible with diagram.Thanks\r\n\r\n1)How many cubes of length 2cm can be made \r\n\r\n from a cubical wood piece of length 30 cm.",
"Solution_1": "Divide the volume of the big cube by the volume of the small cubes.\r\n\r\nSo you get $ \\frac{30^3}{2^3}\\equal{}15^3\\equal{}\\boxed{3375}$.\r\n\r\nYou can make 3375 cubes.",
"Solution_2": "thnx a lot for ur quick reply.\r\n\r\n i've sent u friend reqest.",
"Solution_3": "can you please explain more?",
"Solution_4": "If it was a line of 30 cm that we wanted to break into 2 cm pieces, it's obviously $ \\frac{30}{2}$.\r\n\r\nIf it's a square, it's $ \\left (\\frac{30}{2} \\right )^2$.\r\n\r\nSo if it's a cube, it's $ \\left (\\frac{30}{2} \\right )^3\\equal{}\\boxed{3375}$.",
"Solution_5": "yeah i understand now."
}
{
"Tag": [
"limit",
"logarithms",
"number theory unsolved",
"number theory"
],
"Problem": "When I was partitioning integers I found some nice results which I hope maybe verified.\r\nIn the usual manner, I had\r\n1.P(n)P(n+k) holds for all positive integers n but finietly many values.\r\n\r\nBOmb",
"Solution_1": "Here's a sketch of a \"not-so-combinatorial\" proof of your statement (proving these results rigorously requires some tools from analysis)\r\n\r\nThe infinite product $\\prod_{k=1}^\\infty {1 \\over 1-x^k}$ converges uniformly in the open disk centered at the origin of radius $r$ in the complex plane for any $r<1$. Therefore the infinite sum $\\sum_{n=1}^\\infty p(n)x^n$ must also converge on the whole disk as well. But that implies that $\\lim_{n \\rightarrow \\infty} p(n)^{1 \\over n}=1$. On the other hand, $F(n-k)$ grows exponentially for any fixed $k$.",
"Solution_2": "Could you please elaborate on your proof. Can you please prove it rigorously by analyss, I dont really understand what you've done. Thanks\r\n\r\nBomb",
"Solution_3": "We have $\\sum_{k=0}^\\infty p(k)x^k=F(x)=:\\prod_{k=1}^{\\infty} (1-x^k)^{-1}$. It holds for any positive $x$ less then 1. So, $p(k)\\le x^{-k}F(x)$, hence $\\ln p(n)\\le -n\\ln x-\\sum_{k=1}^\\infty \\ln(1-x^k)$. Now try to optimize RHS by $x$. Using logarithm's expantion, we have $-\\ln(1-x^k)=x^k+x^{2k}/2+x^{3k}/3+\\dots$ Now change the order of summands in RHS, you get $-n\\ln x+y_1+y_2/2+y_3/3+\\dots$, where $y_k=x^k/(1-x^k)=(1-x)^{-1}\\cdot z_k$, where $z_k=x^k/(1+x+x^2+\\dots+x^{k-1})\\le 1/k$. Using all this, we get \r\n\\[\r\n\\ln p(n)\\le -n\\ln x+(1+1/4+1/9+\\dots)(1-x)^{-1}=-n\\ln x+\\pi^2/(6(1-x))\r\n\\]\r\nFor $x=1-c/\\sqrt{n}$, we get $\\ln p(n)\\le (c+\\pi^2/(6c)) \\sqrt{n}+O(1)$, optimize by $c$, we must take $c=\\frac{\\pi}{\\sqrt{6}}$ and finally\r\n\r\n\\[\r\n\\ln p(n)\\le C+\\sqrt{2/3}\\,\\pi \\sqrt{n}\r\n\\]\r\n\r\nthe constant $\\sqrt{2/3}\\,\\pi $ is sharp."
}
{
"Tag": [
"geometry",
"geometric transformation"
],
"Problem": "One day a restless lion roamed about in his circular cage, of radius 10 meters, along a polygonal path of total length 30 km. Thus, coming to the end of one segment in the path. Prove that, counting both turns to the left and turns to the right (considering each angle to be a postive quantity), he must have turned through a total of at least 2998 radians.",
"Solution_1": "Use a triangle of large perimeter. This would mean minimal radians, as triangles have \r\na total angle sum of 180 degrees, or $\\pi$ radians. Say we are dealing with a right triangle inscribed in the circle, as that would give us large perimeter. Now we only have to deal with the radians calculations. As 0 $ then we have:\r\n\r\n$ (x_{1} +x_{2}+ \\ldots +x_{5})^2 \\geq 4(x_{1}x_{2}+x_{2}x_{3}+x_{3}x_{4}+x_{4}x_{5}+x_{5}x_{1}) $\r\n\r\ncheers! :D :D :cool:",
"Solution_1": "Don't you think we have discussed a lot about this inequality?\r\n\r\n[Moderator edit: Discuss about this inequaltiy on http://www.mathlinks.ro/Forum/viewtopic.php?t=5121 .]",
"Solution_2": "well... yes.. but.. perhaps you should move it to the begginers corner! :) \r\nit may be good for them! :)\r\n\r\nsorry again for the duplicate post!",
"Solution_3": "I hope you are joking! This one to the begginers corner? Isn't that too much? Don't forget that 5 is odd and the discution is not trivial in this case. It they were 6 or 8 variables, I would have moved it."
}
{
"Tag": [
"Olimpiada de matematicas"
],
"Problem": "Un triangulo ABC tiene el A=120, llamamos D, E, F al punto de corte de las bisectrices con los lados BC, AC y AB respectivamente. Demuestra que el EDF= 90.",
"Solution_1": "$E$ es excentro de $ABD$\r\n$F$ es excentro de $ADC$\r\n\r\n$\\measuredangle ADF= \\measuredangle FDB =\\alpha$\r\n\r\n$\\measuredangle CDE=\\measuredangle EDA=\\beta$\r\n\r\n$2\\alpha+2\\beta = 180^\\circ \\rightarrow \\alpha+\\beta=90^\\circ \\rightarrow \\measuredangle EDF=90^\\circ$"
}
{
"Tag": [],
"Problem": "A novel has 527 pages (pages 1 \u2013 527). How many digits will it take to number all 527 pages from 1 to 527?",
"Solution_1": "[hide=\"DONT LOOK!!!\"]9 * 1 = 9\n\n90 * 2= 180\n\n527-99= 428; 428 * 3= 1284\n\n9 + 180 + 1284= 1473, Thus, there are [b] 1473 [/b] digits. \n\n[/hide]",
"Solution_2": "[hide]\n\n9 one digits, 90 2 digits, 428 3 digits -- >1473[/hide]"
}
{
"Tag": [
"FTW"
],
"Problem": "Who plays \"For The Win\"",
"Solution_1": "You should have an option, \"Rarely\" or \"Sometimes\", because I rarely go play AoPS \"For the Win\". I only go on so rarely because I haven't experienced much problem solving/MathCounts.",
"Solution_2": "I played it a few times... but stopped\r\nI was not doing too well... Meh",
"Solution_3": "[quote=\"Cyrazeno\"]I played it a few times... but stopped\nI was not doing too well... Meh[/quote]\r\nexactly my case.",
"Solution_4": "I go on all the time. My rating proves it. :)",
"Solution_5": "I go on all the time too, but I'm trying to stop. I should be practicing for stuff, but I just get bored.",
"Solution_6": "Maybe you should put this in the FTW forums... :P But every since going into MOEMS it's helped problem solving a bit more."
}
{
"Tag": [
"search"
],
"Problem": "why is this locked? http://www.mathlinks.ro/Forum/viewtopic.php?t=85011&start=20",
"Solution_1": "maybe because everyone was posting nearly the same thing? :?",
"Solution_2": "Jose, you still haven't read the posting guidelines for the forum, have you?\r\n\r\nEach problem should have its own thread in general.\r\n\r\nIn generaly, people like you ;) keep posting problems beyond the scope of the forum within marathon-like threads. When a problem has its own thread, we can just move it, but in a marathon, that task becomes a moderator's nightmare.\r\n\r\nAlso, it's easier to search for or document problems with their own threads."
}
{
"Tag": [
"trigonometry",
"inequalities",
"geometry",
"trig identities",
"Law of Sines"
],
"Problem": "If in a triangle ABC $2m(B)=m(A)+m(C)$ then $2b\\geq a+c$.",
"Solution_1": "By the Law of Sines, we have $a=\\left (\\frac{sin(A)}{sin(B)}\\right )b$ and $c=\\left (\\frac{sin(C)}{sin(B)}\\right )b$. Plugging these into the inequality yields \r\n\r\n$sin(A)+sin(C)\\le 2sin(B)$.\r\n\r\nSince $2m\\angle B=m\\angle A+m\\angle C$, we have $m\\angle B=60$. Thus the above inequality turns into\r\n\r\n$sin(A)+sin(C)\\le\\sqrt{3}$.\r\n\r\nNow since $m\\angle A+m\\angle C=120$, let $A=60-\\theta$ and $C=60+\\theta$. Then after the use of the $sin(\\alpha+\\beta)$ identity, we have\r\n\r\n$\\sqrt{3}cos(\\theta)\\le\\sqrt{3}$\r\n\r\nwhich is definitely true."
}
{
"Tag": [
"calculus",
"integration",
"function",
"factorial",
"calculus open"
],
"Problem": "Could somebody help me to do the following integral?\r\n\r\n$\\int_0^{2\\pi}{exp(acos(z))dz}$\r\n\r\na is a real constant For me it would be enough expressing it as a known function (e.g. Bessell). I tried to do it with Mapple but it wasn't able either. Is it possible to find a closed form for it?",
"Solution_1": "Translating into complex terms, this is equal to\r\n$\\int_{|z|=1}\\exp(a\\text{Re }z)\\cdot\\frac{dz}{iz} =\\frac1i\\int_{|z|=1}\\exp(\\frac a2(z+\\frac1z))\\cdot\\frac1z\\,dz$.\r\nWe need the residue at zero of $\\exp(\\frac a2(z+\\frac1z))\\cdot\\frac1z$; this is the constant term of the product $(1+\\frac a2z+\\frac{a^2}{2^2\\cdot 2}z^2+\\cdots+\\frac{a^n}{2^n\\cdot n!}z^n+\\cdots)(1+\\frac a{2z}+\\frac{a^2}{8\\cdot z^2}+\\cdots+\\frac{a^n}{2^n\\cdot n!\\cdot z^n}+\\cdots)$, which is\r\n$1+\\frac{a^2}{4}+\\frac{a^4}{64}+\\cdots+\\frac{a^{2n}}{2^{2n}\\cdot (n!)^2}$.\r\n\r\nI'm pretty sure there's no closed form, and I don't know any functions that match it, with that squared factorial in the denominator.\r\n\r\nIt is at least very easy to approximate accurately.",
"Solution_2": "[quote=\"jmerry\"]I don't know any functions that match it, with that squared factorial in the denominator.[/quote]\r\nAh, but there is such a function - a Bessel function.\r\n\r\nThis appears to be $J_0(ia).$",
"Solution_3": "Is that the same than $I_0(a)$?\r\n\r\nThanks a lot for your help"
}
{
"Tag": [
"trigonometry",
"absolute value"
],
"Problem": "1/Find all values of $ \\alpha\\in N$ satisfying: $ tan\\alpha^o\\in Q$\r\n2/Prove that: $ tan34^o>\\frac{2}{3}$",
"Solution_1": "Anyone helps me please! :(",
"Solution_2": "[quote=\"Kyoshiro\"]1/Find all values of $ \\alpha\\in N$ satisfying: $ tan\\alpha^o\\in Q$\n2/Prove that: $ tan34^o > \\frac {2}{3}$[/quote]\r\n\r\n[hide=\"Solution 1\"]\nNow $ \\tan{45^o} \\equal{} 1,\\tan{180n^o} \\equal{} 0 \\in \\mathbb{Q}$,and for all $ x\\in \\mathbb R,\\tan(x \\plus{} 180^o) \\equal{} \\minus{} \\tan{x},\\tan(90^o \\minus{} x) \\equal{} \\frac {1}{\\tan{x}}$,thus it's enough by cousiderd for the case $ \\alpha\\in [1,44]$.\n\nThe tangent addition theorem says $ \\tan(x \\plus{} y) \\equal{} \\frac {\\tan{x} \\plus{} tan{y}}{1 \\minus{} \\tan{x}\\tan{y}}$.\n\nby using this repeatedly,we have the next lemma.\n\nLemma:Let $ n$ be a positive integer,then if $ \\tan(\\alpha^o)$ is rational,$ \\tan(n\\alpha^o)$ is also rational.(but it's necessary that all tangent can be defined.)\n\n1)If $ \\alpha$ is not dividible by $ 3$,then suppose $ \\tan(\\alpha^o)$ is a rational.by the lemma,$ \\tan(30\\alpha^o)$ is also rational,but this absolute value is equal to $ \\tan{30^o} \\equal{} \\frac {1}{\\sqrt {3}}$ or $ \\tan{60^o} \\equal{} \\sqrt {3}$ which are irrationals,contradiction!\n\n2)If $ \\alpha$ is dividible by $ 3$ but not by $ 9$,then suppose $ \\tan(\\alpha^o)$ is a rational.by the lemma,$ \\tan(10\\alpha^o)$ is also rational,but this absolute value is equal to $ \\tan{30^o} \\equal{} \\frac {1}{\\sqrt {3}}$ or $ \\tan{60^o} \\equal{} \\sqrt {3}$ which are irrationals,contradiction!\n\n3)If $ \\alpha$ is not dividible by $ 9$,i.e.,$ \\alpha \\equal{} 9^o,18^o,27^o,36^o$.\nSuppose $ \\tan{9^o}$ or $ \\tan{18^o}$ or $ \\tan{27^o}$ is rational,then,by the lemma and $ \\tan(90^o \\minus{} x) \\equal{} \\frac {1}{\\tan{x}}$,we can say $ \\tan{36^o}$ is also rational,but $ \\tan{36^o} \\equal{} \\sqrt {5 \\minus{} 2\\sqrt {5}}$ is irrational,contradiction!\n\n\ntherefore,we see that such $ \\alpha\\in \\mathbb{N}$ is $ \\alpha \\equal{} 90n \\minus{} 45$ or $ 180n$ ($ n\\in \\mathbb {N}$)[/hide]",
"Solution_3": "[quote=\"Kyoshiro\"]1/Find all values of $ \\alpha\\in N$ satisfying: $ tan\\alpha^o\\in Q$\n2/Prove that: $ tan34^o > \\frac {2}{3}$[/quote]\r\n\r\n[hide=\"Solution 2\"]\nBy the tangent addition theorem we have\n\n$ \\tan{34^o}\\equal{}\\frac{\\tan{30^o}\\plus{}\\tan{4^o}}{1\\minus{}\\tan{30^o}\\tan{4^o}}$.\n\nthen,we see that $ \\tan{34^o}>\\frac{2}{3}$ is equivalent to \n\n$ \\tan{4^o}>\\frac{24\\minus{}13\\sqrt{3}}{23}$.\n\nnow,\n$ \\tan{4^o}\\equal{}\\tan{\\frac{\\pi}{45}}>\\frac{\\pi}{45}>\\frac{3,14}{45}\\equal{}0,0697...>0,0691...\\equal{}\\frac{24\\minus{}13\\cdot 1,73}{23}>\\frac{24\\minus{}13\\sqrt{3}}{23}$,\n\nso that is true.[/hide]"
}
{
"Tag": [
"algebra",
"polynomial",
"function",
"complex analysis",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Hi there!\r\nIt is known that Tchebychev polynomials commute one with other.\r\nAlso the family $ P_{n}(x) = x^{n} $ is a commuting family of polynomials.\r\nI was thinking about this and have composed some problems about\r\ncommuting polynomials. The first of them are very simple or obvious,\r\nbut some of them are interesting, for my opinion.\r\nThe coefficients of polynomials can be considered in $ \\mathbb{C} $\r\nor in $ \\mathbb{R} $, as you wish.\r\nI will consider them in $ \\mathbb{R} $, but all the statements are\r\ntrue also for $ \\mathbb{C} $.\r\n\r\n\r\n1) If a polynomial commutes with linear polynomial $ Q(x) = ax + b $\r\nthen it is linear or constant.\r\n\r\nFrom now on all polynomials are monic.\r\nWe see that $ ( \\mathbb{R}, + ) $ acts on the set of monic polynomials\r\nby $ L_{a}P(x) := P(x + a) - a $, $ a \\in \\mathbb{R} $. It is easy to see that\r\nif polynomials $ P, Q $ commute, so $ L_{a}P , L_{a}Q $.\r\nWe say that polynomials $ P_{1}(x) , P_{2}(x) $ are equivalent if\r\nthere exists $ a \\in \\mathbb{R} $ such that $ P_{2} = L_{a}P_{1} $.\r\n\r\n\r\n2) If for $ n>1 $ a polynomial $ P(x) $ commutes with a polynomial\r\n$ Q(x) = x^{n} $, then it is of a form $ P(x) = x^{m} $ for some natural $ m $.\r\n\r\n3) Prove that for a polynomial $ Q(x) = x^{2} - 2 $ for every degree there exists\r\na polynomial, which commutes with it.\r\n\r\n4) Suppose we have fixed some natural $ n>1$.\r\nProve that almost every polynomial of degree $ n $ does not commute with any\r\npolynomial of degrees between $ 2 $ and $ n-1 $.\r\n\r\n5) Of course every polynomial $ P(x) $ commutes with $ P^{(n)}(x) := P \\circ \\ldots\r\n\\circ P (x) $ ( $ n $ times ), for $ n \\geqslant 0 $. If this set contains\r\nall polynomials which commute with $ P(x) $,then we call such a polynomial\r\npoor. For each degree find many pairwise\r\nnonequivalent polynomials which are poor and prove that.\r\n\r\n6) Prove that if $ P(x) $ commutes wit $ Q(x) $ and $ Q(x) $ commutes with\r\n$ R(x) $ then $ P(x) $ commutes with $ R(x) $.(It is important that all polynomials are monic)\r\n\r\nSo we see that property of commuteness is an equivalence relation ( on monic polynomials).\r\nSo every polynomial $ P(x) $ is contained in some maximal commuting family of\r\npolynomials. Call a maximal commuting family lucky, if it contains polynomials\r\nin each degree. So we know that there exists lucky family\r\n$ P_{n}(x) = x^{n} $ , $ n \\geqslant 1 $. Also we have a \"Tchebychev\" family $ \\{ T_{n}(x) \\}$\r\nfrom problem 3) . Also we have of course for every $ a \\in \\mathbb{R} $ that\r\n$ \\{ L_{a}P_{n} \\} $ is a lucky family and also $ \\{ L_{a}T_{n} \\} $ is a lucky family.\r\n\r\n7) Prove that there do not exists other lucky families.\r\n\r\nHere is one problem which is not connected to others:\r\n( actually it must be appeared in Analysis section)\r\n\r\n8) If an entire function $ f: \\mathbb{C} \\rightarrow \\mathbb{C} $ commutes with\r\na polynomial (you can take it with complex coefficients) of degree at least\r\n$ 2 $, then $ f $ is a polynomial.",
"Solution_1": "Wow! Nice, Leva! :)\r\n\r\nI have a question: in your first problem, don't you need some conditions on $a,b$? What if the polynomial is $P(x)=x^3$ and $Q(x)=-x$? Or maybe you meant to say that $P$ commutes with all linear polynomials? :? I don't think so, because that would make the problem trivial. Actually, it is pretty easy, except for the case $a=-1$, when we might have some trouble like that above.",
"Solution_2": "For the second one:\r\n\r\nWe may assume that $P$ has no roots equal to $0$ (separate the part $x^m|P$ with maximal $m$ from $P$). We have $P(x^n)=P^n(x)$. After we derive that we get $nx^{n-1}P'(x^n)=nP'(x)P^{n-1}(x)$. This means that $x^{n-1}|P'(x)$. From that we get $x^{n(n-1)}|P'(x^n)$, so $x^{n(n-1)}|P'(x)$. We repeat this to find that $x^m|P'(x)$ for all $m$, meaning that $P'(x)=0$, so $P$ is constant. Since we assumed it was monic, we get $P=1$. When we put back the part $x^m$ we had separated, we get $P(x)=x^m$.",
"Solution_3": "Sorry, grobber! :blush: \r\nIndeed, I a little bit huried with the 1-st problem.\r\nIndeed, it is not always true, so I apology, \r\nso the corrected problem is :\r\nFor given $ a,b $, classify all polynomials which commute with $ ax+b $.",
"Solution_4": "Also in problem 6 I assume that the polynomials $ P,Q,R $ have degrees at least $ 2 $.",
"Solution_5": "Ok, thanks for clarifying that.\r\n\r\nFor the third proble, the polynomial $Q$ is a member of the commuting family $2T_n\\left(\\frac x2\\right)$, so we can take our polynomials to be precisely these: $U_n=2T_n\\left(\\frac x2\\right)$ (they're also monic, if that's one of the requirements of the problem).",
"Solution_6": "Cool! :lol:",
"Solution_7": "Why nobody wants to solve my problems? :( :( :(",
"Solution_8": "Your questions are very interesting.For more this type of problems you can check the book\"Polynomials\" by E.J. Barbeau.Its a springer publication & a very nice book.",
"Solution_9": "[quote=\"Leva1980\"]Why nobody wants to solve my problems? :( :( :([/quote]\r\nThey are very nice, just too many for one post. I especially like #8. Try posting them one at a time. ;)",
"Solution_10": "Not so many people on this forum have three degrees in math! ;)",
"Solution_11": "But the problems are elementary, as their solutions\r\n( except the problem 8, which Fedja solved) :)",
"Solution_12": "Also, Ihave only 2 degrees, I am doing the 3rd now."
}
{
"Tag": [
"calculus",
"integration",
"function",
"trigonometry",
"Galois Theory",
"calculus computations"
],
"Problem": "I would like to know about some impossible integrals (to be expressed as finite sum/product of elementary functions)\r\nfor example $\\int e^{x^{2}}, \\int e^{-x^{2}}, \\int\\sqrt{a^{2}\\sin^{2}x+b^{2}\\cos^{2}x}, a>b$.\r\nCan someone help me to find more about this subject? Maybe you can recommend me some books which treat this kind of subject. I would like to know the proof that they are not expressible as elementary functions and maybe some ways to aproximmate them when the integrals are definite.\r\n\r\nThank you.",
"Solution_1": "You can start with Wikipedia and continue with the references mentioned there.\r\n\r\n[url=http://en.wikipedia.org/wiki/Differential_Galois_theory]Differential Galois theory[/url]: to prove that something is not expressible as an elementary function.\r\n\r\n[url=http://en.wikipedia.org/wiki/Quadrature_%28mathematics%29]Numerical integration[/url]: to approximate definite integrals.",
"Solution_2": "goto\r\nhttp://mathworld.wolfram.com/Erf.html\r\nhttp://mathworld.wolfram.com/Erfi.html\r\nhttp://mathworld.wolfram.com/EllipticIntegral.html",
"Solution_3": "I will refer to another of [b]greengrass[/b]'s [url=http://www.artofproblemsolving.com/Forum/viewtopic.php?p=617028#617028]posts[/url] -- this one has links to three papers that may interest you. :)\r\n\r\nBy the way this has been discussed many times, and I think the current thread has the most links out of all the ones I've seen, so I'm going to stick in some terms to make it more searchable:\r\n\r\nelementary antiderivative\r\nelliptic integral\r\nclosed form\r\n\r\nHopefully this helps people find the thread in the future. ;)",
"Solution_4": "it seems that improper or definite integrals of the second\r\nform you asked can be solved\r\nthe solution can be found [url=http://www.joma.org/images/cms_upload/delgado09010344816.pdf]here[/url]\r\nthe bibliography that is presented there has solutions to your first\r\nintegral as well(in the same conditions as the second one)",
"Solution_5": "The values of improper integrals $\\int_{0}^{\\infty}e^{-x^{2}}dx$ and $\\int_{-\\infty}^{\\infty}e^{-x^{2}}dx$ are well-known. It's not clear to me what other \"definite integrals of the second\r\nform\" you had in mind. And the improper integral of $e^{x^{2}}$ is divergent."
}
{
"Tag": [
"calculus",
"derivative",
"LaTeX",
"function",
"trigonometry",
"calculus computations"
],
"Problem": "Find the four second partial derivatives for f(x,y) = y^2e^x + xycosx\r\n\r\nI am stuck on the last part... here's what I got so far:\r\n\r\nZx = y^2e^x - ysinx\r\nZxx = y^2e^x - ycosx\r\nZxy = ?\r\n\r\nZy = 2ye^x + xcosx\r\nZyy = 2e^x + xcosx\r\nZyx = ?\r\n\r\n\r\nI need help with solving for xy. Both should end up being the same. :huh:",
"Solution_1": "[quote=\"p4nda\"]Find the four second partial derivatives for f(x,y) = y^2e^x + xycosx\n\nI am stuck on the last part... here's what I got so far:\n\nZx = y^2e^x - ysinx\nZxx = y^2e^x - ycosx\nZxy = ?\n\nZy = 2ye^x + xcosx\nZyy = 2e^x + xcosx\nZyx = ?\n\n\nI need help with solving for xy. Both should end up being the same. :huh:[/quote]\r\n\r\nWhy not?",
"Solution_2": "You've made mistakes both in calculating $ Z_{x}$ and in calculating $ Z_{yy}$. Try it again, and think carefully about every derivative of every term.\r\n\r\n\r\nAlso, your posts will be much nicer to read if you use LaTeX. An FAQ on LaTeX in the forum is at http://www.artofproblemsolving.com/Forum/viewtopic.php?t=123690",
"Solution_3": "Okay, from solving Zy, Zyy, Zyx... I got 2e^xy. But I'm having a difficult time solving for Zx, Zxx, Zxy, even though it ends up the same as Zyx. I need help on finding the derivative \"xycosx\" as if the y is treated as a constant.",
"Solution_4": "1) There is no partial derivative of that function which is equal to either $ 2e^{xy}$ or $ 2ye^{x}$.\r\n\r\n2) Can you calculate $ \\frac{d}{dx}\\left( x \\cdot \\cos x\\right)$?\r\n\r\n3) Seriously, use LaTeX. Or at least use parentheses."
}
{
"Tag": [
"induction",
"arithmetic sequence"
],
"Problem": "The odd natural numbers are arranged as follows:\r\n (1)\r\n (3,5)\r\n (7,9,11)\r\n (13,15,17,19)\r\n (21,23,25,27,29)\r\n ..........................\r\nFind the sum of the nth row",
"Solution_1": "Correct me if I'm wrong but isn't it n^3\r\n\r\nidk how though",
"Solution_2": "[hide]The sum of the first $ n$ odd numbers is $ n^2$, easily proven since $ n^2 \\minus{} (n \\minus{} 1)^2 \\equal{} 2n \\minus{} 1$.\n\nSo the nth row sums to:\n$ T_n ^2 \\minus{} T_{n \\minus{} 1}^2 \\equal{} \\left(T_{n \\plus{} 1} \\plus{} T_{n \\plus{} 2} \\right) \\left( T_{n \\plus{} 1} \\minus{} T_{n \\plus{} 2} \\right)$\n$ \\equal{} \\left(\\frac {n(n \\plus{} 1)}{2} \\plus{} \\frac {n(n \\minus{} 1)}{2} \\right) \\left( \\frac {n(n \\plus{} 1)}{2} \\minus{} \\frac {n(n \\minus{} 1)}{2} \\right)$\n$ \\equal{} n^2 \\cdot n \\equal{} n^3$,\n\nwhere $ T_n$ is the nth triangular number.[/hide]",
"Solution_3": "You could also prove that the nth row starts with\r\n1+n(n-1)\r\neither by induction or using arithmetic sequences, and since the nth row has n elements that constitute an arithmetic sequence, we can calculate the sum, arriving at the same formula."
}
{
"Tag": [
"geometry",
"3D geometry",
"sphere"
],
"Problem": "What is the volume of a sphere with a diameter of $\\pi$? :D",
"Solution_1": "diameter = $\\pi$ $\\longrightarrow$ radius = $\\frac{\\pi}{2}$\r\nvolume = $\\frac{4}{3}\\pi r^3 = \\frac{4}{3}\\pi (\\frac{\\pi}{2})^3 = \\frac{\\pi^4}{6}$ :)",
"Solution_2": "[quote=\"236factorial\"]What is the volume of a sphere with a diameter of $\\pi$? :D[/quote]\r\n\r\n[hide]Volume of sphere $=\\frac{4}{3}*\\pi*\\left(\\frac{\\pi}{2}\\right)^3$\n\n$=\\frac{4\\pi}{3}*\\frac{\\pi^3}{8}$\n\n$=\\frac{4\\pi^4}{24}$\n\n$=\\frac{\\pi^4}{6}$[/hide]",
"Solution_3": "delete this post.\r\ni accidentaly hit submit instead of preview. :wallbash_red:",
"Solution_4": "[hide]the volume is $\\frac{4}{3}\\pi \\cdot r^3$\nin this case the radius is $\\frac{\\pi}{2}$\nso now the volume is$\\frac{4}{3}\\pi \\left(\\frac{\\pi}{2}\\right)^3$\n$\\frac{\\pi}{2}^3$\nis $\\frac{\\pi^3}{8}$\n$\\frac{\\pi^3}{8} \\cdot \\frac{4}{3} \\cdot \\pi=\\frac{4 \\cdot \\pi^4}{24}=\\boxed{\\frac{\\pi^4}{6}}$\n\n\n\n\n\n\n\n\n\n\n :) :) :) :) :) :) :) :) :) :) :) :) \n[/hide]",
"Solution_5": "I love to double post too, math92 ;) \r\n\r\nSay, Drunken_Math, where did you get the formula for the sphere? $\\frac{\\pi}{2}$ for the radius?",
"Solution_6": "hey, to the genius who came up with the title: it's '[i]radii[/i]' not '[i]radiuses[/i]'.",
"Solution_7": "[quote=\"gman\"]hey, to the genius who came up with the title: it's '[i]radii[/i]' not '[i]radiuses[/i]'.[/quote]\r\n\r\nSave that babble and solve the problem first ;)"
}
{
"Tag": [
"calculus",
"integration",
"trigonometry",
"calculus computations"
],
"Problem": "show that $ {\\int_{0}^{\\infty}}{\\sin{x^2} dx}$ convergents",
"Solution_1": "Substitute $ x^a\\equal{}y$ in the integral $ \\int_0^\\infty \\sin x^a dx$ ($ a>1$) and then integrate by parts...."
}
{
"Tag": [
"geometry",
"inequalities",
"trigonometry",
"inequalities unsolved"
],
"Problem": "Prove that in each triangle we have the inequalities\r\n1)$\\sqrt{bc(p-a)}+\\sqrt{ca(p-b)}+\\sqrt{ab(p-c)}\\leq\\ 3R\\sqrt p$.\r\n2)$\\sum\\cos\\frac{B}{2}\\cos\\frac{C}{2}\\leq\\frac{9}{4}$",
"Solution_1": "at least the first one is easy if i'm not mistaken(the hour is late so it is posible to make silly mistakes).i haven't got the time to post a full solution but i think only the main ideas will be enough.use $R=abc/4S$then substitute $a=(x+y)/2,b=(y+z)/2,c=(x+z)/2$use am-gm for the radical and it will be obvious. ;) if you need more details just say and i will try to post it tomorrow morning",
"Solution_2": "q: cos(A/2)cos(B/2)+cos(B/2)cos(C/2)+cos(C/2)cos(A/2) <= 9/4.\r\na: The lhs <= cos(A/2)^2+cos(B/2)^2+cos(C/2)^2 = (1+cos(A))/2+(1+cosB)/2+(1+cosC)/2 = (3+cosA+cosB+cosC)/2 <= (3+3cos(A+B+C)/3)/2 = (3+3/2)/2 = 9/4."
}
{
"Tag": [
"algebra unsolved",
"algebra"
],
"Problem": "I know that :\r\n\\[\\sum_{k = 0}^{2n}{C_{2n}^{k}}t^{k}= (t+1)^{2n}\\]\r\nfind out how much is :\r\n\r\n\\[\\sum_{k = 0}^{2n}{C_{4n}^{2k}}t^{k}= ? \\]",
"Solution_1": "\\[\\frac{(1+\\sqrt t)^{4n}+(1-\\sqrt t)^{4n}}2 \\]",
"Solution_2": "This Topic is just making easy this problem http://www.mathlinks.ro/Forum/viewtopic.php?t=125070"
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "find all natural $x,y$ such that:$2^x-11^y=5$",
"Solution_1": "Isn't it [url=http://www.mathlinks.ro/Forum/viewtopic.php?t=41824]here[/url]?"
}
{
"Tag": [
"geometry",
"3D geometry",
"function",
"quadratics",
"complex numbers",
"algebra"
],
"Problem": "Ok so i was playing around with complex numbers today and i forund something iteresting and soo wierd :huh: :huh: ok\r\n\r\n$\\sqrt-1=i$\r\n$i^{2}=-1$\r\nSo now how does\r\n$(\\sqrt-1) (\\sqrt-1) =-1$\r\n\r\nwhen you normaly mutiply rotts you multiply the radicands to get teh answer but the answer should be 1 not -1 :|",
"Solution_1": "[quote=\"Jeremic\"]Ok so i was playing around with complex numbers today and i forund something iteresting and soo wierd :huh: :huh: ok\n\n$\\sqrt-1=i$\n$i^{2}=-1$\nSo now how does\n$(\\sqrt-1) (\\sqrt-1) =-1$\n\nwhen you normaly mutiply rotts you multiply the radicands to get teh answer but the answer should be 1 not -1 :|[/quote]\r\nThink about it this way. $\\sqrt{x}*\\sqrt{x}=\\sqrt{x^{2}}=x$",
"Solution_2": "[quote=\"Jeremic\"]Ok so i was playing around with complex numbers today and i forund something iteresting and soo wierd :huh: :huh: ok\n\n$\\sqrt-1=i$\n$i^{2}=-1$\nSo now how does\n$(\\sqrt-1) (\\sqrt-1) =-1$\n\nwhen you normaly mutiply rotts you multiply the radicands to get teh answer but the answer should be 1 not -1 :|[/quote]\r\n\r\nLet $x$=$-1$.\r\nLet us start like this:\r\n\r\n$\\sqrt{x}\\cdot\\sqrt{x}$\r\n$(\\sqrt{x})^{2}$\r\n$\\sqrt{x^{2}}$\r\n$=x$\r\nPlug in $-1$ for $x$.\r\n\r\noh wait, this is really weird...",
"Solution_3": "that works but i dont understand one thing\r\n\r\nif you have $\\sqrt-1^{2}=-1$ but if you evaluate that $-1^{2}=1$ and $\\sqrt1=1$",
"Solution_4": "[quote=\"Jeremic\"]that works but i dont understand one thing\n\nif you have $\\sqrt-1^{2}=-1$ but if you evaluate that $-1^{2}=1$ and $\\sqrt1=1$[/quote]\r\nWell actually, the property $\\sqrt{a}\\cdot\\sqrt{b}=\\sqrt{ab}$ only holds if $a,b\\geq0$, which is why you can't use it with $\\sqrt{-1}$. (That's why it got you an incorrect answer.)",
"Solution_5": "ohhh ok so i have to use $\\sqrt-x=x$ ok thank you for all teh help",
"Solution_6": "[quote=\"Jeremic\"]ohhh ok so i have to use $\\sqrt-x=x$ ok thank you for all teh help[/quote]\r\n\r\nWhere did you get $\\sqrt{-x}=x$ from?",
"Solution_7": "[quote=\"DanK\"][quote=\"Jeremic\"]ohhh ok so i have to use $\\sqrt-x=x$ ok thank you for all teh help[/quote]\n\nWhere did you get $\\sqrt{-x}=x$ from?[/quote]\r\n\r\ni think he forgot to add a square...",
"Solution_8": "[quote=\"anirudh\"]$\\sqrt{x^{2}}$\n$=x$[/quote]\r\nThis is invalid, because $\\sqrt{x^{2}}=|x|$, although $\\sqrt{x}^{2}=x$",
"Solution_9": "Well, $\\sqrt{x}\\cdot \\sqrt{x}= x$ for all real x. In this case, x=-1, which is a real number.",
"Solution_10": "[quote=\"nutz_for2.718281828\"]Well, $\\sqrt{x}\\cdot \\sqrt{x}= x$ for all real x. In this case, x=-1, which is a real number.[/quote]\r\n\r\nAs stated earlier, by nebula42, the case is only true when $\\sqrt{x}\\in\\mathbb{R}$, implying that $x\\geq0$.\r\n\r\nAlso, $\\sqrt{x^{2}}=|x|$, so $\\sqrt{\\imath^{2}}=|\\imath|$...",
"Solution_11": "bflat, thats EXACTLY what I said. But he's talking about $\\sqrt{x}^{2}$",
"Solution_12": "Yes, I did mess up on the part about $\\sqrt{x^{2}}$ instead of $(\\sqrt{x})^{2}$, but I still don't think you get the point.\r\n\r\n$(\\sqrt{x})\\in\\mathbb{R}$, not what you said. You only said that $x\\in\\mathbb{R}$, which causes $(\\sqrt{x})\\not\\in\\mathbb{R}$, if $x<0$. The identity is true only if $x\\geq0$.",
"Solution_13": "He's wondering why $(\\sqrt{-1})^{2}=-1$ and not 1. For ALL real x, $(\\sqrt{x})^{2}=x$ even when x is negative. Or at least the AoPS book Vol. 1 says.",
"Solution_14": "To stop freaking yourself out, you need to be more cautious with the symbol $\\sqrt{\\cdot}.$\r\n\r\nEvery positive real number has two square roots. Every nonzero complex number has two square roots (and three cube roots, and four fourth roots, and so on). What does the symbol $\\sqrt{\\cdot}$ mean? In order for it to define a function, we must have an agreed-upon convention. That convention is this:\r\n\r\nFor nonnegative real numbers $x,$ the symbol $\\sqrt{x}$ denotes the unique nonnegative number $y$ such that $y^{2}=x.$\r\n\r\nNow, what about complex numbers? The symbol $\\sqrt{-1}$ lies utterly outside the reach of that convention. Trying to \"define\" $i$ as $i=\\sqrt{-1}$ falls woefully short of being a valid mathematical definition. Once we have the complex numbers constructed, we will have that $i$ is [b]a[/b] square root of $-1,$ but that's not the same thing as saying that it is [b]the[/b] square root of $-1.$ There are two complex numbers with equal claim to being the square root of $-1.$ One of them is $i$; the other one is $-i.$\r\n\r\nWe do need to construct the complex numbers somehow - perhaps as ordered pairs with a defined arithmetic. That's essentially that same as saying that complex numbers are objects that can be written in the form $z=x+iy$ where $x$ and $y$ are real and a central feature of the arithmetic is that $i^{2}=-1.$ It's not enough to say that; you need to grub through the proof that this gives you a [i]field[/i]. That is, we need to show that addition is commutatitve and associative and has an identity and inverses; that multiplication is commutative and associative and that there's an identity and that nonzero numbers have multiplicative inverses, and that multiplication distributes over addition.\r\n\r\nNow: suppose $z^{2}=u$ and $w^{2}=u$ where $z,w,$ and $u$ are complex numbers. (That is, $z$ is a square root of $u$ and $w$ is a square root of $u.$) Can we conclude that $z=w?$ No, we can't. This is a quadratic equation and has two roots. The most we can conclude is that either $z=w$ or $z=-w.$\r\n\r\nAnd there's no [i]continuous[/i] way do make up a rule that specifies which of the two possibilities is \"the\" square root of a complex number."
}
{
"Tag": [
"geometry",
"3D geometry",
"algebra",
"factorization",
"difference of cubes",
"special factorizations"
],
"Problem": "Factor 250x^3-2\r\n\r\n\r\ni got an answer but I heard this was extremely complicated.\r\nI don't get why...",
"Solution_1": "[quote=\"8parks11\"]Factor 250x^3-2\n\n\ni got an answer but I heard this was extremely complicated.\nI don't get why...[/quote]\r\n\r\n[hide]Notice that this is just 2(125x^3-1).\n\nAnd remember that a^3-b^3=(a-b)(a^2+ab+b^2)\n\nGoes from there...[/hide]",
"Solution_2": "thats what I got but i heard this was some weird theory and i searched it and got nothing.. ( i heard i had to use that 5^3=125 and stuff)",
"Solution_3": "Well, you do. $250x^{3}= 2(5x)^{3}$.",
"Solution_4": "Yeah, and of course 1^3=1...",
"Solution_5": "Look at it like this:\r\n$250x^{3}-2$\r\nFactor out a two:\r\n$(2)(125x^{3}-1)$\r\nUse the method for factoring a difference of cubes:\r\n$2(5x-1)(25x^{2}+5x+1)$\r\nAnd we are finished."
}
{
"Tag": [
"geometry",
"circumcircle",
"incenter",
"geometry proposed"
],
"Problem": "Let $ (O) $ be the circumcircle of the triangle $ ABC $ . $ (I) $ is the incircle of the triangle $ ABC $ . $ (I) $ tangents $ BC, CA , AB $ , respectively , at $ D , E , F $ . A circle $ (O_1) $ externally tangents $ (I) $ at $ D $ and internally tangents $ (O) $ at $ K $. A circle $ (O_2) $ externally tangents $ (I) $ at $ E $ and internally tangents $ (O) $ at $ M $. A circle $ (O_3) $ externally tangents $ (I) $ at $ F $ and internally tangents $ (O) $ at $ N $ .$ A_1 $ is the midpoint of the arc $ BC $ (not contains A) . $ B_1 $ is the midpoint of the arc $ AC $ (not contains B) . $ C_1 $ is the midpoint of the arc $ BA $ (not contains C) . $ KA_1 $ intersects $ BC $ at $ A_2 $ . $ MB_1 $ intersects $ AC $ at $ B_2 $ . $ NC_1 $ intersects $ BA $ at $ C_2 $ . Prove that : $ A_2 , B_2 , C_2 $ are collinear",
"Solution_1": "I will prove that $A_2, E, F $ are collinear.\r\n\r\nFrom here directly $AD$ is polar of $A_2$, \u2026\r\n(three polars intersect in Gergonne point) hence $A_2,B_2,C_2$ are collinear.\r\n\r\nWLOG suppose $b > c$. Then $B$ lies between $A_2$ and $C$\r\n\r\nLet $A_3$ be other midpoint of arc $BC$. $A_3A_1$ is diameter.\r\n\r\nFrom previous problems on ML we have $A_3, D, K$ are collinear.\r\n\r\nLet $X$ be the midpoint of $BC$. $A_3, X, A_1$ are collinear.\r\n\r\n $A_3K \\perp A_2K$ $A_1X \\perp BC$. So $A_1XDK$ concyclic.\r\n\r\n$A_2B . A_2C = A_2K . A_2A_1= A_2D . A_2X$\r\n\r\n$A_2X^2 - \\frac {a^2}{4}= A_2X . (A_2X - DM)$\r\n\r\n$DM = (b - c)/2$ and $A_2X (b - c)/2 = a^2/4$, \r\n\r\n$A_2X = \\frac {a^2}{2(b - c)}$\r\n\r\n$A_2B = A_2X - a/2 = \\frac{a}{2(b-c)} (a+c-b)$,\r\n\r\n$A_2C =A_2X +a/2 = \\frac{a}{2(b-c)}(a-c+b)$\r\n\r\n$\\frac {A_2B}{A_2C} = \\frac{p-b}{p-c}$\r\n\r\nApplying Menelaos for $ABC$ and points $E,F,A_2$ we get that $E, F,A_2 $ are collinear.",
"Solution_2": "Please look: ( http://groups.yahoo.com/group/Hyacinthos/message/6208 ). Many thanks to Darij for this proof.\r\n Now: $AK,BM,CN$ concur at $P$ (see: http://www.mathlinks.ro/Forum/topic-36681.html ) and $AA_1,BB_1,CC_1$ (bisectors) concur at $Q$ (incenter of $ABC$) in other words $KMN$ and $A_1B_1C_1$ are the circumcevian triangles of the two points $P$ and $Q$. Now the cut points $A_2,B_2,C_2$ of $(KA_1,BC),(MB_1,AC),(NC_1,BA)$ are collinear.",
"Solution_3": "Could you please tell me what the circumcevian triangles are ? :?:",
"Solution_4": "Please look:\r\nhttp://mathworld.wolfram.com/CircumcevianTriangle.html"
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "Let $ a, b, c$ be three positive real numbers such that $ \\frac{1}{a}\\plus{}\\frac{1}{b}\\plus{}\\frac{1}{c}\\equal{}3$. Find the max value of $ Q\\equal{}\\frac{ab}{a^3\\plus{}b^3}\\plus{}\\frac{bc}{b^3\\plus{}c^3}\\plus{}\\frac{ca}{c^3\\plus{}a^3}$.\r\n\r\nThank.",
"Solution_1": "$ \\frac{ab}{a^3 \\plus{} b^3} \\leq \\frac{\\frac{1}{a} \\plus{} \\frac{1}{b}}{4}$ is equivalent to $ 4a^2 b^2 \\leq (a\\plus{}b)(a^3 \\plus{} b^3 )$ which is trivially true.\r\n\r\nThis yields $ Q \\leq \\frac{\\frac{1}{a} \\plus{} \\frac{1}{b} \\plus{} \\frac{1}{c}}{2} \\equal{} \\frac{3}{2}$",
"Solution_2": "[quote=\"BillClin\"]Let $ a, b, c$ be three positive real numbers such that $ \\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c} \\equal{} 3$. Find the max value of $ Q \\equal{} \\frac {ab}{a^3 \\plus{} b^3} \\plus{} \\frac {bc}{b^3 \\plus{} c^3} \\plus{} \\frac {ca}{c^3 \\plus{} a^3}$.\n\nThank.[/quote]\r\n[b]Here is another solution[/b]:\r\nIt is very wel- known that $ a^3 \\plus{} b^3\\ge ab(a \\plus{} b)$ \r\n$ \\Longrightarrow Q \\equal{} \\frac {ab}{a^3 \\plus{} b^3} \\plus{} \\frac {bc}{b^3 \\plus{} c^3} \\plus{} \\frac {ca}{c^3 \\plus{} a^3} \\le \\frac {1}{a \\plus{} b} \\plus{} \\frac {1}{b \\plus{} c} \\plus{} \\frac {1}{c \\plus{} a} \\le \\frac {1}{2}(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c}) \\equal{} \\frac {3}{2} \\Longrightarrow max Q \\equal{} \\frac {3}{2}$."
}
{
"Tag": [
"calculus",
"derivative",
"function",
"real analysis",
"real analysis solved"
],
"Problem": "let $f$ be a continuous function on the open interval $]a,b[$.\r\nSuppose that for every $x \\in ]a,b[$, $lim_{h->0} \\frac{f(x+h)+f(x-h)-2f(x)}{h^2}=0$.\r\nProve that $f=ax+b$",
"Solution_1": "It is a direct consequence from http://www.mathlinks.ro/Forum/viewtopic.php?t=20566 , since we obtain that $f$ is convex and concave in the same time."
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Let \\[ x^2 \\plus{} y^2 \\plus{} xy \\equal{} 1\r\n\\]\r\n\r\nDenote \\[ f(x,y) \\equal{} |(m \\plus{} 1)x^2 \\plus{} (m \\plus{} 2)y^2 \\plus{} (m \\minus{} 1)xy|\r\n\\]\r\n\r\nFind the $ Min(Max(f(x,y)))$ there Min means the minimal value and Max means the maximal one",
"Solution_1": "[quote=\"Allnames\"]Let\n\\[ x^2 \\plus{} y^2 \\plus{} xy \\equal{} 1\n\\]\nDenote\n\\[ f(x,y) \\equal{} |(m \\plus{} 1)x^2 \\plus{} (m \\plus{} 2)y^2 \\plus{} (m \\minus{} 1)xy|\n\\]\nFind the $ Min(Max(f(x,y)))$ there Min means the minimal value and Max means the maximal one[/quote]\r\nNo one gives me a help. Where are you, mathlinkers being good at inequality ?"
}
{
"Tag": [],
"Problem": "Se consider\u0103 paralelogramul ABCD, cu unghiul D>90. Perpendiculara din D pe AC taie cercul ABC \u00een dou\u0103 puncte. S\u0103 se arate c\u0103 unul dintre acestea este simetricul lui D fa\u0163\u0103 de AC, iar cel\u0103lalt este ortocentrul triunghiului ADC.\r\n\r\nCaut mai exact o solu\u0163ie sintetic\u0103...cu un pic de trigonometrie mi-a ie\u015fit.",
"Solution_1": "Vectorial iese imediat !",
"Solution_2": "Sunt si eu curios! Levi, puteti presenta o solutie in acest sens?\r\n :D",
"Solution_3": "See http://www.mathlinks.ro/Forum/viewtopic.php?t=56001",
"Solution_4": "Solu\u0163ia - sintetic\u0103, ce-i drept - nu m\u0103 mul\u0163ume\u015fte 100%. Cu ani \u00een urm\u0103, exista \u00een manualul de clasa a IX-a o \"problem\u0103 cu p\u0103trat\" care suna cam a\u015fa: \"\u00een interiorul unui p\u0103trat ABCD se ia un punct M astfel \u00eenc\u00eet unghiurile \r\nMCD \u015fi MDC s\u0103 fie ambele de m\u0103sur\u0103 15. S\u0103 se arate c\u0103 triunghiul MAB este echilateral\".\r\n\r\n Solu\u0163ia cea mai la \u00eendem\u00een\u0103 era considerarea unui punct M' astfel \u00eenc\u00eet triunghiul M'AB s\u0103 fie echilateral, urmat\u0103 de demonstrarea faptului c\u0103 M'=M. \u00cen ata\u015fament (format .PDF), iat\u0103 \u015fi o alt\u0103 solu\u0163ie - accesibil\u0103 \u015fi elevilor de clasa a VI-a - \u015fi care mie cel pu\u0163in \u00eemi place mult mai tare. \u00cen spiritul acesteia sper s\u0103 existe o solu\u0163ie \u015fi pentru problema 11342...",
"Solution_5": "[u]Principiul demonstratiei \"directei\" cu ajutorul \"reciprocei\" (care deseori este mai simpla sau chiar imediata ca in cazul nostru) prin utilizarea unei functii injective.[/u]\r\n\r\nSa reluam problema din clasa a IX - a care m-a fascinat si pe mine prin anii 1957 (eram in clasa a VIII - a), mai ales cand profesorul nostru ne-a oferit solutia (identica cu cea mentionata de tine, [i]Bpisai[/i]):\r\n\" Fie patratul $ABCD$ si un punct $M$ in interiorul acestuia si pe mediatoarea laturii $[AB]$.. Sa se arate: $m(\\widehat {MCD})=15^{\\circ}\\Longleftrightarrow MA=AB$.\"\r\n\r\n[u]Solutie.[/u] Notam mijloacele $M_1$, $M_2$ ale laturilor $AB$, $CD$ respectiv. Sa dam un grad de libertate punctului $M$ prin deplasarea acestuia pe mediatoarea laturii $[AB]$ catre latura $[CD]$. In timp ce punctul $M$ se apropie de latura $[CD]$, masura unghiului $\\widehat {AMB}$ descreste strict de la valoarea $\\pi$ cand $M: =M_1\\in [AB]$ pana la valoarea $2\\arctan\\frac 12<\\frac{\\pi}{3}<\\pi$ cand $M: =M_2\\in [CD]$. Deci va exista o singura pozitie a punctului $M$ pentru care $m(\\widehat {AMB})=\\frac{\\pi}{3}$. Si acum vom putea folosi reciproca acestei probleme care este evidenta: $MA=AB\\Longrightarrow m(\\widehat {MCD})=15^{\\circ}$. Functia folosita $f\\ : \\ [M_1M_2]\\rightarrow R,\\ M\\in [M_1M_2]$, $f(M)=m(\\widehat {AMB})$ este injectiva.\r\n\r\n[u]Observatie.[/u] Se pot oferi numeroase exemple de probleme de geometrie (si nu numai) cre se rezolva in aceasta maniera. Iata un exemplu semnificativ care este o problema cu adevarat dificila:\r\n\r\n\" Fie triunghiul $ABC$ pentru care $A=50^{\\circ}$ si exista un punct $P$ in interiorul sau astfel incat $PB=PC$, $E\\in PB\\cap AC$, $F\\in PC\\cap AB$, $m(\\widehat {PEF})=50^{\\circ}$, $m(\\widehat {PFE})=30^{\\circ}$. Sa se determine $m(\\widehat {ABE})$.\"\r\n\r\n[u]Rugaminte.[/u] In ipoteza fericita ca ati rezolvat-o (nu neaparat cu principiul mentionat mai sus) postati solutia in continuare.",
"Solution_6": "Eu stiu o solutie prin metoda reducerii la absurd la problema asta cu patratul. Apropo, ea apartine lui Ivan Block. ;)",
"Solution_7": "Dar repede ai aparut, Cezar ! Nici nu am apucat sa termin redactarea corecta a reply-iului ca ai si strecurat un reply. Ce parere ai de acest principiu de rezolvare ? Poti oferi o aplicatie al lui tot in geometrie ? Daca da, te rog sa-l asezi tot aici. Cel care m-a invatat cu adevarat matematica si totodata m-a facut sa o iubesc este acest profesor de liceu de la care de fapt am invatat acest principiu."
}
{
"Tag": [
"function",
"algebra unsolved",
"algebra"
],
"Problem": "Find all functions $f: \\mathbb R^{+} \\to \\mathbb R^{+}$ satisfying \n\\[ (1\\plus{}yf(x))(1\\minus{}yf(x\\plus{}y))\\equal{}1, \\quad \\forall x,y \\in \\mathbb R^\\plus{}.\\]",
"Solution_1": "Let $ g(x)\\equal{}xf(x)$, then we get $ g(x\\plus{}y)\\equal{}\\frac{(x\\plus{}y)g(x)}{x\\plus{}yg(x)}$.\r\nIf $ y\\equal{}zx$, then $ g(x(1\\plus{}z))\\equal{}\\frac{(1\\plus{}z)g(x)}{1\\plus{}zg(x)}$ or \\[ g(t)\\equal{}\\frac{t}{t\\plus{}g_1(x)}, g_1(x)\\equal{}x\\frac{1\\minus{}g(x)}{g(x)},t\\equal{}x(1\\plus{}z).\\]\r\nIt mean $ g_1(x)\\equal{}const\\equal{}a$. It give $ g(x)\\equal{}\\frac{x}{x\\plus{}a},f(x)\\equal{}\\frac{1}{x\\plus{}a}.$",
"Solution_2": "[quote=\"\u222bFaiLurE\u222e\"]Let $ R^ \\plus{}$ be the set of all positive real numbers. Find all functions $ f: R^ \\plus{} \\to R^ \\plus{}$ satisfying \n$ (1 \\plus{} yf(x))(1 \\minus{} yf(x \\plus{} y)) \\equal{} 1$ ,$ \\forall x,y \\in R^ \\plus{}$[/quote]\r\n\r\nSet $ g(x) \\equal{} \\frac {1}{f(x)} \\minus{} x$. The given equation is equivalent to $ f(x) \\equal{} f(x \\plus{} y) \\plus{} yf(x)f(x \\plus{} y)$ or $ g(x \\plus{} y) \\equal{} g(x)$ for all $ x,y \\in \\mathbb{R_ \\plus{} }$. It follows that $ g$ is constant, hence $ f(x) \\equal{} \\frac {1}{x \\plus{} k}$ for some arbitrarily chosen $ k \\ge 0$.",
"Solution_3": "[quote=\"FelixD\"] ... hence $ f(x) \\equal{} \\frac {1}{x \\plus{} k}$ for some arbitrarily chosen $ k > 0$.[/quote]\r\n\r\nTo be precise, $ k\\equal{}0$ fits too. :)",
"Solution_4": "[quote=\"pco\"][quote=\"FelixD\"] ... hence $ f(x) \\equal{} \\frac {1}{x \\plus{} k}$ for some arbitrarily chosen $ k > 0$.[/quote]\n\nTo be precise, $ k \\equal{} 0$ fits too. :)[/quote]\r\n\r\nOf course it does^^... Thanks, pco."
}
{
"Tag": [
"geometry",
"rectangle",
"trigonometry",
"function",
"AMC",
"USA(J)MO",
"USAMO"
],
"Problem": "Hmm...is it customary for people to establish their own personal end of proof symbols? :D",
"Solution_1": "[quote=\"Fierytycoon\"]Hmm...is it customary for people to establish their own personal end of proof symbols? :D[/quote]\r\n\r\nI much prefer consistency in end of proof symbols as in other matters of mathematical notation. The \"Halmos square\" (a black filled-in rectangle) is all right, except that I see it used for purposes other than marking the end of a proof, even in math textbooks. I personally would end proofs with Q.E.D. if I were writing a math book myself (and would explain what \"Q.E.D.\" means) but I have nothing against other conventions for showing the end of a proof.",
"Solution_2": "I make my students write \"QED\" at the end of their proofs. For a forum like this, using a smiley or such is fine, but it's probably not a good idea to use \"invented notation\" on a regular basis - it confuses everyone else needlessly.\r\n\r\nRichard Feynman talks about this in one of his books - he had invented his own notation for trig functions and log functions which he used for awhile, until he had to waste too much time explaining to others what he was doing.",
"Solution_3": "Sometimes when I'm bored I try to come up with original end-of-proof symbols, but I tend to stick with the Halmos symbol. Sometimes when I finish a proof online, I write $\\blacksquare$.",
"Solution_4": "i dont see why you should not make an end of proof symbol. If your proof has been finished, your grader will realize the fact that the its over, and the end symbol is only there for fun, or personality, or whatever. ive heard of people making up end of proof symbols at imo (gabriell carrol drew a gastropod i think) so i dont see what the problem is. a contest is not a book, it is a lot more personal, so i think it makes sense to end with your own symbol.",
"Solution_5": "Neal's into QED a lot... i'll let him tell the rest of the story..",
"Solution_6": "I don't think you should use a personalized end of proof symbol on an olympiad. The graders should not have any idea who is writing the proof since it may influence the score.",
"Solution_7": "I meant for casual work, not for the USAMO or something.",
"Solution_8": "well, if it were casual work, i would just spend 20 mintues drawing a very cool design!",
"Solution_9": "I personally write the last two lines like this:\r\n\u2234(for those of you who can't see this, it's the \"therefore\" sign with the three dots.) (statement)\r\n(statement)\r\n\r\nFor example:\r\n\u2234 2x=y\r\nx=y/2\r\n\r\nor something like that. The whole point of using the same mathematical symbols in every country in the world (such as =, numbers, etc.) is to reduce all the work translation is going to take. I mean, just by one country having it's own system of measurement, a whole rocketship or something like that :) fell from the sky. Think about what'll happen if some country described the number 1 as 1 and another country wrote 2 for it.\r\n\u2234 I think we should stick with the orthodox symbols. Q.E.D. :)",
"Solution_10": "Q.E.D doesn't look that nice in print, but so does the halmos symbol. In my opinion.",
"Solution_11": "Mindspa, is that a photo of you?",
"Solution_12": "Yes, I think. He looks Swedish enough to me! :)\r\nOn the same note, tokenadult...*abruptly stops talking* :D",
"Solution_13": "Yes, that's me.",
"Solution_14": "I feel kind of weird after thinking for like 2 month that you're Indian...\r\nIt's kind of odd when an avatar suddenly goes from Indian to Swedish... :)\r\n\r\nOh well, better than being mistaken for a girl [b]6[/b], count'em, [b]6[/b] times because of avatar :lol: (no offense to girls, just trying to make a point.)",
"Solution_15": "that avatar that mindspa had was of ramujan (sorry if i misspelled it, especially since i am indian), i thought mindspa was indian also, but then i was corrected a few months ago",
"Solution_16": "Yes, I did recognize his face because his picture was in my textbook.\r\nbut still, since avatar represents who you are, I thought he was Indian.\r\n\r\nI was also surprised when you told me you were Indian. I thought you were American (or Caucasian) because of your avatar.\r\nSo it's kind of like a flip between you two. :)",
"Solution_17": "Well probably no one had any doubt on what nationality I am :D",
"Solution_18": "lol\r\n\r\n\r\ntare: my avatar is from the game chrono cross, the main character serge",
"Solution_19": "lol[b]t[/b] :) \r\nit'd be funni[b]er[/b] if you weren't actually Chinese :lol: :P \r\n\r\nI have never heard of that game. My philosophy on games, music, movies, etc.: Watch it, then decide whether you want to pay or not. :twisted:",
"Solution_20": "tare i thought you were a girl for a really long time. especially from the pumpkin with pi carved out, that made me particularly confident.",
"Solution_21": "Um, okay...the toll is up to nine now \n\n[hide]at this rate, I could really fool a lot of people...no one knows that I'm a...[/hide]\n\n\n\nbesides, the pumpkin was made by a male sculptor, along with some other pumpkins i can not discuss here. It was pretty famous a long time ago when someone spammed it all over the place, but now, no ones knows the legacy"
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "Give $ k \\equal{} \\frac {n \\plus{} 1}{\\sqrt {n^2 \\minus{} 2n \\minus{} 1}}$ with $ n \\in N*, n > 2$\r\nand $ x_1,..,x_n$ are $ n$ positive numbers such that $ x_i \\in [\\frac {1}{k},k]$\r\nProve that $ \\frac {x_1}{x_2 \\plus{} x_3} \\plus{} \\frac {x_2}{x_3 \\plus{} x_4} \\plus{} ... \\plus{} \\frac {x_n}{x_1 \\plus{} x_2} \\geq \\frac {n}{2}$",
"Solution_1": "Too hard ??"
}
{
"Tag": [
"analytic geometry",
"LaTeX"
],
"Problem": "The point $A(4, 0)$ is a vertex of regular hexagon $ABCDEF$, whose side is 8 and whose interior lies completely within Quadrant I. If vertex $D$ has coordinates $(x, y)$, compute the ordered pair $(x, y)$.",
"Solution_1": "[hide]\nIn order for all interior of hexagon to be in Q1, B has to be at least on x-axis and that force B(or F) to be on (0,4sqrt3)\n\nD is the one opposite of A, so it is at (12, 8sqrt3)\n\n\nsorry I cant use LaTex YET.\n\n[/hide]",
"Solution_2": "[hide]the hexagon can only fit if it has one leg along the x axis and one vertex touching the y axis, forming a 30-60-90 right triangle with legs $4, 4\\sqrt{3}, 8$. $D$ is twice the height of the triangle up, and 8 across, so $(4,0)+(8,8\\sqrt{3})=\\boxed{(12,8\\sqrt{3})}$[/hide]\r\n\r\nah, i see geehoon beat me to it :lol:",
"Solution_3": "Okay, you guys are now certifiably ready to succeed at NYSML. Break a leg!"
}
{
"Tag": [
"blogs"
],
"Problem": "Okay so whats the deal with people wanting to run? \r\n\r\nSomebody mentioned that a mentor might lead runs early in the morning?\r\n\r\nI was hoping to do around 35 miles a week. Maybe more if the heat isn't so bad.",
"Solution_1": "I'm up for it. Though I'm rather out of shape; I haven't run in the past week or so due to various reasons.",
"Solution_2": "you guys like running a lot, don't you",
"Solution_3": "[quote=\"mdk\"]you guys like running a lot, don't you[/quote]\r\nWhen whatever God you believe in was handing out brains, you must have thought he said trains and asked for a slow one.\r\n\r\n*zing*",
"Solution_4": "LOL...\r\n\r\nbut a bit harsh.\r\n\r\nI like running to the extent that I want to do it consistently. The main reason that I run is to stay in shape. Also my friends are destroying me in summer XC [in terms of training]. So I want to at least try to 'keep up.' Also running feels good...at times...I also want to have a good XC season. I did well as a soph [like top 10 at my school for my age]. But then as a junior I was lazy as hell. Now I want to at least be average. Hence, I want to do some serious running over the summer.",
"Solution_5": "blargh I still don't get why you would want a slow train\r\n\r\nlike HEY THERE'S A TRAIN I HOPE IT'S PRETTY SLOW",
"Solution_6": "I actually like running and run at least 5 fives a week, otherwise how can I play 30 minutes every game and in the track season win about every race I participate in.",
"Solution_7": "haha, I'm down for early morning runs... just hope it doesn't get too hot. ><",
"Solution_8": "oh trust me. it will be too hot. you'll just have to deal.",
"Solution_9": "Well, I went last year and there was one mentor who brought us over to a cement loop about a third of a mile long and we ran there. It was like from 7:15- 7:45 and then class started 9:00. I'm actually not too sure about that time frame there... It wasn't too hot in the morning... only 98 degrees by the end of the run :P . And by the end of the three weeks last year, it was just me running... Anyways, I need to run everyday probably since I'm, well, a track/xc devotee... I'm actually considering running after dinner from 7:00-8:30 because I might need at least an hour for running and I'm not the most fond of waking up crazy early. Yup. That's it. I'm not the greatest at math, but that's ok. :lol:",
"Solution_10": "@humDdum: Well, I am motivated to run the whole 3 weeks. So you won't be alone. I am not too amped for running in a circle everyday. I'd like to run like 5 miles a day-that translates into 15 laps. Also cement is not so good for xc training...we need grass or something. Running late at night might work...but I am not sure. \r\n\r\nI'm doing okay in like 85 degrees here.\r\n\r\nAND I GOT TOTALLY SOAKED TODAY RUNNING. It was like...hey its raining...I am 2 miles away from home...",
"Solution_11": "[quote=\"Altheman\"]@humDdum: Well, I am motivated to run the whole 3 weeks. So you won't be alone. I am not too amped for running in a circle everyday. I'd like to run like 5 miles a day-that translates into 15 laps. Also cement is not so good for xc training...we need grass or something. Running late at night might work...but I am not sure. \n\nI'm doing okay in like 85 degrees here.\n\nAND I GOT TOTALLY SOAKED TODAY RUNNING. It was like...hey its raining...I am 2 miles away from home...[/quote]\r\n\r\nSo one day, I'm like a mile from home, right. \r\n\r\nAnd there is this one street that goes straight to my house.\r\n\r\nAnd on the side of the street I was downwind from, it was garbage night. I still have the shirt I wore, and it still smells.",
"Solution_12": "[quote=\"PenguinIntegral\"][quote=\"Altheman\"]@humDdum: Well, I am motivated to run the whole 3 weeks. So you won't be alone. I am not too amped for running in a circle everyday. I'd like to run like 5 miles a day-that translates into 15 laps. Also cement is not so good for xc training...we need grass or something. Running late at night might work...but I am not sure. \n\nI'm doing okay in like 85 degrees here.\n\nAND I GOT TOTALLY SOAKED TODAY RUNNING. It was like...hey its raining...I am 2 miles away from home...[/quote]\n\nSo one day, I'm like a mile from home, right. \n\nAnd there is this one street that goes straight to my house.\n\nAnd on the side of the street I was downwind from, it was garbage night. I still have the shirt I wore, and it still smells.[/quote]\r\n\r\nWOW THAT IS PRETTY INTERESTING",
"Solution_13": "I'm a sprinter first and foremost, but I'm doing cross country next season in high school so I'm considering running 3 miles a day... if I do I will SUCK though. Wait so are you saying we should like run in a group? Too many different ages... \r\n\r\nAltheman, what grade are you going intO?",
"Solution_14": "Yea...I am not sure if group running is going to fly... I am almost sure that I will be running shirtless everyday [not only to conserve shirts, but because it will be hot]. I agree that the speed differences are hard for a group run. I also generally like to run alone unless it is like 90+ mins. \r\n\r\nI will be a senior in high school. I am like average on our team. At the end of xc season last year, my time was like 19:02 for 3 miles [which was worse than my sophomore time of 18:30].\r\n\r\nOh and I was running today in like 90 degree weather. When I was going out, it wasn't so bad because there was a breeze. But if you are running in the heat with the sun beating down on you with no breeze, it is pretty bad. I had to cut my run short by like half a mile and just walk home.",
"Solution_15": "I think it should be broken up into semi-equal sections, so that we do a loop of a landmark or something than come back to campus, so some people can stop, then the rest go and do a loop around another landmark, and come back and let people off, etc.\r\n\r\n\r\nAlso, another stipulation : if we, being manly math runners, run with our shirts off and hot girls chase after us at the sight of our six packs and huge biceps, and are like \"Hey I haven't seen you around here, why are you here?\", the answer is NOT \"math camp\".",
"Solution_16": "I can do a mile in 3:20.\r\n\r\n\r\nThat's 3 hours and 20 minutes, half way through I usually set camp and cook a meal.",
"Solution_17": "Why the - is this forum filled with totally unhelpful posts? Like we are actually trying to figure stuff out and then we just get cluttered with spam. Direct spam to the blogs :D",
"Solution_18": "I have not heard of a mentor doing running. Last year at AMP there were no mentors, only 1-4 students running.\r\n\r\nAnyways, I'll run in the mornings.",
"Solution_19": "If we were to run, I'd prefer like 6:30 to 7:00 or something obscenely early if possible... \r\n\r\nanyhow just ran two miles, 15:25 =/ so not sure how many people will be that slow.",
"Solution_20": "I really should get a watch to like give myself a wake-up call for my slow ---. Anyway, I think it is pretty obvious that we need to have it so people can just stop. The thing is that there is no shame in stopping if you are overheating. Like usually it is good to push yourself when running, but pushing yourself in heat is the major exception. \r\n\r\nI have been reading a bit about running in the heat. I guess the main point is that [b]you need to hydrate well[/b]. That means like drinking like a pint of water before you start. Then drinking some water or w/e every 15-20 mins. Other considerations are: a light sunscreen like SPF 15, avoiding direct sunlight, wearing sunglasses, wearing special running clothing. In my totally unqualified opinion, these last few things are not that important since I run without them and I am fine. To jump on the 'PenguinIntegral' bandwagon, we might wear sunglasses to look awesome. :lol:\r\n\r\nI guess for timing, we should get out pretty early. I might just like start running at like 7:15. I'd like to find a mile loop or something. That way I could just like run 5 of them and be done. Obviously that would get boring, but we will see.",
"Solution_21": "it looks like the first week is going to be nice and cool (low 90s), but I am sure that it will get into the 100's by the time we leave (yay Texas weather)",
"Solution_22": "Actually it looks like it's going to be Seattling (raining) on us for the first week or so as it's been doing almost all summer.Hydration is important, and I would definitely recommend some sort of wicking material to run in. No sunscreen should ever be necessary unless you're albino, and if you're running in the morning, chances are the sun will be staring at you or, more likely, vice versa, so sunglasses will come in handy.",
"Solution_23": "My violin teacher recommends we bring boric acid to kill giant tree roaches.\r\n :huh:",
"Solution_24": "no roaches in Dallas.",
"Solution_25": "sweet. That's reassuring. Math is hard enough, let alone math while killing giant bugs.\r\nthanks =)",
"Solution_26": "sorry to interrupt, but that is pretty slow",
"Solution_27": "I'm starting to think that night running will definitely be better than morning running. Definitely.",
"Solution_28": "mdk, were you talking to me?\r\n\r\n14:29 yesterday... dunno how i got so much faster. first day my knees hurt though, so bad I had to run on the grass for a few stretches of like 100 meters.\r\n\r\nEdit: fine I know there are like beasts that do 3 miles in 16 minutes, but\r\nA) I'm a sprinter\r\nB) I'm entering 9th grade",
"Solution_29": "I e-mailed the UTD xc coach. Provided that he responds, I will do what he suggests.",
"Solution_30": "In my school's xc league, 16 minutes for 3 miles is quite fast for guys, 19 minutes is quite fast for girls. We have four or five guys at 16:20 and three or four girls under 19.",
"Solution_31": "hey, that's like my MC countdown speed :D\r\nthree problems in 16 minutes is pretty good xP",
"Solution_32": "maybe I can get someone to sneak us out on wednesdays to run with the greenhill team...zach maybe?",
"Solution_33": "btw im a girl. just thought you needed to know. :whistling: shirts on would be quite nice.",
"Solution_34": ":D mkay, no i did not realize that\r\n\r\nco-ed running doesn't tend to fly...as i mentioned earlier...but w/e, we'll see",
"Solution_35": "[quote=\"Fanatic\"]mdk, were you talking to me?\n\n14:29 yesterday... dunno how i got so much faster. first day my knees hurt though, so bad I had to run on the grass for a few stretches of like 100 meters.\n\nEdit: fine I know there are like beasts that do 3 miles in 16 minutes, but\nA) I'm a sprinter\nB) I'm entering 9th grade[/quote]\r\n\r\nyes, I was talking to you, but that's okay, you're still young"
}
{
"Tag": [
"factorial"
],
"Problem": "The are x apples on a tree. Jack takes (x-3)! - 118 apples. Jill then takes half of the remaining. The number of apples remaining is (x/2)^(1/2) + 1.\r\n\r\nHow many apples were originally on the tree.\r\n\r\nTry not to use brute force to solve this. It is about a #12 or #13 on the AMC 12 level question, so not very hard or very easy, but some where in the middle and requires work to do it without simply plugging in numbers and is just for fun. I'll give the answer after people have a jab at it.",
"Solution_1": "My guess [hide=\"is\"]8. Since x-3 has to be 5 if the equation at least makes sense, there has to be 8 there originaly. We can check: (x/2)^(1/2) + 1=(8/2)^(1/2)+1=4^(1/2)+1=3, which is 6/2 as well.[/hide]\r\n\r\nEDUCATED guess.",
"Solution_2": "[quote=\"neelnal\"]The are x apples on a tree. Jack takes (x-3)! - 118 apples. Jill then takes half of the remaining. The number of apples remaining is (x/2)^(1/2) + 1.\n\nHow many apples were originally on the tree.\n\nTry not to use brute force to solve this. It is about a #12 or #13 on the AMC 12 level question, so not very hard or very easy, but some where in the middle and requires work to do it without simply plugging in numbers and is just for fun. I'll give the answer after people have a jab at it.[/quote]\r\n\r\n[hide]The answer has to be at least 8, because $4!$<118, and we can't have negative apples. I quickly see that 5 is the answer, because $5!=120$, $120-118=2$ $8-2=6$, $6/2=3$, $8-2-3=3$,\n$8/2=4$, 4^1/2=2, $2+1=3$[/hide]\r\n\r\nDoes that count as brute force? Sorry :rotfl:",
"Solution_3": "mustafa I was wondering could you calculate 6! as well?\r\nAs it comes to 720, which means 602were taken; which is over the original number and as the factorial increases, it increases more. Also, as the answer cannot be below 5! and the factorial has to be an integer as there is no fraction or negative factorial, the only answer is 5! during the working?\r\nJust a thought. This may mix up the problem but definitely proves the solution. This doesn't seem to be that brutal as we are using numbers given to us, not guessing random digits. Also this and your solution shows workings.\r\n\r\nI'm looking forward to the solution you have neelnal.",
"Solution_4": "mustafa's got it.",
"Solution_5": "[quote=\"neelnal\"]The are x apples on a tree. Jack takes (x-3)! - 118 apples. Jill then takes half of the remaining. The number of apples remaining is (x/2)^(1/2) + 1.\n\nHow many apples were originally on the tree.\n\nTry not to use brute force to solve this. It is about a #12 or #13 on the AMC 12 level question, so not very hard or very easy, but some where in the middle and requires work to do it without simply plugging in numbers and is just for fun. I'll give the answer after people have a jab at it.[/quote]\r\n[hide=\"non brute force\"]\n$\\frac{(x-3)!-118}{2}=\\sqrt{.5x}+1$\n$(x-3)!-118=\\sqrt{2x}+2$\n$(x-3)!-120-\\sqrt{2x}=0$\nat which point we go to the graphing calculator :P[/hide]\r\nEDITED SOL. Just realized I made an error",
"Solution_6": "Or we just note that $\\sqrt{{x\\over 2}}$ must be integer, hence $x$ is of the form $2k^{2}$. For $k=1$ the number is too small, for $k=3$ is already too large ($(x-3)!=15!$), and $k=2$ is just right.\r\n\r\n(Boy, does this sound like Goldilocks...)",
"Solution_7": "[quote=\"Farenhajt\"]Or we just note that $\\sqrt{{x\\over 2}}$ must be integer, hence $x$ is of the form $2k^{2}$. For $k=1$ the number is too small, for $k=3$ is already too large ($(x-3)!=15!$), and $k=2$ is just right.\n\n(Boy, does this sound like Goldilocks...)[/quote]\r\n\r\n :rotfl:"
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "Let $a,$ $b$ and $c$ are positive numbers. Prove that\r\n$\\sqrt[3]{2(a+b+c)\\left(\\frac{1}{a}+\\frac{1}{b}+\\frac{1}{c}\\right)+9}\\geq\\sqrt[3]{\\frac{a}{b}}+\\sqrt[3]{\\frac{b}{c}}+\\sqrt[3]{\\frac{c}{a}}.$",
"Solution_1": "Cubing both sides with $x^{3}= \\frac{a}{b}$, etc, reveals that it is equivalent to \r\n$\\sum x^{4}yz+2x^{3}y^{3}+3x^{2}y^{2}z^{2}\\geq \\sum 6x^{3}y^{2}z$, where sums are symmetric, i.e. Schur times $xyz$ and Schur on $xy, xz, yz$, so it seems pretty strong; not sure how to actually do the problem yet, of course (I'm playing around with weighted Jensen on $f(t) = \\sqrt[3]{t}$?)",
"Solution_2": "[quote=\"MysticTerminator\"] \n$\\sum x^{4}yz+2x^{3}y^{3}+3x^{2}y^{2}z^{2}\\geq \\sum 6x^{3}y^{2}z$[/quote]\r\nR.H.S. must be $\\sum_{sym}3x^{3}y^{2}z$.",
"Solution_3": "[quote=\"mathpk\"][quote=\"MysticTerminator\"] \n$\\sum x^{4}yz+2x^{3}y^{3}+3x^{2}y^{2}z^{2}\\geq \\sum 6x^{3}y^{2}z$[/quote]\nR.H.S. must be $\\sum_{sym}3x^{3}y^{2}z$.[/quote]\r\n\r\nNo, MysticTerminator is right. Note that LHS has sym sum also, not cyc sum.",
"Solution_4": "Oops sorry :blush:",
"Solution_5": "How to do this one arqadi? :D",
"Solution_6": "[quote=\"mathpk\"]How to do this one arqadi? :D[/quote]\r\nIn fact, MysticTerminator has solved this problem and in accuracy by mine way: \r\n$\\sqrt[3]{2(a+b+c)\\left(\\frac{1}{a}+\\frac{1}{b}+\\frac{1}{c}\\right)+9}\\geq\\sqrt[3]{\\frac{a}{b}}+\\sqrt[3]{\\frac{b}{c}}+\\sqrt[3]{\\frac{c}{a}}\\Leftrightarrow$\r\n$\\Leftrightarrow\\sum_{cyc}\\left(\\frac{a}{b}+\\frac{2a}{c}+3\\right)\\geq3\\cdot\\sum_{cyc}\\left(\\sqrt[3]{\\frac{a^{2}}{bc}}+\\sqrt[3]{\\frac{bc}{a^{2}}}\\right)\\Leftrightarrow$\r\n$\\Leftrightarrow\\sum_{cyc}(x^{3}-x^{2}y-x^{2}z+xyz)+2\\cdot\\sum_{cyc}(u^{3}-u^{2}v-u^{2}w+uvw)\\geq0,$\r\nwhere $x=\\frac{1}{u}=\\sqrt[3]{\\frac{a}{b}},$ $y=\\frac{1}{v}=\\sqrt[3]{\\frac{b}{c}}$ and $z=\\frac{1}{w}=\\sqrt[3]{\\frac{c}{a}}.$ :)",
"Solution_7": "I think, the problem for two variables in very famous, it's from Hojoo Note!?!"
}
{
"Tag": [
"geometry",
"trigonometry"
],
"Problem": "$ PQR$ is a triangle with $ PQ = 33$, $ PR = 21$ and $ QR = m$, an integer.\r\n\r\nThere are points $ D$, $ E$ on the sides $ PQ$, $ PR$ respectively such that $ PD = DE = ER = n$, an integer.\r\n\r\nFind $ m$.",
"Solution_1": "The answer is 30.",
"Solution_2": "[quote=\"zhou_qiang\"]The answer is 30.[/quote]\r\n\r\ndo you care to show the method?",
"Solution_3": "[quote=\"zhou_qiang\"]The answer is 30.[/quote]\r\n\r\nPlease show why it has to be 30, since I haven't found an answer yet.",
"Solution_4": "[quote=\"xxxyyyy\"]$ PQR$ is a triangle with $ PQ = 33$, $ PR = 21$ and $ QR = m$, an integer.\n\nThere are points $ D$, $ E$ on the sides $ PQ$, $ PR$ respectively such that $ PD = DE = ER = n$, an integer.\n\nFind $ m$.[/quote]\r\n\r\nConsider the triangle $ PDR$ with $ PD=DE=ER=n$ and $ PR=q$. \r\nLet $ \\theta = \\angle ERD = EDR$. Then $ \\angle DEP = \\angle DPE = 2\\theta$\r\n\r\n[img]10058[/img]\r\nLet $ F\\in PD$ such that $ RF=RP$. Then $ \\triangle RPF \\sim \\triangle DPE\\Rightarrow \\frac{PF}{RP}= \\frac{PE}{DP}\\Rightarrow \\frac{PF}{q}= \\frac{q-n}{n}\\Rightarrow$\r\n\r\n$ PF = q\\cdot \\frac{q-n}{n}= \\frac{q^{2}}{n}-q \\ \\ (1)$\r\n\r\n[img]10059[/img]\r\n\r\nLet $ w$ be the circle centered at $ R$ and with radius $ q$\r\n\r\nLet $ Q$ on the ray $ PF$ and set $ QP=r$ and $ QR=m$\r\n\r\n$ \\blacksquare$ [color=blue][b]case 1)[/b] $ Q$ is in the extension of the segment $ PF$[/color]\r\n\r\nThen $ QF \\cdot QP = QR^{2}-q^{2}\\Rightarrow QF = \\frac{m^{2}-q^{2}}{r}$\r\n\r\n$ PF = PQ-QF \\Rightarrow PF = r+\\frac{q^{2}-m^{2}}{r}$\r\n\r\n\r\n$ \\blacksquare$ [color=blue][b]case 2)[/b] $ Q$ is in the interior of the segment $ PF$[/color]\r\n\r\nThen $ QF \\cdot QP =-(QR^{2}-q^{2}) \\Rightarrow QF = \\frac{q^{2}-m^{2}}{r}$\r\n\r\n$ PF = PQ+QF \\Rightarrow PF = r+\\frac{q^{2}-m^{2}}{r}$\r\n$ \\hline$\r\nAt any position of $ Q$ on the ray $ PF$ we have $ \\boxed{PF = r+\\frac{q^{2}-m^{2}}{r}}\\ \\ (1)$ (even if $ Q\\equiv F$)\r\n\r\nFrom $ (1)$ and $ (2)$ we find that\r\n\r\n$ \\frac{q^{2}}{n}-q = r+\\frac{q^{2}-m^{2}}{r}\\Rightarrow$\r\n\r\n$ q^{2}r-qnr = r^{2}n+(q^{2}-m^{2})n\\Rightarrow$\r\n$ q^{2}r = (r^{2}+q^{2}+qr-m^{2})n\\Rightarrow \\boxed{n = \\frac{q^{2}r}{r^{2}+q^{2}+qr-m^{2}}}\\ \\ (3)$\r\n$ \\hline$\r\n\r\nApplying to $ (3)$ the data of the problem we get that $ n = \\frac{21^{2}\\cdot 33}{33^{2}+21^{2}+21\\cdot 33-m^{2}}= \\frac{14553}{2223-m^{2}}$\r\n\r\nSo we have to find $ m$ such that $ 2223-m^{2}\\mid 14553$\r\n\r\nObserving within the possible values of $ m\\in\\{13,14,...,53\\}$ we find that only $ m=30$ gives solution, with corresponding $ n=11$",
"Solution_5": "First use cosine rule. We will get n^2=n^2+(21-n)^2-2(n)(21-n)cos(angleP). Evaluate the expression of cosP in terms of n.Then apply cosine rule again to tiangle PQR, and since m is an integer, we can factor it out."
}
{
"Tag": [
"AMC",
"AMC 10"
],
"Problem": "How did NJ'ers do?\r\n\r\nI got 139.5 (amc 10a) and 6 aime.",
"Solution_1": "126/8=206 :mad: \r\n\r\n2 perfect scorers on AMC10 A were both from AAST :D",
"Solution_2": ":| 133.5/10 .. kind've disappointed as a senior"
}
{
"Tag": [
"inequalities",
"inequalities open"
],
"Problem": "Hi,\r\n\r\ncan someone suggest me a text or a book which introduces to the different topics of inequalities?",
"Solution_1": "Here is it :wink: \r\n\r\nhttp://www.animath.fr/"
}
{
"Tag": [
"Support",
"geometry",
"parallelogram"
],
"Problem": "Sa se construiasca cu ajutorul unei rigle si al unui compas triunghiul [b]ABC[/b] daca se stiu pozitiile a trei puncte: [b]D[/b] , mijlocul inaltimii [b]AH[/b], [b]E[/b], mijlocul inaltimii [b]CP[/b] si [b]M[/b] mijlocul laturii [b]AC[/b]",
"Solution_1": "@mihneaA: double post :wink: Ar trebui sa stergi unul din ele\r\nProblema pare interesanta :)",
"Solution_2": "[hide=\"cheia problemei\"]diametralul opus lui $M$ in cercul $MDE$ este ortocentrul lui $\\triangle{ABC}$ si unghiul $\\angle{DME}=\\angle{ABC}$, de aici constructia se usureaza considerabil.[/hide]\n\n[hide=\"solutie completa\"]acum deci se cunosc punctele $D$, $E$, $M$ si $X$, cu $X$ ortocentru, deci se cunosc dreptele suport ale inaltimilor din $A$ si $C$, adica $DX$, respectiv $EX$., dar $M$ mijloc lui $AC$, deci putem gasi $A$ si $C$: fie $Y$ simetricul lui $X$ fata de $M$, si astfel prin $Y$ ducem paralela la $DX$ care taie $EX$ in $C$, si paralela la $EX$ care taie $DX$ in $A$ (fapt dovedit de faptul ca $AXCY$ este paralelogram , prin constructie).\n\nsi s-a cam terminat acum, ducem paralele din $A$ respectiv $C$ la $EM$ , respectiv $DM$ , care se vor intersecta in $B$. astfel se respectva si conditia ca $D,E$ sunt mijloace, caci sunt acelea linii mijlocii, si ca $\\angle{ABC}=\\angle{DME}$.\nPS: in rezolvare am folosit ca $DM \\perp DX$, $EM\\perp EX$ , rezultat evident.[/hide]"
}
{
"Tag": [
"combinatorics unsolved",
"combinatorics"
],
"Problem": "Find the sum of all 3-digit natural numbers which contain at least one odd digit and at least one even digit.",
"Solution_1": "denote them by $ a_1a_2a_3$\r\nnow try to find how many times each integer appears as $ a_1$ , $ a_2$ and $ a_3$ the rest is easily done since $ a_1a_2a_3\\equal{}10^2a_1\\plus{}10a_2\\plus{}a_3$",
"Solution_2": "the answer is 370775",
"Solution_3": "i dink i got 448800 as the answer",
"Solution_4": "Sum of all $ \\ 3$ digit numbers $ \\equal{}\\frac {900}{2}(100\\plus{}999)\\equal{}494550$\r\n\r\nSum of all $ \\ 3$ digit numbers in which all digits are even is\r\n\r\n$ 10^2 \\times 25\\times (2\\plus{}4\\plus{}6\\plus{}8)\\plus{}10 \\times 20\\times (2\\plus{}4\\plus{}6\\plus{}8)\\plus{}20\\times (2\\plus{}4\\plus{}6\\plus{}8)\\equal{}54400$\r\n\r\nSum of all $ \\ 3$ digit numbers in which all digits are odd is\r\n\r\n$ 10^2 \\times 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\plus{}10 \\times 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\plus{} 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\equal{}69375$\r\n\r\nHence required sum is\r\n\r\n$ \\equal{}494550\\minus{}54400\\minus{}69375\\equal{}\\boxed{370775}$\r\n :)",
"Solution_5": "I got 370775 as the answer too \r\nI summed up all 3 digit numbers and subtracted all numbers with 3 even and 3 odd.\r\nI did by listing down every number with 3 even/odd digits out of each 100 numbers i.e 100 - 199 n 200 - 299 n so on in the RMO anwer sheet. Then for summing up I made into an AP with additional numbers n summed it up. Then I subtracted from TOTAL sum. \r\n\r\nThe question is will I get Marks for this trial n error sort of method?",
"Solution_6": "Hey makar i am new to all this combinatorics so could you help an ignorant guy like me understand how you summed up all those numbers with only odd digits. It would really help. thanks\r\nI skipped that problem anyway on the exam",
"Solution_7": "[quote]Hey makar i am new to all this combinatorics so could you help an ignorant guy like me understand how you summed up all those numbers with only odd digits. It would really help. thanks \nI skipped that problem anyway on the exam[/quote]\r\nrequired sum of numbers= sum of all 3 digit numbers-sum of 3 numbers with only odd digits-sum of 3 digit numbers with only even digits",
"Solution_8": "[quote=\"aadil\"][quote]Hey makar i am new to all this combinatorics so could you help an ignorant guy like me understand how you summed up all those numbers with only odd digits. It would really help. thanks \nI skipped that problem anyway on the exam[/quote]\nrequired sum of numbers= sum of all 3 digit numbers-sum of 3 numbers with only odd digits-sum of 3 digit numbers with only even digits[/quote]\r\n\r\nsee i got that far i just didnt get how to calculate the sum of all the numbers with only odd digits and only even digits. :oops:",
"Solution_9": "arrange the numbers with only even digits in a column.there are 4*5^2 such numbers.now if you consider the sum of the one's digits,it will contain the same no of 2s,4s,6s,8s,0s.therefore the sum is 4*5(2+4+6+8+0).similarly the sum of the ten's digits is 4*5(2+4+6+8+0) and sum of the hundred's digits is 5*5(2+4+6+8).\r\nso the number is 54400",
"Solution_10": "[quote=\"Manmather\"]Hey makar i am new to all this combinatorics so could you help an ignorant guy like me understand how you summed up all those numbers with only odd digits. It would really help. thanks\nI skipped that problem anyway on the exam[/quote]\r\n\r\nTo sum all three digit number in which all digits are odd:\r\n\r\nOdd digits can be $ \\ 1,3,5,7,9$ \r\n\r\nNow calculate how many times 1 will occur at unit place :idea: \r\nAs $ \\ 100^{th}$ place can be filled in $ \\ 5$ ways (by using 1 or 3 or 5 or 7 or 9) and $ 10^{th}$ place also can be filled in $ \\ 5 $ ways $ \\implies 1$ will occur at unit place $ 5\\times 5\\equal{}25$ times \r\n\r\nSimilarly $ \\ 3,5,7,9$ each will occur $ \\ 25$ times at unit place\r\n\r\nHence sum of digits at unit place in all such numbers will be $ 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)$ :idea: \r\nSimilarly we can find sum at $ \\ 10^{th}$ place and $ \\ 100^{th}$ . obviously it will be same as $ 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)$\r\n\r\nFinally sum of all such numbers in which all digits are odd is\r\n\r\n$ 10^{2}\\times 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\plus{}10\\times 25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\plus{}25\\times (1\\plus{}3\\plus{}5\\plus{}7\\plus{}9)\\equal{}69375$\r\n\r\nSimilarly you can find sum of numbers in which all digits are even\r\nin this case because of digit $ \\ 0$ you have to take some precaution basically $ \\ 0$ can not occur at $ \\ 100^{th}$ place\r\n\r\nI think this will clear your doubt\r\n :wink:",
"Solution_11": "thanks alot everyone"
}
{
"Tag": [
"geometry"
],
"Problem": "Compute the area of the circle that passes through all the intersection poits of 4x 2 +11y 2 =29 and x 2 -6y 2 =6.\r\n\r\nThanks!",
"Solution_1": "[hide]\nAdd the equations to get $x^2 + y^2 = 7$, so the area is $7\\pi$\n[/hide]"
}
{
"Tag": [
"real analysis",
"advanced fields",
"advanced fields unsolved"
],
"Problem": "Let $ m^*$ be the Lebesgue outer measure. Then one can show that it is \"subadditive\" or that $ m^*(\\bigcup A_i)\\leq \\sum m^*(A_i)$. Is there an easy case where equality doesn't hold? My book does not have any example of this and intuitively it seems as if it equality [i]should[/i] hold...",
"Solution_1": "For an \"easy\" case in which equality does not hold, take $ A_i$ that are not pairwise disjoint (and some of the overlaps have positive measure).\r\n\r\nIf you add \"pairwise disjoint\" as a hypothesis, then equality holds when the $ A_i$ are measurable. To get a case in which equality does not hold, you will need non-measureable sets.",
"Solution_2": "Oh duh! I was thinking that in the hypothesis they were disjoint for some reason. Thanks."
}
{
"Tag": [
"function",
"topology",
"algebra unsolved",
"algebra"
],
"Problem": "Let $f: R \\to R$ be a continuous function such that : $\\forall a1$.\r\n\r\n\\[ \\begin{aligned} 2^n-1=2^{ab}-1&=(2^a-1)(2^{a(b-1)}+2^{a(b-2)}+\\cdots+2^{a(b-(b-1))}+1)\\\\ &=(2^b-1)(2^{b(a-1)}+2^{b(a-2)}+\\cdots+2^{b(a-(a-1))}+1)\\\\ \\end{aligned} \\]\r\n\r\nHence, $2^n-1$ is composite.",
"Solution_5": "I still don't really get it... :?",
"Solution_6": "In the problem, he states that if $n$ is composite, it can be represented by the product of two factors $a*b$. He then writes $2^n - 1$ as $2^{ab} - 1$ which can be factored as he demonstrated, meaning that $2^n-1$ cannot be prime. This means that for $2^n-1$ to be prime, $n$ must also be prime.\r\n\r\nAnd the way he factored it, was basically using the fact that\r\n$a^n - b^n = (a-b)(a^{n-1} + a^{n-2}b + \\dots - b^{n-1})$\r\n\r\nSo you can write\r\n$(2^{a})^b - 1^{b} = (2^{a}-1)((2^{a})^{b-1} - (2^{a})^{b-2}(1) + \\dots - 1^{b-1})$"
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "$a,b,c>0$ and $a+b+c=3$ Prove that\r\n$\\frac{a}{b^2+1}+\\frac{b}{c^2+1}+\\frac{c}{a^2+1}\\geq\\frac{3}{2}$",
"Solution_1": "See http://www.mathlinks.ro/Forum/viewtopic.php?t=13943 .\r\n\r\n darij",
"Solution_2": "Thank you."
}
{
"Tag": [
"LaTeX",
"inequalities unsolved",
"inequalities"
],
"Problem": "for all $ a,b,c,d$ in $ (0,1/2]$\r\npf that\r\n$ abcd$/$ (1 \\minus{} a)(1 \\minus{} b)(1 \\minus{} c)(1 \\minus{} d)$ $ < \\equal{}$ $ a^4 \\plus{} b^4 \\plus{} c^4 \\plus{} d^4/(1 \\minus{} a)^4 \\plus{} (1 \\minus{} b)^4 \\plus{} (1 \\minus{} c)^4 \\plus{} (1 \\minus{} d)^4$\r\nsrry for my bad latex skill",
"Solution_1": "See here:\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?highlight=2002+taiwan&t=41543"
}
{
"Tag": [
"MATHCOUNTS",
"induction",
"quadratics"
],
"Problem": "hmmmm. im soo bored that im now posting questions. what a sin. lool. \n\n\n\non to the question in the not so exact words\n\n\n\nx+1/x=4\n\nwhat does x:^3:+1/x:^3: equal\n\n\n\nit caused quite a hustle bustle on the late mathcounts forum. i have no idea why, it wasnt hard. especially since i think its in the aops book 1. at least i think. im pretty sure. if you want to know the answer, highlight below\n\n\n\n[hide](x+1/x):^3:=x:^3:+3x+3/x+1/x:^3:=64=x:^3:+3(x+1/x)+1/x:^3:=\n\nx:^3:+3(4)+1/x:^3:=64\n\nso...x:^3:+1/x:^3:=64-12=52. cool beans.[/hide]",
"Solution_1": "There was a similar problem at NYSML once.\r\n\r\nIf cos(x)+sec(x)=3, find (sec^10(x)+1)/sec^5(x). It has a solution like Syntax Error's question, and it also has a very elegant solution (that I found because I didn't want to multiply that much).",
"Solution_2": "how very interesting. god im soo bored.",
"Solution_3": "Use your boredom as a time to learn mathematics. That's one of my favorite time-spending methods.",
"Solution_4": "i am doing math. its not the math thats boring me. very interesting indeed. i just am a mix of tired hungry stressed angry, and i should do some other math problems. cause im doing recycled ones. hmmmmm. maths very fun, dont get me wrong. i love math",
"Solution_5": "Let x be a real number such that x + 1/x is an integer. Prove that x^n + 1/x^n is an integer, for all positive integers n. (Germany, 1995). This shouldn't be tough for most of you to show.",
"Solution_6": "How very interesting... yeah... I would use [hide]induction[/hide] <-- Highlight to find out... That should be all you need...\n\n\n\nPeace.\n\nIW.OAV",
"Solution_7": "In mindspa's question, doesnt x have to be 1? If it does, then that is really easy and doesnt even need induction.",
"Solution_8": "No, that is not necessary. Suppose that x+1/x=5, for example. Then (by solving a quadratic), we can have x=(5+sqrt(21))/2.",
"Solution_9": "[quote=\"Syntax Error\"]\nx+1/x=4\nwhat does x:^3:+1/x:^3: equal\n[/quote]\r\n\r\n[hide=\"solution\"]\n(x + 1/x)^3 = 4^3\nx^3 + 3x + 3x^-1 + x^-3 = 64\nx^3 + x^-3 + 3(x + 1/x) = 64\nx^3 + x^-3 + 12 = 64\nx^3 + x^-3 = 52\n[/hide]\r\n\r\nI love these kinds of questions in math competitions, as I know that some of my competitors will do it the long way xD ."
}
{
"Tag": [
"trigonometry",
"Asymptote",
"function",
"MIT",
"college",
"inequalities",
"algebra proposed"
],
"Problem": "Problem 1: Prove that the equation:\r\n\\[ 4^x(4x^2\\plus{}1)\\equal{}1\\]\r\nhave only 3 real different roots :) \r\n\r\nProblem 2: Solve\r\n\\[ \\frac{1}{1\\minus{}x^2}\\plus{}1>\\frac{3x}{\\sqrt{1\\minus{}x^2}}\\]",
"Solution_1": "Obviously $ x\\equal{}0$ is one root.\r\n\r\nA quick sketch of $ y\\equal{}4^{x}$ and $ y\\equal{}\\frac{1}{4x^{2}\\plus{}1}$ shows that there seem to be 2 other roots and that both of them are negative (about $ \\minus{}\\frac{1}{2}$ and $ \\minus{}2$, I think.)",
"Solution_2": "Using Rolle's theorem. I don't think that -2 is root.",
"Solution_3": "[quote=\"zaizai-hoang\"]Problem 1: Prove that the equation:\n\\[ 4^x(4x^2 \\plus{} 1) \\equal{} 1\n\\]\nhave only 3 real different roots :) \n[/quote]\nThere exist $ t \\in (\\minus{}\\frac{\\pi}{2},\\plus{}\\frac{\\pi}{2})$ s.t. $ x\\equal{}\\frac{\\tan(x)}{2}$. It implies $ 2^{\\tan(t)}\\equal{}\\cos^2(t)$.\nMay be helpful?\n\n\n[quote=\"zaizai-hoang\"]Problem 2: Solve\n\\[ \\frac {1}{1 \\minus{} x^2} \\plus{} 1 > \\frac {3x}{\\sqrt {1 \\minus{} x^2}}\n\\]\n[/quote]\r\n\r\nSince $ |x| < 1$ there exist $ t \\in [ \\minus{} \\pi,\\pi)$ s.t. $ x \\equal{} \\cos(t)$. Then $ \\frac {1}{1 \\minus{} x^2} \\plus{} 1 > \\frac {3x}{\\sqrt {1 \\minus{} x^2}} \\implies \\frac {1 \\plus{} \\sin^2(t)}{\\sin^2(t)} > 3 cotan(t)$. So $ 1 \\plus{} (\\frac {1 \\minus{} \\cos(2t)}{2}) > \\frac {3}{2}\\sin(2t) \\implies \\cos(2t) \\plus{} 3\\sin(2t) \\minus{} 3 < 0$ that can be rewritten as $ \\sin(2t \\plus{} arsin(\\frac {1}{\\sqrt {10}})) < \\frac {3}{\\sqrt {10}}$,from here it is straightforward.",
"Solution_4": "Again I've done a rough sketch.\r\n\r\n$ y\\equal{}\\frac{1}{1\\minus{}x^{2}} \\plus{} 1$ has asymptotes $ x\\equal{}\\pm1$ and $ y\\equal{}1$ and has a local minimum at $ (0, 2)$.\r\n\r\n$ y\\equal{} \\frac{3x}{\\sqrt{1\\minus{}x^{2}}}$ is only defined for $ \\minus{}1 \\dfrac{3\\sin\\theta}{\\cos\\theta}$ \n\n$ \\sec^2\\theta \\plus{} 1 > 3\\tan\\theta$ \n\n$ \\tan^2\\theta \\minus{} 3\\tan\\theta \\plus{} 2 > 0$\n\n$ (\\tan\\theta \\minus{} 1)(\\tan\\theta \\minus{} 2) > 0$ \n\n$ \\tan\\theta < 1$ OR $ \\tan\\theta > 2$ \n\n$ x \\equal{} \\sin\\theta < \\dfrac{1}{\\sqrt {2}}$ OR $ x \\equal{} \\sin\\theta > \\dfrac{2}{\\sqrt {5}}$ \n\nSo, the values of $ x$ that satisfy the inequality are $ \\boxed{x \\in \\left( \\minus{} 1,\\dfrac{1}{\\sqrt {2}}\\right) \\bigcup \\left(\\dfrac{2}{\\sqrt {5}},1\\right)}$\n[/hide]"
}
{
"Tag": [
"conics",
"ellipse",
"trigonometry",
"advanced fields",
"advanced fields unsolved"
],
"Problem": "Points P,Q and R lie on the ellipsoid of equation$\\frac{x^2}{a^2}+\\frac{y^2}{b^2}+\\frac{z^2}{c^2}=1$ such that segments OP, OQ, OR are pairwise perpendicular, where O denotes the centre of the ellipsoid. Prove that the distance of plane PQR from O is independent of the position of the points P, Q, R.",
"Solution_1": "oh ! why I posted it in this topic ,moderator plz replace it :( :blush:",
"Solution_2": "Basicly, we do this (no details for now :)):\r\n\r\nWe show that a similar thing happens on the ellipse: if it's centered at $O$ and $M,N$ are points on the ellipse s.t. $OM\\perp ON$, then the distance from $O$ to $MN$ is constant. Then we look at the intersection of a plane with the ellipsoid: this is an ellipse. Now take $Q,R$ on this ellipse s.t. $OQ\\perp OR$. In this case $OP$ is fixed because it passes through $O$ and it's $\\perp (OQR)$. The locus of $T$, the projection of $O$ on $QR$, is a circle (we use the property for the ellipse). $OP$ is perpendicular to the plane of this circle and passing through its center, so the locus of the projection of $O$ on $PT$ is a circle. \r\n\r\nThis means that each plane cuts the locus of the projection of $O$ on $(PQR)$ in a circle. It's well-known that this means that the locus is, in fact, a sphere.",
"Solution_3": "In order to prove the property for an ellipse, we just use the polar equation of the ellipse: $r^2=\\frac {a^2b^2}{b^2\\cos \\theta+a^2\\sin \\theta}$. It's easy to see that $\\frac 1{OM^2}+\\frac 1{ON^2}$ is constant, where $M,N$ are points on an ellipse centered at $O$ s.t. $OM\\perp ON$, and this, in turn, is equivalent to the stateed property: the distance $d(O,MN)$ is constant."
}
{
"Tag": [
"quadratics",
"modular arithmetic",
"number theory unsolved",
"number theory"
],
"Problem": "Let $p$ be a odd prime number and $m>1,h$ be natural numbers such that $h<2^m,n=2^mh+1$ if $n^{\\frac{p-1}{2}}\\equiv -1$ $(\\mod p)$ .prove that if $n$ wants to be a prime number It's nessecery and suffient that:\r\n\r\n \r\n $p^{\\frac{n-1}{2}} \\equiv -1$ $(\\mod n)$",
"Solution_1": "$n=4k+1$. If $n$ is a prime, then the quadratic reciprocity law says that $\\left(\\frac pn\\right)\\left(\\frac np\\right)=1$. Since $\\left(\\frac np\\right)=-1$, we get $\\left(\\frac pn\\right)=-1$, as desired.\r\n\r\nConversely, let $q|n$ be a prime. We have $p^{n-1}\\equiv 1\\pmod q$, and $p^{q-1}\\equiv 1\\pmod q$, so $p^{(n-1,q-1)}\\equiv 1\\pmod q$. If $2^m\\not|\\ q-1$, then $(n-1,q-1)=\\left(\\frac{n-1}2,q-1\\right)$, so $p^{\\frac {n-1}2}\\equiv 1\\pmod q$, contradiction. This means that $2^m|q-1$, so $q=2^mt+1$. Since this is true for all prime factors of $n$, if $q'|\\frac nq$ is another prime factor, then $n\\ge qq'>(2^mt+1)\\cdot 2^m>(2^mt+1)\\cdot h\\ge 2^mh+1=n$, so $n=q$, and we're done.",
"Solution_2": "It's Pocklinton's theorem!",
"Solution_3": "What is the Pocklinton's theorem? Can you explain to me",
"Solution_4": "Pocklinton's theorem as follow:For n be interger and s.t.: [tex]n-1=FR(R 16 \\equal{} 64 \\minus{} x^2 \\minus{} y^2 <\\equal{}> x^2 \\plus{} y^2 \\equal{} (4 \\sqrt{3})^2$. The intersection is then a circunference with radius $ 4 \\sqrt{3}$.[/hide]\n\n2) The projection of the surface found in (a) on the xy plane is the region R in which you are to integrate.\n\n3) With the geometric interpretation of a double integral as the volume of a solid in mind, set up the double integral, with the appropriate limits of integration, that will give you the volume of the solid. Use cilindrical or spherical coordinates if appropriate.\n\n[hide]$ V \\equal{} \\int_{\\minus{} \\pi}^{\\pi} \\int_0^{4 \\sqrt{3}} [(64 \\minus{} \\rho ^2) \\minus{} 16] \\rho \\, d \\rho \\,d \\theta$.[/hide]"
}
{
"Tag": [
"geometry"
],
"Problem": "Our students are learning to write geometry proofs. Some of them, however, find the statements that they're asked to prove to be so obvious that they don't appreciate what proofs are really for.\r\n\r\nI'm intersted in finding some geometry statements that seem to be true, but that are actually false. I'm also looking to find some simple, yet beautiful, proofs of geometric facts that are surprising (or at least, not immediately obvious). The proofs and statements should be simple enough that they could be discovered and/or understood by an \"average\" student with no more than a few subtle hints, and presented by a teacher in say, under 10 minutes.",
"Solution_1": "Don't know where to find counterexamples but this is good: http://www.amazon.co.uk/Crossing-Bridge-Gerry-Leversha/dp/1906001065\r\n\r\nYou probably know this (I haven't seen it but have come across some nice \"Proof without words\" on the internet: http://books.google.co.uk/books?id=Kx2cjyzTIYkC&dq=Proofs+without+words&printsec=frontcover&source=bn&hl=en&ei=ug_cStCeJZWF4QbkmI31Bg&sa=X&oi=book_result&ct=result&resnum=4&ved=0CBIQ6AEwAw#v=onepage&q=&f=false\r\n\r\nSorry I can't be more helpful, but I'll watch for other replies to your post as this seems quite interesting.",
"Solution_2": "Around what level are they?\r\n\r\nOne example I can think of is the proof that Given a point P in an equilateral triangle, the sum of the perpendiculars from that point to the sides is equal to the altitude. If you want, I can write up a solution for it."
}
{
"Tag": [
"geometry",
"Pythagorean Theorem"
],
"Problem": "I found this test on http://www.geocities.com/grissom_math , GEOMETRY 2003 \r\n\r\nAT the end I lookd at the answer key, and I don't know where to find stuff about chords and inscribed things\r\n\r\nI still have problems with number 14,15,16,17,19,21,22, and 24. Can some people help me? I am terribly frustrated.",
"Solution_1": "#10.\r\n\r\nMake the center O. Draw a line from O to C. The arc formed by OC and OB is the same degree as that formed by AB and AC.\r\n\r\nOB=OA=OC=radius=AC\r\n\r\nHence, triangle OAC is equilateral. Thus, A = s(3)^(1/2)\r\n\r\n(1/2B)^2 = (s/2)^2 + s^2 --> B= s(5)^(1/2)\r\n\r\nHence, \r\n\r\nB-A = (5^(1/2)-3^(1/2))s\r\n\r\nAnswer is B.\r\n\r\nI don't have much time. You'll have to wait. I'll post one or two a day.",
"Solution_12": "okay thanks\r\n\r\nI'll learn 1 or 2 a day, it will help a lot",
"Solution_13": "7 more left i think yay =)",
"Solution_14": "started out with 10 i got wrong or something like that\r\n\r\n guessed on 2 of em and got right",
"Solution_15": "Sorry, but the rest are kind of hard to explain here without drawing everything in front of you.",
"Solution_16": "okay, im gonna make a new thread with lots of various problems from tests ive done that i dont get",
"Solution_17": "Some advice to now a ranger, don't make consecutive posts, instead edit your previous posts.",
"Solution_18": "practice=http://www.satmathpro.com/Practice.html\r\nbut only the free stuff :wink: \r\nother than that if you just want to do problems without much expl. try mathcounts.org's link saab!",
"Solution_19": "thanks. I'll look into it :roll:"
}
{
"Tag": [
"geometry",
"3D geometry",
"prism"
],
"Problem": "I was wondering how u would count the number of paths from vertex A to vertex B on an a X b X c rectangular prism if points A and B are endpoints of one of the space diagonals (they are as far away as possible from each other). Is there a specific method?\r\n\r\nThanks",
"Solution_1": "Simply extend the reasoning used to count paths on a 2-dimensional grid. :)",
"Solution_2": "Never mind i got it"
}
{
"Tag": [
"USAMTS",
"calculus",
"Harvard",
"college",
"Princeton",
"AMC 10"
],
"Problem": "Hi, I'm wondering what I should do to better my application for college and whether I will be admitted to any of the Ivies.\r\n\r\nCredentials:\r\n\r\nI haven't took the SAT yet but I will, (approx. scores are 470 for Critical Reading, 500 for Writing, and 770 for Math). However, I did took three Subject Tests and is considering to taking a fourth one soon (770 for Mathematics I, 790 for Mathematics II, 700 for Chemistry, and considering Biology)\r\n\r\nI have participated in sports like Indoor Track (with only three honors) and will participate in Tennis.\r\n\r\nI play two instruments (violin, self taught piano), and I also sing in a selected choir.\r\n\r\nMy overall GPA so far is 4.0 (actual average is about 95.6) and I am ranked #1 in my graduating class.\r\n\r\nI am involved in the musical (and will be involved in the play next year)\r\n\r\nI am doing the USAMTS contest (hopefully I will achieve a bronze status) Current Score is 34/50 (not bad for a first timer)\r\n\r\nI am currently taking AP Chemistry, and AP US History (and considering of taking BC Calculus, AP Biology, AP Microeconomics, AP Statistics, and maybe AP English or AP Psychology)\r\n\r\nI have leadership (currently Treasurer in a selected orchestra and a Director in a voluntary activity.) I am considering of being Treasurer for National Honor Society as well.\r\n\r\nAlso I have at least 50 hours of service from Freshman year to the middle of Junior year.\r\n\r\nLastly, I am involved in Spanish Club and will be taking the AMC 12 this year. (I have taken AMC 10 in the past two years and I'm unsure about my scores.)\r\n\r\nSorry about my lengthy achievements in high school... but I am really considering being admitted in the Ivies, especially Dartmouth and Cornell... please reply with your opinions :)",
"Solution_1": "Sorry to tell you but I doubt it. A 1740 is terrible for top 20 schools (<10 percentile I'm sure) and your EC's don't stand out. Unless you're a recruited athlete or you have some amazing EC you didn't mention, I don't see much of a chance.",
"Solution_2": "if you get better SAT scores (the above ones are really bad), then you have a decent chance of getting into one of the lower-ranked ivies. you pretty much have no shot at harvard or princeton or any similarly competitive school though.",
"Solution_3": "Are you a native speaker? Those Critical Reading and Writing SATs really hurt. As much as I don't like to emphasize test scores, it's very unlikely that you can get away with scores that low without something else spectacular, and nothing stands out.",
"Solution_4": "yeah, my first language is Chinese, and it was very hard for me to understand the English language =/",
"Solution_5": "well, I guess I should begin reading more books, and learn more SAT words before May... hopefully I will get something beneficial from it",
"Solution_6": "Then you should take TOEFL. I think TOEFL is more important and the SAT is considered only if it is above 680 or something.",
"Solution_7": "You are required to take the TOEFL if English is not your native language, and to be on the safer side, it would be good if you can get your SAT's up by at least another 200-250 points.",
"Solution_8": "Yeah, I took a practice test for the SATs and this is now what I got:\r\n\r\nCritical Reading: 540 - 600\r\nMath: 720 - 800\r\nWriting: 580 - 710\r\nTotal: 1840 - 2110\r\n\r\nWith these scores, I hope I can get to one of the Ivies... any more tips on improving my credentials?",
"Solution_9": "I guess a 2110 is decent if thats ure best. Take the ACT also. Plus, you should volunteer more and do more AP's. Im assuming ure a junior currently?",
"Solution_10": "Have you done anything special with your EC's? Otherwise, it just looks like you're a serial joiner and colleges hate that.",
"Solution_11": "[quote=\"iironiic\"]yeah, my first language is Chinese, and it was very hard for me to understand the English language =/[/quote]\r\n\r\nMake sure that you spend time around people with good grammar. Getting the sound of correct language into your ears and grammatical intuition into your heart will help you a lot. Being with people who speak well also tends to lead to more useful social connections."
}
{
"Tag": [
"calculus",
"integration",
"logarithms",
"parameterization",
"real analysis",
"real analysis unsolved"
],
"Problem": "Compute $ \\int_0^\\infty\\frac{e^{\\minus{}ax}\\minus{}e^{\\minus{}bx}}{\\ln x}\\,dx.$\r\n\r\nCould it be computed applyin' double integration? I've tried to find a suitable parameter but I've couldn't.\r\n\r\nI dunno the answer, that's why I haven't mentioned the constraint for $ a,b.$",
"Solution_1": "This diverges unless $ a\\equal{}b,$ and $ a\\equal{}b$ isn't exactly an interesting case.\r\n\r\nThe problem arises not at zero or infinity but from the singularity at $ x\\equal{}1.$"
}
{
"Tag": [
"AMC",
"AIME",
"USA(J)MO",
"USAMO"
],
"Problem": "Does anyone know why they didn't send out an answer booklet for the teacher with the tests this year?",
"Solution_1": "They will be mailed along with the AIME results for your school.",
"Solution_2": "oh thanks. I don't know why my teachers were expecting to have one now then.",
"Solution_3": "For the AMC, a solutions booklet was actually mailed out with the test. So as soon as the test was over, students could basically check their answers against the official ones. Now, I'm not sure about the reasoning behind not attaching a solutions booklet to the AIME, but it does decrease the possibility of cheating. Qualifying for the USAMO is a highly respectable achievement, so I'm assuming there would be at least some temptation if answers were handed out.",
"Solution_4": "[i]That's why [/i]people in other schools knew their scores the day after the test. Anyways i agree completely with Mits...",
"Solution_5": "Or you could just happen to go to a school with a good enough math team that they can figure out all the answers right after the test.",
"Solution_6": "[quote=\"RC-7th\"]Or you could just happen to go to a school with a good enough math team that they can figure out all the answers right after the test.[/quote]\r\n\r\nOr you could come here =).",
"Solution_7": "or you could be so smart that you already know all 15 answers..",
"Solution_8": "[quote=\"RC-7th\"]Or you could just happen to go to a school with a good enough math team that they can figure out all the answers right after the test.[/quote]\r\n\r\nYes that works. You could also befriend enuf USAMO HM, who are many times as good as the answer key is. :-) (Well, at least several of them put together are.)",
"Solution_9": "who writes the official solutions?",
"Solution_10": "THE committee :) (of the AIME) i believe"
}
{
"Tag": [
"inequalities",
"trigonometry",
"geometry",
"incenter",
"geometric transformation",
"reflection",
"circumcircle"
],
"Problem": "In triangle $ABC$, if $L,M,N$ are midpoints of $AB,AC,BC$. And $H$ is orthogonal center of triangle $ABC$, then prove that \\[LH^{2}+MH^{2}+NH^{2}\\leq\\frac14(AB^{2}+AC^{2}+BC^{2})\\]",
"Solution_1": "[size=117]Only if the triangle $ABC$ is acute or right ![/size][quote=\"Omid Hatami\"][color=darkred]In [b][u]acute or right[/u][/b] triangle $ABC$, if $L,M,N$ are midpoints of the its sides and $H$ is its orthocenter, then show that $\\boxed{\\ LH^{2}+MH^{2}+NH^{2}\\leq\\frac14\\cdot \\left(a^{2}+b^{2}+c^{2}\\right)\\ }$[/color][/quote][color=darkblue][b]Proof.[/b] It is well-known that $HA=2R|\\cos A|$, $a=2R\\sin A$ a.s.o. and $\\Delta\\equiv\\sum \\cos 2A=-1-4\\prod\\cos A\\ .$ Therefore,\n$4\\sum HL^{2}-\\sum a^{2}=$ $\\sum\\left(4\\cdot HL^{2}-a^{2}\\right)=$ $\\sum\\left[2\\left(HB^{2}+HC^{2}\\right)-2a^{2}\\right]=$ $4\\sum HA^{2}-2\\sum a^{2}=$\n$16R^{2}\\sum \\cos^{2}A-8R^{2}\\sum \\sin^{2}A=$ $8R^{2}\\sum (1+\\cos 2A)-4R^{2}\\sum (1-\\cos 2A)=$ $8R^{2}(3+\\Delta )-4R^{2}(3-\\Delta )=$\n$4R^{2}(6+2\\Delta-3+\\Delta)=$ $12R^{2}(1+\\Delta )=$ $-48R^{2}\\cos A\\cos B\\cos C\\le 0$ $\\Longrightarrow$ $\\sum HL^{2}\\le\\frac{1}{4}\\cdot\\sum a^{2}\\ .$\n\n[b]Remark.[/b] I posted long ago the following inequality : $\\boxed{\\ \\sum IM^{2}\\ge r(R+r)\\ }\\ ,$ where $I$ is the incenter of the triangle $ABC$. Try solve it ![/color]",
"Solution_2": "A correction : $\\sum IM^{2}\\ge r(R+r)$",
"Solution_3": "I have wondered this for a long time: what does you a.s.o stand for? Is it something along the lines of \"and symmetric others\"?",
"Solution_4": "$\\left\\{\\begin{array}{c}4\\cdot IM^{2}=2\\left(IB^{2}+IC^{2}\\right)-a^{2}\\\\\\\\ IB^{2}=\\frac{ac(p-b)}{p}\\\\\\\\ IC^{2}=\\frac{ab(p-c)}{p}\\end{array}\\right\\|$ $\\Longrightarrow$\r\n\r\n$4p\\cdot IM^{2}=2ac(p-b)+2ab(p-c)-pa^{2}=a[c(a+c-b)+b(a+b-c)-ap]=$ \r\n\r\n$a\\left[(b-c)^{2}+a(b+c)-ap\\right]=a(b-c)^{2}+a^{2}(p-a)$ $\\Longrightarrow$\r\n\r\n$\\left\\{\\begin{array}{c}4p\\cdot\\sum IM^{2}=\\sum \\left[a(b-c)^{2}\\right]+\\sum a^{2}(p-a)\\\\\\\\ \\sum a^{2}(p-a)=4S(R+r)\\end{array}\\right\\|$ $\\Longrightarrow$\r\n\r\n$4p\\cdot \\sum IM^{2}=\\sum \\left[a(b-c)^{2}\\right]+4pr(R+r)$ $\\Longrightarrow$ $\\boxed{\\ \\sum IM^{2}\\ge r(R+r)\\ }\\ .$",
"Solution_5": "Forall $P$ on plane then $PA'^{2}+PB'^{2}+PC'^{2}\\ge\\frac{a^{2}+b^{2}+c^{2}}{12}\\ge\\frac{r(4R+r)}{3}\\ge r(R+r)$ with $A',B',C'$ are midpoint of $BC,CA,AB$ resp\r\nThus the result of Virgil Nicula is true for all $P$ on plane! And more here we can show min of $\\sum PA'^{2}$ with a fix triangle $ABC$ and really it is only simple result!",
"Solution_6": "[quote=\"Omid Hatami\"]In triangle $ABC$, if $L,M,N$ are midpoints of $AB,AC,BC$. And $H$ is orthogonal center of triangle $ABC$, then prove that \\[LH^{2}+MH^{2}+NH^{2}\\leq\\frac14(AB^{2}+AC^{2}+BC^{2})\\][/quote]\n\n\ni tried in complex co-ordinat and the result became\n$2abc+ \\sum (a^2b+b^2a) \\ge 0$ is this true?\n\n$a,b,c$ are co-ordinat of $A,B,C$ and $h=a+b+c$ $l=(a+b)/2$ $m=(a+c)/2$ $n=(b+c)/2$",
"Solution_7": "[quote=\"tsemng\"][quote=\"Omid Hatami\"]In triangle $ABC$, if $L,M,N$ are midpoints of $AB,AC,BC$. And $H$ is orthogonal center of triangle $ABC$, then prove that \\[LH^{2}+MH^{2}+NH^{2}\\leq\\frac14(AB^{2}+AC^{2}+BC^{2})\\][/quote]\n\n\ni tried in complex co-ordinat and the result became\n$2abc+ \\sum (a^2b+b^2a) \\ge 0$ is this true?\n\n$a,b,c$ are co-ordinat of $A,B,C$ and $h=a+b+c$ $l=(a+b)/2$ $m=(a+c)/2$ $n=(b+c)/2$[/quote]\nHmm, I am getting a different result.\nDenote by $H$, the origin; and give $A,B,C$ the affixes $u,v,w.$ So we see that $l=\\frac{v+w}{2}, m=\\frac{w+u}{2},n=\\frac{u+v}{2}.$\nTherefore $AB^2+BC^2+CA^2=(u-v)^2+(v-w)^2+(w-u)^2$ and\n$LP^2+MP^2+NP^2=\\frac 14\\left[(w+u)^2+(u+v)^2+(v+w)^2\\right].$\nAlso, $u+v+w=0$; so that \n\\[\\begin{aligned}\\frac 14\\left[AB^2+BC^2+CA^2\\right]&=\\frac 14\\left[(u-v)^2+(v-w)^2+(w-u)^2\\right]\\\\&=\\frac 14\\left[2(u+v+w)^2+(u-v)^2+(v-w)^2+(w-u)^2\\right]\\\\&=\\frac 14\\left[2(u^2+v^2+w^2)+(u+v)^2+(v+w)^2+(w+u)^2\\right]\\\\&\\geq \\frac 14\\left[(u+v)^2+(v+w)^2+(w+u)^2\\right];\\end{aligned}\\]\nAnd we are done. $\\Box$\n:)",
"Solution_8": "And so what is the equality case ? Is it when the triangle is reduced to a point ?\nThanks in advance.",
"Solution_9": "Only if $ABC$ is acute or right. Equality holds if it is right.\nWe use the fact that reflection of the orthogonal center with the respect to the midpoints of the triangle lay on the circumcircle. \nDenote $P,Q$ intersection of $HL$ with circumcircle. ($P$ is reflection of $H$ wrt $L$)\nLooking at power of the point $L$ wrt to the circumcircle and noticing that $LP=HL \\leq LQ$ we get\n$\\frac{AB^2}{4}=AL*BL=LQ*LP \\geq HL^2$\nApplying this to two other sides we get the desired result.",
"Solution_10": "From Leybnits's identity it's equivalent to $ cosA \\cdot cosB \\cdot cosC \\ge 0 $.It's true if and only if $ ABC $ is non-obtuse!",
"Solution_11": "[quote=Virgil Nicula][size=117]Only if the triangle $ABC$ is acute or right ![/size][quote=\"Omid Hatami\"][color=darkred]In [b][u]acute or right[/u][/b] triangle $ABC$, if $L,M,N$ are midpoints of the its sides and $H$ is its orthocenter, then show that $\\boxed{\\ LH^{2}+MH^{2}+NH^{2}\\leq\\frac14\\cdot \\left(a^{2}+b^{2}+c^{2}\\right)\\ }$[/color][/quote][color=darkblue][b]Proof.[/b] It is well-known that $HA=2R|\\cos A|$, $a=2R\\sin A$ a.s.o. and $\\Delta\\equiv\\sum \\cos 2A=-1-4\\prod\\cos A\\ .$ Therefore,\n$4\\sum HL^{2}-\\sum a^{2}=$ $\\sum\\left(4\\cdot HL^{2}-a^{2}\\right)=$ $\\sum\\left[2\\left(HB^{2}+HC^{2}\\right)-2a^{2}\\right]=$ $4\\sum HA^{2}-2\\sum a^{2}=$\n$16R^{2}\\sum \\cos^{2}A-8R^{2}\\sum \\sin^{2}A=$ $8R^{2}\\sum (1+\\cos 2A)-4R^{2}\\sum (1-\\cos 2A)=$ $8R^{2}(3+\\Delta )-4R^{2}(3-\\Delta )=$\n$4R^{2}(6+2\\Delta-3+\\Delta)=$ $12R^{2}(1+\\Delta )=$ $-48R^{2}\\cos A\\cos B\\cos C\\le 0$ $\\Longrightarrow$ $\\sum HL^{2}\\le\\frac{1}{4}\\cdot\\sum a^{2}\\ .$\n\n[b]Remark.[/b] I posted long ago the following inequality : $\\boxed{\\ \\sum IM^{2}\\ge r(R+r)\\ }\\ ,$ where $I$ is the incenter of the triangle $ABC$. Try solve it ![/color][/quote]\n\nSorry,I don,t know why LH^2-a^2=2(BH^2+CH^2)-2a^2\nCan you explain?",
"Solution_12": "$ O $ the center of $ \\odot (ABC) $ and WLOG let $\\odot (ABC)$ be unit circle vectors\nBy Euler line we get $ H=a+b+c $ \n\n$ LH^{2}+MH^{2}+NH^{2}\\leq\\frac14(AB^{2}+AC^{2}+BC^{2}) \\Longleftrightarrow \\sum_{cyc} (\\frac{a+b+2c}{2})^2 \\le \\sum_{cyc} (\\frac{a-b}{2})^2 $\n\nWe use that $ a^2=b^2=c^2=1 $\n\n$ \\Leftrightarrow -1 \\ge ab+bc+ca $\n\n By scalar product ,$ ab= cos(2C) $,\n\n$ \\Leftrightarrow -1 \\ge cos(2A)+cos(2B)+cos(2C) $\n\n$ \\Leftrightarrow -1 \\ge 2cos(A+B)cos(A-B)+2cos^2(A+B)-1 $\n\n$ \\Leftrightarrow 0 \\ge +4cos(A+B)cosAcosB $\n\n$ \\Leftrightarrow 0 \\ge -4cos(C)cos(A)cos(B) $ \n\nThis is true for acute or right triangle."
}
{
"Tag": [
"email",
"FTW"
],
"Problem": "anyone already planning an april 2008 one?",
"Solution_1": "Me; I am working on a website that will basically have some similar features as #H34N1's site did.\r\nWhen the time comes, I will let everyone know how and where to signup.\r\nAlso, if anyone would like to run it this time, just send me a PM. :)\r\nminicon",
"Solution_2": "You may now register your team [url=http://aopsftw.freehostia.com/home/]here[/url].\r\nBut first, you must create an account there. Mail server does not work, so if it is a long password, remember it or write it down. :D\r\nThis is an individual tournament.\r\nAvatars must be private messaged to me. Attachments currently do not work on that site.\r\nThank you for your cooperation,\r\nminicon",
"Solution_3": "So...\r\nMail server not working + required email validation = no account created :D.",
"Solution_4": "im not getting the emails",
"Solution_5": "just a curiousity, what's the point of the team name and avatar if it's individuals?",
"Solution_6": "Incase you want to be called something different + user may not have avatar. Its easier to see it with avatars.\r\n[b]Everyone who has registered, please re-register. I fixed the email settings[/b]",
"Solution_7": "You may now register your team via a form located at\r\nhttp://aopsftw.freehostia.com/home/?q=node/3\r\nThank you.\r\nminicon\r\n:)",
"Solution_8": "For your information, [b]the site aopsftw.freehostia.com/home/ will be down tonight for maintenance[/b]. If you wish to register, please post here temporarily. Hopefully the site will be back up and running tomorrow. I will post when that happens.\r\nThank you for your cooperation.",
"Solution_9": "Site is now running on 5.7 .\r\nEverything should be same; no files were deleted.\r\n:)",
"Solution_10": "I accidentally signed up without for it without an avatar. :| \r\nWhat should i do?",
"Solution_11": "Just sign up again with all info you can fill out.\r\nI won't add you twice. \r\n:)",
"Solution_12": "how many people are in this tournament",
"Solution_13": "I think it will be max. 16 people, unless there is a giant rush before then, in which case it will be max. 32.",
"Solution_14": "I am planning to have 16 people and then have a qualification round where 1/2 of the people will be eliminated.\r\n:)",
"Solution_15": "I will probably not make it to over half of the scheduled tournaments, so could I file as a substitute?",
"Solution_16": "Yeah, if someone can't make it to the whole tournament (or part of it), you can fill in if you can make it; if no-one can make it, the opponent gets an automatic bye.\r\n :P",
"Solution_17": "Are there only two teams or many more?",
"Solution_18": "The April Tournament is an individual tournament. MustItAlwaysBeMe's tournament is a team tournament. But all spots are taken for those. If you want, try sending a private message to MustItAlwaysBeMe and he might try to add you in; all of his slots aren't full, it has just been past the deadline but no games have been played.\r\n:)",
"Solution_19": "Thanks. :) I think I want to join the individual tournament, sounds more challenging. (As a substitute, of course) Where is the current player list?",
"Solution_20": "Currently [url=http://aopsftw.freehostia.com/home/?q=node/10]here[/url].\r\nIt will probably move to another place as I am planning on changing the hosts tonight. A link will be provided to the new site when up though. :lol:\r\nEdit: Don't fill out any forms, I can't access them as logging in doesn't work. :(",
"Solution_21": "Darn it, I keep on missing those new tournaments... Team April tourney, anybody?",
"Solution_22": "For the tournament, if you don't play the three games who gets the bye? because I'm playing rd5493 and he's not responding to my private messages. :|",
"Solution_23": "Yeah, AIMEman and rd5493 have not gotten my messages. I'll take care of this, but first let's wait a couple of days. If he doesn't respond by Wednesday night, he will be replaced.",
"Solution_24": "When's the May tournament? Because its May already. And is it a team or individual tournament?",
"Solution_25": "I am taking May off. :( But I will probably run another in June. I don't exactly know what I'm doing yet though.",
"Solution_26": "I've been busy, and there is going to be no May tournament.\r\n\r\nTechnically, I should be doing it and I should have wrapped up the April Tournament, so sorry guys. As for the next one, I have no clue about it.",
"Solution_27": "Wait, does that mean minicon's running two tournaments in a row? (April and June)",
"Solution_28": "I don't know. I might do the June tournament. No matter what, the next one will definitely go smoother. A lot smoother, in fact.",
"Solution_29": "Um...It's June, and there's no tournament......Could one of you please run it?"
}
{
"Tag": [],
"Problem": "The standard tip in a restaurant is $ 15 \\%$ of the bill before tax. Many people in California find the tip by doubling the sales tax, which is $ 8.25 \\%$. By how many cents does this exceed the standard tip on a bill of $ \\$60$ before tax?",
"Solution_1": "Tax is $ .0825\\times 60\\equal{}4.95$ so tip is $ 9.90$. Alternatively, $ .15\\times 60\\equal{}9$, so the difference is 90 cents.",
"Solution_2": "Faster way is:\r\n$ .0825*2\\equal{}.165\\equal{}$ the actual % tipped.\r\n$ .165\\equal{}.15\\equal{}.015\\equal{}$ the extra % tipped\r\n$ .015*\\$60\\equal{}\\$.9\\equal{}90$ cents = the extra amount tipped."
}
{
"Tag": [
"induction",
"strong induction",
"combinatorics proposed",
"combinatorics"
],
"Problem": "Consider a convex polygon having $n$ vertices, $n\\geq 4$. We arbitrarily decompose the polygon into triangles having all the vertices among the vertices of the polygon, such that no two of the triangles have interior points in common. We paint in black the triangles that have two sides that are also sides of the polygon, in red if only one side of the triangle is also a side of the polygon and in white those triangles that have no sides that are sides of the polygon.\r\n\r\nProve that there are two more black triangles that white ones.",
"Solution_1": "Here is my solution :\n\nWe prove by induction that there are at least two black triangles, and that $b = w+2$ where $b$ (resp. $w$ ) denote the number of black (resp. white) triangles.\n\nFirst for $n=4$ there are two black triangles and no red and white. Thus the result holds.\nNow, let $n > 4$ be given, and assume the result holds for any $k$ such that $4 \\leq k < n$.\n\nLet's consider a $n$-gon as colored as in the statement of the problem.\nLet $d$ be any diagonal (not side) which has been drawn for the triangulation.\nThen $d$ separates the $n$-gon into two disjoint polygons. If any of these polygons is a triangle then it is a black one in the original coloring. If none of these polygons is a triangle then use the induction hypothesis for each of the two subpolygons. They gives us at least 2 black triangles on each side of $d$. And, for each side at most one of these black triangles have $d$ as a side. Thus, in any case, there is at least one black triangle on each side of $d$, which ensures that there are a least two black triangles for the $n$-gon.\n\nNow, let $T =ABC$ be a black triangle (which does exist from above), with vertices in that order on the $n$-gon. Consider the $n-1$-gon obtained by deleting $T$ and replacing it by the edge $AC$.\nThe coloring induced on this $(n-1)$-gon is the same as the initial one except for the triangle $T'$ having $AC$ as side. Note that since $AC$ is a diagonal of the $n$-gon, then $T'$ was red or white in the original coloring, and since it is a side of the $(n-1)$-gon it is now black or red, respectively.\nFrom the induction hypothesis, we have $b'=w'+2$ for the $(n-1)$-gon.\n- If $T'$ is black for the induced coloring then it was red for the initial one. Adding $T$, it leads to $b=b'-1+1=b'$ and $w=w'$.\n- If $T'$ is red for the induced coloring then it is white for the initial coloring. Adding $T$ leads to $b = b'+1$ and $w=w'+1$.\n\nThus in any case, we have $b=w+2$ as desired, which ends the induction step and the proof.\n\nPierre.",
"Solution_2": "Shortlist's solution of problem 4\r\n\r\nSolution. Denote by b, r, w the number of black, red white triangles respectively. It is easy to prove that the polygon is divided into n-2 triangles, hence\r\nb + r + w = n \u2013 2.\r\neach side of the polygon is a side of exactly one triangle of the decomposition, and thus\r\n2b + r = n.\r\nSubtracting the two relations yields w = b \u2013 2, as needed.",
"Solution_3": "This is a very interesting solution for pb 4, because it does not need induction to prove the result. :) \r\n\r\nPierre.",
"Solution_4": "[quote=\"pbornsztein\"]This is a very interesting solution for pb 4, because it does not need induction to prove the result.[/quote]\r\n\r\nI don't think it is really induction-free. Actually, I don't know any proof of the \"easy to prove\" fact that the number of triangles used in the triangulation is n - 2 but the one using induction.\r\n\r\n Darij",
"Solution_5": "Yet, this is easy to do :\r\n\r\nThe sum of the internal angles of the convex $n$-gon is $(n-2) \\pi$, which is also the sum of the angles of the triangles of the subdivision according to the statement of the problem, hence there are $n-2$ triangles :) \r\n\r\nMoreover, the fact that the sum of the internal angles of the convex $n$-gon is $(n-2) \\pi$ is easy to prove without induction again : Just divides the convex $n$-gon into $n-2$ triangles by all the diagonals from one arbitrary vertex, and use again that the sum of the internal angles is the sum of the angles of all these triangles. \r\n\r\nIt is nice that one particular triangulation (the last above) gives the key for any triangulation.\r\n\r\nPierre.",
"Solution_6": "And here is my solution:\n\nAt first we will show a rather simple lemma:\n\n[b]Lemma.[/b] Let $n\\geq 4$ be an integer. Every triangulation of a convex $n$-gon contains at least $2$ black triangles.\n\n[i]Proof.[/i] The proof will use induction after $n$:\n\n[i]Induction base:[/i] The case $n = 4$ is not what most of you would call difficult...\n\n[i]Induction step:[/i] Now assume that we have proved our lemma for all $n\\leq k$, where $k$ is a certain number with $k\\geq 4$, and now we are going to prove it for $n = k + 1$.\n\nIn fact, consider a triangulated convex $\\left(k + 1\\right)$-gon $P$. Assume that it doesn't contain $2$ black triangles. Every side of $P$ belongs either to a black triangle or to a red triangle. Since $k\\geq 4$, we have $k+1\\geq 5$, and our $\\left(k + 1\\right)$-gon $P$ has, at least, $5$ sides. Only $2$ of these sides can belong to black triangles (else, we would have at least $2$ black triangles, contradicting our assumption!). Hence, we will always find, at least, one side of $P$ which belongs to a red triangle $T$. If we remove this red triangle, we are left with two triangulated convex polygons $Q$ and $Q^{\\prime}$ having one common vertex (namely the third vertex of the removed red triangle). Each of these two polygons $Q$ and $Q^{\\prime}$ has $\\leq k$ vertices. Now, by our induction assumption, each of these two polygons must have at least two black triangles. Hence we have four black triangles altogether. Now, if we put in the removed red triangle $T$ again, then at most two of these black triangles may possibly become red triangles (in fact, every of our two polygons $Q$ and $Q^{\\prime}$ had a side bordering with the removed red triangle $T$; now, if such a side belongs to a black triangle in $Q$ or in $Q^{\\prime}$, then, after putting in the red triangle $T$ again, this black triangle becomes a red triangle). But at least $4 - 2 = 2$ black triangles remain. Hence, we have found $2$ black triangles in our $\\left(k + 1\\right)$-gon $P$, contradicting our assumption. Hence, our $\\left(k + 1\\right)$-gon $P$ must have $2$ black triangles (at least). This completes the induction step, and the Lemma is proven.\n\nNow to the solution of the problem:\n\nBy the [i]niveau[/i] of a triangle in our triangulation, we will mean the number defined according to the following rule: If the triangle is white, then its niveau is -1; if the triangle is red, then its niveau is 0; if the triangle is black, then its niveau is 1.\n\nThe [i]summary niveau[/i] of a triangulated polygon will denote the sum of the niveaus of all triangles involved in the triangulation. Equivalently, it is the number of the black triangles minus the number of the white triangles.\n\nNow we will prove that the summary niveau is always equal to $2$ (when $n\\geq 4$). This will be obviously enough to solve the problem.\n\nI will prove my assertion by induction. For $n = 4$, it is more than trivial. Now we assume that it holds for a certain $n\\geq 4$ (i. e. it holds for all $n$-gons), and we have to show that it holds for $n + 1$ (i. e. for all $\\left(n + 1\\right)$-gons).\n\nConsider a triangulated $\\left(n + 1\\right)$-gon $P$. Clearly, $n\\geq 4$ yields $n+1\\geq 5$. On the other hand, after the Lemma, the triangulated $\\left(n + 1\\right)$-gon $P$ contains at least two black triangles. Let $D$ be one of them. Now remove this black triangle $D$. What you then get is a triangulated $n$-gon $P^{\\prime}$. All black, white, red triangles in the triangulation of $P$ preserve their color in $P^{\\prime}$, except for the initial black triangle $D$, which disappears, and the triangle $D^{\\prime}$ directly bordering with $D$. In fact, if $D^{\\prime}$ was a red triangle for $P$, then it becomes a black triangle for $P^{\\prime}$, and if $D^{\\prime}$ was a white triangle for $P$, then it becomes a red triangle for $P^{\\prime}$. (You can easily figure out that $D^{\\prime}$ could not have been a black triangle for $P$; since otherwise, all three sides of $D^{\\prime}$ would be sides of $P^{\\prime}$, and thus, the polygon $P^{\\prime}$ would be a triangle, and $P$ would be a quadrilateral, contradicting $n+1\\geq 5$.)\n\nNow, we see that the triangulated polygons $P$ and $P^{\\prime}$ have the same summary niveau. This is because when we remove the black triangle $D$, then on one hand, we lose a niveau of $1$ (since $D$ had a niveau of $1$), but on the other hand, we gain a niveau of $1$ again (in fact, if $D^{\\prime}$ was a red triangle for $P$, then it becomes a black triangle for $P^{\\prime}$, so that the niveau changes from $0$ to $1$, and if $D^{\\prime}$ was a white triangle for $P$, then it becomes a red triangle for $P^{\\prime}$, so that the niveau changes from $-1$ to $0$). Hence, the sum of the niveaus of all triangles remains constant, i. e., the summary niveaus of the triangulated polygons $P$ and $P^{\\prime}$ are equal. But since $P^{\\prime}$ has $n$ edges, then by our induction assumption, its summary niveau is $2$, and hence the summary niveau of $P$ is also $2$. Proof complete.\n\n Darij",
"Solution_7": "We use strong induction on $n$.\nConsider a white triangle in the middle of the polygon. It then divides the problem into three separate subpolygons.\n\nLemma: In each of the subpolygons, $B = W + 1$.\nThis is easy to see, because red triangles are counted as black and white polygons are counted as red. Then we apply the induction hypothesis.\n\nTherefore, summing over the subpolygons, $B = W + 3$. But we must include the white triangle in the center, so $B = W +2$, as desired.",
"Solution_8": "First we shall state a $lemma$:\n\n$Lemma$: A convex polygon with $n$ sides can be arbitrarily decomposed into exactly $n-2$ triangles having all the vertices among the vertices of the polygon, such that no two of the triangles have interior points in common.\n\n$Proof of Lemma$: \nSum of internal angles of a convex polygon with $n$ sides is $(n-2) \\pi$. \nSince triangles are using all of the vertices of the polygon, the aggregate sum of angles of all the\ntriangles must equal the sum of internal angles of a convex polygon = $(n-2) \\pi$.\nThus, number of triangles = $(n-2) \\pi / \\pi = n-2$.\n\n\n$Proof of problem$:\nLet there be a total of $x Black$ triangles in $n$ sided convex polygon $P1$.\n\nNow consider the polygon $P2$ formed after removing all the $x$ black triangles from polygon $P1$. It's not hard to see that $P2$ has exactly $(n - x)$ sides.\n\nNow, in polygon $P2$, we are left with exactly $(n - x) - x = n - 2x$ sides which are still part of the original Polygon $P1$.\n\nNo two adjacent sides of these $(n - 2x)$ can be joined to form another $Black$ triangle (since we have assumed a total of $x Black$ triangles exist).\n\nSo that leaves us with $(n - 2x)$ $Red$ triangles in total.\n\nSo, now all the triangles that remain must be $White$ triangles.\n\nFrom lemma, $P2$ has exactly $(n - x - 2)$ triangles satisfying the conditions of the problem, so the number of $White$ triangles = $(n - x - 2) - (n - 2x) = x - 2$\n\nThus total number of $White$ triangles = $x - 2$ which is $2$ less than total number of $Black$ triangles.",
"Solution_9": "We induct on $n$, with trivial base cases $n=4$ and $n=5$. We also note that 0 white triangles imply 2 black triangles and the rest are red. Hence we assume the existence of at least 1 white triangle, which we draw in. The remainder of the polygon is split into 3 smaller polygons, from which we claim the following:\n\n[color=#00f][b]Claim:[/b][/color] In each of these polygons, we have $B-W=1$.\n[list]\n[*] Obvious if the polygon is a triangle.\n[*] Otherwise, we utilize our inductive hypothesis, which says $B-W=2$ with respect to the sides of our smaller polygon. However, we must consider the triangle with one side on our main white triangle.\n\nIf the triangle is black in our smaller polygon, it is red in our original polygon, and if it is red in our smaller polygon, it is white in our original polygon. Thus the difference $B-W$ decreases by 1. ${\\color{blue} \\Box}$\n[/list]\n\nThus we conclude by summing over the 3 new polygons and adding back our white triangle. $\\blacksquare$",
"Solution_10": "[hide=\"Storage:\"] \n\nWe'll prove by induction two statements at the same time : First, there exists a black triangle. Second, there are two more black triangles than white ones. For $n=4$ both are trivial, for we are forced to connect two diagonally opposite vertices of the quadrilateral.\n\nNow take $n\\ge5$ and consider a convex $n$-gon. Supposing that the first statements are true for a $k$-gon st $4 \\leq k <\n n$, then we can take any arbitrary triangle in the triangulation of the $n$-gon. IF it's black, we are done. Otherwise, it divides the polygon in two. Take any half, use $1$st inductive statement on it. If the new black triangle is also black in the $n$-gon, we are done. Otherwise, it contains a side of the original triangle. Repeat this operation until we clearly eventually get to a black triangle. This proves the first statement.\n\nFor the second statement, take a black triangle in the $n$-gon.Consider the $(n-1)$-gon formed by removing the two sides of the original $n$-gon that were lying on said black triangle, with the vertex connecting them with it. By induction, there are $2$ more \\emph{new} black triangle than \\emph{new} white ones. Consider the triangle containing the third segment of the original black triangle. If this triangle is \\emph{new} black, then its \\emph{old} red. If it's \\emph{new} red, then its \\emph{old} white. In any case, taking in account the first black triangle, we notice that the difference stays constant.\n\nSo by induction, there are two more black triangles than white ones.\n\n$$\\mathbb{Q.E.D.}$$\n[/hide]"
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "[b]Pro1[/b] Let $ x,y,z\\in N^*$. Solve equation\r\n$ 12^x\\plus{}4^y\\equal{}2008^z$\r\n\r\n[b]Pro2[/b] Find $ x,y\\in N$ such that\r\n$ (x^2\\plus{}y)(y^2\\plus{}x)\\equal{}(x\\minus{}y)^3$\r\n\r\n[b]Pro3[/b] $ p$ is odd prime number. $ x,y\\in N$\r\nSolve equation\r\n$ x^p\\plus{}y^p\\equal{}p.((p\\minus{}1)!)^p$",
"Solution_1": "It might be more convenient if these problems were posted as three separate topics."
}
{
"Tag": [
"AMC",
"USA(J)MO",
"USAMO",
"induction",
"number theory"
],
"Problem": "Prove that for every positive integer $n$ there exists an $n$-digit number divisible by $5^n$ all of whose digits are odd.",
"Solution_1": "When replying to the problem, I ask that you make posts for solutions and submit comments, jokes, smilies, etc. separately. Please use LaTeX for posting solutions. Thanks.",
"Solution_2": "Don't post in solved problems, please. This (and all the others) should go in the proposed problems section.\r\n\r\nOoh and this was the very first problem I solved on an official olympiad ... yay :).\r\n\r\n[hide]It follows by induction on $n$; clearly for $n = 1$, $5$ satisfies the given conditions. Suppose the proposition works for $n$, i.e. there exists some number $N$ satisfying the giving conditions. Consider the five numbers $i*10^n + N$, where $i = 1, 3, 5, 7, 9$. It is clear that these five numbers all have $n+1$ digits with all digits odd, from the inductive hypothesis. Also from the inductive hypothesis, they are all divisible by $5^n$. If one were to divide them by $5^n$ however, they would differ by $2^n$ between consecutive numbers, i.e. they would all have different residues mod $5$. Thus one exists whose residue mod $5$ is $0$, or in other words, the original number was divisible by $5^{n+1}$ and thus satisfies the conditions for $n+1$.[/hide]",
"Solution_3": "It follows by induction on $n$; clearly for $n = 1$, $5$ satisfies the given conditions. Suppose the proposition works for $n$, i.e. there exists some number $N$ satisfying the giving conditions. Consider the five numbers $i*10^n + N$, where $i = 1, 3, 5, 7, 9$. It is clear that these five numbers all have $n+1$ digits with all digits odd, from the inductive hypothesis. Also from the inductive hypothesis, they are all divisible by $5^n$. If one were to divide them by $5^n$ however, they would differ by $2^n$ between consecutive numbers, i.e. they would all have different residues mod $5$. Thus one exists whose residue mod $5$ is $0$, or in other words, the original number was divisible by $5^{n+1}$ and thus satisfies the conditions for $n+1$.",
"Solution_4": "Who can show that the first digit number of the k-digit number can be never equal to 1 ?\r\n\r\nDavron",
"Solution_5": "It is easy to construct this number by induktion.\r\n$k_0=1, N_0=k_0*5^n=a_{n-1,0}a_{n-2,0}\\dots a_{0,0}, a_{0,0}=5 \\ is \\ odd.$ \r\nLet we have $k_l<2^{m_l+1},k_l*5^n=N_l=a_{n-1,l}a_{n-2,l}\\dots a_{0,l}$ were m_l+1 digits $a_{m_l,l}\\dots a_{0,l}$ are odd.\r\nLet we have $m_{l+1}>m_l$, suth that $a_{m_{l+1},l}$ is even. We construct \r\n$k_{l+1}=k_l+2^{m_{l+1},N_{l+1}=k_{l+1}*5^n}$ with property: digits $a_{i,l+1}$ are odd for $i\\le m_{l+1}$.\r\nThis number is unique.",
"Solution_6": "[quote=\"Rust\"]It is easy to construct this number by induktion.\n$k_0=1, N_0=k_0*5^n=a_{n-1,0}a{n-2,0}\\dots a_{0,0}, a_{0,0}=5 \\ is \\ odd.$ \nLet we have $k_l<2^{m_l+1},k_l*5^n=N_l=a_{n-1,l}a_{n-2,l}\\dots a_{0,l}$ were m_l+1 digits $a_{m_l,l}\\dots a_{0,l}$ are odd.\nLet we have $m_{l+1}>m_l$, suth that $a_{m_{l+1},l}$ is even. We construct \n$k_{l+1}=k_l+2^{m_{l+1},N_{l+1}=k_{l+1}*5^n}$ with property: digits $a_{i,l+1}$ are odd for $i\\le m_{l+1}$.\nThis number is unique.[/quote]\r\n\r\nThank you very much Rust please and is this the answer to my question ?\r\n\r\nDavron",
"Solution_7": "I think you are wrong. Check it by programming. It is easy calculate for n<1000.",
"Solution_8": "[quote=\"Rust\"]I think you are wrong. Check it by programming. It is easy calculate for n<1000.[/quote]\r\n\r\nnope but we have all the terms as odd numbers...?\r\n :)",
"Solution_9": "the first number to satisfy ur condition is 375=3*5^3",
"Solution_10": "We will proceed by induction to prove that there always exists a $n$ digit number with all odd digits $A_n$ such that $5^n|A_n$.\n\nBase Case: $n=1$, let $A_1=5$ and $5^1|A_1=5$.\n\nInductive Hypothesis: There exists a $k$ digit number with all odd digits such that $5^k|A_k$.\n\nNow let $a\\in\\{1, 3, 5, 7, 9\\}$ and $A_k=b_k\\cdot 5^k$, then let $A_{k+1}=a\\times 10^{k}+bA_k$ and we shall show that you can always pick some $a$ and $b$ such that $5^{k+1}|A_{k+1}$. Notice that $A_{k+1}$ is $A_k$ with an odd digit augmented to beginning of the number. Thus, $A_{k+1}$ contains only odd digits.\n\\[A_{k+1}=5^{k}(a\\cdot 2^k+b_k)\\]\nSo if we can choose some $a$ such that $a\\cdot 2^k+b_k$ is divisible by $5$, then $5^{k+1}|A_{k+1}$. However, this is obviously true because $2^k a$ can take on any possible residue modulo $5$ since $a$ can take on any residue modulo $5$ and we are done. $\\blacksquare$",
"Solution_11": "Induction: Clearly true for $n=1$. Now assume true for $n=k$, where $d_i$ are the digits in the decimal representation, i.e. $5^k\\mid{\\overline{d_1d_2\\cdots{d_k}}_{10}}$ with all $d_i$ odd. Then the next number in the list can potentially be any one of:\n$\\overline{(2a+1)d_1d_2\\cdots{d_k}}_{10}$, where $0\\leq{a}\\leq{4}$; we want to show that one of these has to be divisible by $5^{k+1}$. \nBy the inductive hypothesis: $(2a+1)\\cdot{10^k}+5^kt$; $t\\in{\\mathbb{Z^+}}$.\nFactoring (I-V) we have: $5^k(2^k(2a+1)+t_1)=5^k(2^{k+1}a+2^k+t_1)$. \nBut wait! $(2^{k+1}a+2^k+t_1)$ forms a complete set of residue classes $\\mod{5}$ as it runs across the $a$'s (this is obvious, see myticterminator's proof of it above), so we're done. $\\blacksquare$",
"Solution_12": "Quick\n[hide=Solution]\nWe use induction on $n$. For $n=1$ pick $5$, completing the base case. Suppose the result is true for $n=k$, we will show it is for $k+1$.\n\nLet $\\overline{a_ka_{k-1}\\cdots a_1}$ be an $k$ digit number such that all $a_i, 1 \\le i \\le k$ and $5^k|\\overline{a_ka_{k-1}\\cdots a_1}$. Suppose $\\frac{\\overline{a_ka_{k-1}\\cdots a_1}}{5^k}=x$. \nConsider the set of numbers $x_1=\\overline{1a_ka_{k-1}\\cdots a_1}, x_3=\\overline{3a_ka_{k-1}\\cdots a_1}, x_5=\\overline{5a_ka_{k-1}\\cdots a_1}, x_7=\\overline{7a_ka_{k-1}\\cdots a_1}, x_9=\\overline{9a_ka_{k-1}\\cdots a_1}$. Trivially, $5^n|10\\cdots 0$ where there are $n$ zeroes. \nNote that when $1 \\le i \\le 5, x_i=5^k(i+x)$. $x \\pmod{5}$ can be $0, 1, 2, 3, 4$ but $i$ ranges through the whole set of residues$\\pmod{5}$, so we can choose an appropriate $i$ such that $5^{k+1}|x_i$. This completes the inductive step so done.\n[/hide]",
"Solution_13": "We prove the set of $n$-digits integers whose digits are odd form a complete set of residue modulo $5^n$. There are $5^n$ such integers. It suffices to show they are distinct.\n\n[hide=Proof]\nLet $A,B$ be 2 distinct $n$-digits integers whose digits are odd. For sake of contradiction, suppose $A\\equiv B\\pmod{5^n}$. Let $0\\le k\\le n-1$ be the smallest integer s.t. $A$ and $B$ differs on the $k^{\\text{th}}$ digit (from the right). I.e.\n$$A=\\overline{a_{n-1}\\ldots a_ka_{k-1}\\ldots a_0}\\;\\wedge\\;B=\\overline{b_{n-1}\\ldots b_ka_{k-1}\\ldots a_0}$$\nHence $5^n\\mid A-B\\iff 5^{n-k}\\mid\\overline{a_{n-1}\\ldots a_k}-\\overline{b_{n-1}\\ldots b_k}=C\\implies 5\\mid C$.\nSo the last digit $c_0$ of $C$ is either $0$ or $5$. Note $c_0\\equiv a_k-b_k\\pmod{10}$ and both $a_k$ and $b_k$ are odd, so $c_0$ is even ie $c_0=0$. Finally, $a_k=b_k$. Contradiction.\n[/hide]\n\nPS : I suppose this problem inspired [url=https://www.artofproblemsolving.com/community/c6h148016p836869]Croatian National MO 2005[/url].",
"Solution_14": "[hide=Solution]We shall induct on $n$.\n\nBase Case: $n=1$.\n\nTake the $1$-digit number $5$.\n\nInductive Step: Suppose the $k$ digit number $a_{k-1}a_{k-2}\\cdots a_1a_0$ is divisible by $5^k$ with all odd digits.\n\nClaim: One of $1a_{k-1}a_{k-2}\\cdots a_1a_0$, $3a_{k-1}a_{k-2}\\cdots a_1a_0$, $5a_{k-1}a_{k-2}\\cdots a_1a_0$, $7a_{k-1}a_{k-2}\\cdots a_1a_0$ or $9a_{k-1}a_{k-2}\\cdots a_1a_0$ will be divisible by $5^{k+1}$. It is easy to see that all of these are still divisible by $5^k$ since $10^k\\equiv 0\\pmod{5^k}$. Say the odd number for the first digit we pick is $2m+1$, then let $a_{k-1}a_{k-2}\\cdots a_1a_0=5^k\\cdot n$. We have $(2m+1)\\cdot 10^k+5^k\\cdot n=5^k(2^{k+1}\\cdot m+2^k+n)$. So it suffices to show that $2^{k+1}\\cdot m+2^k+n$ is divisible by $5$ for one of $m=0,1,2,3,4$. However this is basically the same as showing that $2^{k+1}\\cdot m$ each produce a different residue mod $5$ as $m$ varies which is obvious. Thus, at least one of these values for $m$ will give the expression to be divisible by $5$.\n\nThus, our induction is complete and our claim is true.[/hide]",
"Solution_15": "We proceed by induction on $n$. When $n=1$, then number is $5$.\n\\\\$~$\\\\\nSuppose $\\overline{c_{n-1}c_{n-2}\\dots c_1c_{0}}=5^n\\cdot k$ is the $n$-digit number described in the conditions, then we need \\[5^{n+1}\\mid c_n\\cdot 10^n+5^n\\cdot k=5^n(2c_n+k)\\implies 5\\mid (2c_n+k).\\] Since $\\gcd(2,5)=1$, there is a solution $c_n\\equiv c\\pmod 5$ where $0\\le c\\le 4$. Now, pick either $c$ or $c+4$ and one of them will be odd. Therefore, we are done.",
"Solution_16": "We induct on $n$; the base case $n=1$ is obvious. Now, consider a valid number $\\overline{a_1a_2\\cdots a_n}$ with all odd digits divisible by $5^n$. Then, one of the five numbers given by $\\overline{k a_1a_2\\cdots a_n}$, with $k$ an odd integer between $1$ and $9$, must be divisible by $5^{n+1}$, as $5^{n+1} \\nmid 2\\overbrace{00\\cdots 0}^{n \\text{ zeroes}}$ by Pigeonhole, which completes the induction.",
"Solution_17": "Use induction\nwhen $n=1$, $5$ works\nif $n$ works, then $a*10^n+\\text{previous value that worked}$ is divisible by $5^n$. There are $5$ possible values and each of them has a different remainder when divided by $5^{n+1}$ because of pigeonhole, so one of the must divide $5^{n+1}$",
"Solution_18": "USSR 1967 or All-Russian 1967\n[url]https://artofproblemsolving.com/community/c893771h1859931p12577115[/url]",
"Solution_19": "When $n = 1$, we see that $5$ works. We will now induct on $n$. Let $a_k$ be our desired positive integer with $k$ digits, where we write $a_k = 5^kr$ for some odd positive integer $r$. Then we claim that we can construct $a_{k + 1}$ by appending some odd digit $d$ to the start of $a_k$, so we want \n$$a_{k + 1} = d \\cdot 10^k + a_k = 5^k(d \\cdot 2^k + r). $$\nto be divisible by $5^{k + 1}$, or $d \\cdot 2^k + r$ to be divisible by $5$. But since $2^k$ and $5$ are relatively prime for all positive integers $k$ and $1$, $3$, $5$, $7$, and $9$ all have distinct residues modulo $5$, we will always be able to find exactly one suitable $d$ such that $5 \\mid (d \\cdot 2^k + r)$. This completes the proof. \\\\",
"Solution_20": "Failed on this some months ago. Tried again and did it!\n\n[i]Solution:[/i] We will perform induction on $n$. Some base cases are \n[list]\n [*] For $n = 1$, take 5,\n [*] For $n = 2$, take 25, \n [*] For $n = 3$, take 375,\n [*] For $n = 4$, take 9375.\n[/list]\nAs an induction hypothesis, assume there exists a number (say $X$) with $k$ digits divisible by $5^k$ with all odd digits. Write $X$ as \n \\[X = \\sum_{n = 0}^{k}10^{n-1}\\cdot a_n\\]\nwhere $a_i$ is a digit. For the induction step, consider number $X' = X + 10^kb$ where $b$ is a digit. We need to show that there exists an odd $b$ such that $X'$ is divisible by $5^{k+1}$. Reduce $X'$ modulo $5^{k+1}$ with the fact that $X = 5^k \\cdot \\alpha$ for positive integer $\\alpha$. \n \\[X' \\equiv 0 \\pmod{5^{k+1}} \\iff b\\cdot 2^k + \\alpha \\equiv 0 \\pmod{5}\\]\nSince $5 \\nmid 2^k$, there will always be a solution to the congruence for $b$. If the smallest positive residue of $b$ is odd, then we are done, otherwise add 5 to it to make it odd. Everything will obviously work by tracing the work backwards. $\\blacksquare$",
"Solution_21": "Sketch: \n\nWe can prove this inductively. \n\nBase Case: $n = 5$ clearly abides. \n\nInduction: Take the number $k$ digit number that is divisible by $5^k$. Then to construct a $k+1$ digit number that is divisible by $5^{k+1}$, we can either append $1, 3, 5, 7, 9$ to the left of the number. One of these selections is legal (we can see this through algebra and some basic principles of modular arithmetic).",
"Solution_22": "proceed by induction. Base case n=1 is trivial. Now, we'll append an odd digit onto the left side of the old $n$ digit number $a_n$. $\\frac{a_n}{5^n}=b_n\\implies 5^n(2^nk+b_n)$. If $b_n$ is divisible by 5 we can just choose k=5; if not, noting that $$5\\nmid 2^n,b_n\\implies b_n=\\{1,3,5,7,9\\}\\equiv 2^nk\\pmod 5,$$ since 1,3,5,7,9 is a complete residue set and so is $2^nk$, hence we can always choose a $k\\le 10$ such that this works.",
"Solution_23": "We will show that if there exists a $k$ digit integer divisible by $5^{k}$ there exists a $k+1$ digit integer divisible by $5^{k+1}$.\nIf we add a digit $a$ to the far left of are $k$ digit integer we will show there exists an $a$ such that this is divisible by $5^{k+1}$\nLet are $k$ digit integer be $c$ then we require\n $$c+a10^{k}\\equiv 0 \\pmod{5^{k+1}}$$\n$$\\frac{c}{5^{k}}+2^{k}a \\equiv 0 \\pmod{5}$$\nSince we get a complete residue set mod $5$ as possibilities for $a$ the desired number exists.\nSince $5$ has $1$ digit and is divisible by $5$ we have completed our induction.",
"Solution_24": "We prove this statement by induction. The base case for $n = 1$ is trivial, as $5^1 \\mid 5$.\n\nAssume that for all $N \\leq n$, we have $k \\cdot 5^n$ has all of its $n$-digits as odd. (Observe that $k \\in \\{1, 3, 5, 7, 9\\}$). Consider the $(n+1)$-digit integer $\\ell \\cdot 10^n + 5^n$. Factoring yields, \\[5^n(\\ell \\cdot 2^n + k), \\phantom{ccc}\\ell \\in \\{1, 3, 5, 7, 9\\}.\\] Notice that the inside expression goes through all residues modulo $5$. Hence for every value of $\\ell$, there is some $k$ for which $(\\ell \\cdot 2^n + k)$ is divisible by $5$, so that the entire expression including the factor of $5^n$ is divisible by $5^{n+1}$, which finishes the induciton. $\\blacksquare$",
"Solution_25": "Consider induction on $n$. When $n=1$, the desired number is $5$, and when moving forward, consider a working $n$-digit number to be $N$. Then, simply notice that exactly one of the elements in \n\n\\[S = \\{10^n+N, 3 \\cdot 10^n+N, 5 \\cdot 10^n+N, 7 \\cdot 10^n+N, 9 \\cdot 10^n+N\\}\\] \n\nwill satisfy the conditions for an $(n+1)$-digit number, so we are done. $\\square$",
"Solution_26": "We proceed with induction. The base case, $n=1$ is obvious by just taking $5$ as our number. For our inductive step, assume $k$ is an $n$-digit number divisible by $5^n$ with all odd digits. We know that $k\\equiv 0 \\pmod {5^n}$, so $k$ can be congruent to either $0, 5^n, 2(5^n), 3(5^n)$, or $4(5^n)$ modulo $5^{n+1}$. Notice that the $n+1$ digit number $x(10^n) = x\\cdot 2^n (5^n)$ as $x$ ranges over $1, 3, 5, 7, 9$ can only be $0, 5^n, 2(5^n), 3(5^n)$, or $4(5^n)$ modulo $5^{n+1}$. In fact, each of these five residues appears exactly once due to $\\gcd(5, 2^n) = 1$. Thus, given a $k$, we can always pick an $x(10^n)$ that makes $k+x(10^n)$ an $n+1$ digit number with all odd digits and $k+x(10^n)\\equiv 0\\pmod {5^{n+1}}$. This completes our induction, so we're done. $\\blacksquare$",
"Solution_27": "We will use induction.\n$\\newline$\nLet our $n$-digit number be $k_n$.\nThen our base case of $n = 1$ works for $k_1 = 5$.\n$\\newline$\n\nInductive Step:\n$\\newline$\n\nAssume that $k_n$ works. Then we can prove that $k_{n+1}$ works.\nLet $k_{n+1} = a \\cdot 10^n + k_{n}$ for some $a \\in \\{1, 3, 5, 7, 9\\}$.\nAnd since all of $\\{1, 3, 5, 7, 9\\}$ are distinct modulo $5$, then there must be some $a$ that makes $k_{n+1} \\equiv 0\\pmod{5^{n+1}}$ since both $10^n$ and $k_n$ are divisible by $5^n$, so we are done. \n",
"Solution_28": "[quote=Davron]Who can show that the first digit number of the k-digit number can be never equal to 1 ?\n\nDavron[/quote]\n\n\nConsider 193,359,375",
"Solution_29": "we will use induction\nwe first have that there exists $(2x_0+1)*10^0$ that is divisible by $5^1$\nnext, assume there exists $(2x_0+1)*10^{n-1}+...+(2x_n+1)*10^0$ divisible by $5^n$\nmultiply by $10$ to get $2(x_1*10^n+...+x_n*10)+\\frac{10^{n+1}-10}{9}$\nthis is divisible by $5^{n+1}$\nadd one to get $2(x_1*10^n+...+x_n*10)+\\frac{10^{n+1}-1}{9} \\equiv 1 (\\text{mod }5^{n+1})$\nwe need $2(y_1*10^n+...+y_n) \\equiv -1 (\\text{mod }5^{n+1})$\nthe existence of this is trivial since $5^{n+1}-1$ is even \ntherefore, we are done"
}
{
"Tag": [
"geometry",
"abstract algebra",
"complex numbers"
],
"Problem": "If $a , b , c$ are dinstict complex numbers such that $a^2 + b^2 + c^2 = ab +bc +ca$ , prove that the images of $a ,b , c$ form a equilateral triangle.\r\n\r\n [b]NOTE[/b]\r\n\r\n Virgil has a nice solution there. Please , try to find - if possible - a solution using modules and simple algebra. Not roots of unit , or trigonometrical form. I repeat:If possible. Perhaps such a solution is impossible.I also try to find it.\r\n\r\n Thanks a lot !\r\n\r\n [u]Babis[/u]",
"Solution_1": "http://www.artofproblemsolving.com/Forum/viewtopic.php?p=284898&highlight=#284898",
"Solution_2": "Actually here is a nice solution to this one.\r\n\r\nThe relation that Stergiu wrote is equivalent with this one\r\n \\[ (a-b)^{2}=(b-c)(c-a)\\Leftrightarrow \\]\r\n \\[ |a-b|^{2}=|b-c||c-a|. \\]\r\nWriting the analogues, and by adding them up, we get \\[ |a-b|^{2}+|b-c|^{2}+|c-a|^{2}=|a-b||b-c|+|b-c||c-a|+|c-a||a-b|. \\]\r\n From here we easily get $|a-b|=|b-c|=|c-a|$. :) ;)"
}
{
"Tag": [
"AMC"
],
"Problem": "Has the difficulty of the AMC 12 changed over the years or has it stayed relatively constant? I don't know if it is just me but it seems as though the tests after 2007 are easier (because of no calculators allowed now?)...",
"Solution_1": "This is to some extent, true(about the easiness thing). This works for all AMCs. For example, take the 2009 AMC10. You'll find it extremely easy compared to past years, mainly because they outlawed calculators.",
"Solution_2": "I got a 126 on the 2006 A AMC 12, but I didn't use a calculator. Does anyone have an idea of to what score this would correspond on a non-calculator AMC 12?",
"Solution_3": "[quote=\"MathWise\"]I got a 126 on the 2006 A AMC 12, but I didn't use a calculator. Does anyone have an idea of to what score this would correspond on a non-calculator AMC 12?[/quote]There really is no set number, it depends on your mental math skills and how much hard mental math there was at that time.",
"Solution_4": "[quote=\"MathWise\"]I got a 126 on the 2006 A AMC 12, but I didn't use a calculator. Does anyone have an idea of to what score this would correspond on a non-calculator AMC 12?[/quote]\r\n\r\nAs Ihatepie said, this corresponds with your own mathematical ability. Basically, the advantage with calculators is that you save time, and considering that you got 126 without a calc, with a calc you could probably boost this score up by 6 points (maybe higher, again depending on your mathematical ability).",
"Solution_5": "so would any of you agree that the AMC 12s after 2007, for the most part, are easier than the ones of 2007 and earlier?"
}
{
"Tag": [
"combinatorics unsolved",
"combinatorics"
],
"Problem": "Suppose we have $ 2n$ integers between $ \\minus{}n$ and $ n$ whose sum equals one (Where $ n$ is a natural number bigger than one). Prove that the sum of a subset of these $ 2n$ integers equals zero.",
"Solution_1": "[quote=\"theDeviation\"]Suppose we have $ 2n$ integers between $ \\minus{} n$ and $ n$ whose sum equals one (Where $ n$ is a natural number bigger than one). Prove that the sum of a subset of these $ 2n$ integers equals zero.[/quote]\r\n\r\nFirst of all i assume you wanted to include -n and n while choosing 2n numbers, or else there are only 2n-1 numbers in between.\r\nNext check that sum of all no. from -n to n = 0. also there are 2n+1 no. Hence we must choose all no. except -1. The subset {0} works. :wink:",
"Solution_2": "I don't think he assumes the numbers to be distinct...\r\n\r\n darij",
"Solution_3": "For the case $ n\\equal{}1000$, this was the third problem of the Canadian Mathematical Olympiad in the year 2000.\r\n\r\nHere's the solution: [url]http://www.math.ca/Competitions/CMO/solutions/sol_2000.pdf[/url]"
}
{
"Tag": [],
"Problem": "Determine $ a,b\\in \\mathbb{Z}$ such that $ a^4\\plus{}b^4$ is a prime number and $ 2a^4\\plus{}2b^4\\minus{}1$ is a perfect square .",
"Solution_1": "There aren't any.\r\n\r\nSuppose $ a^4\\plus{}b^4 \\equal{} p$, a prime number, and $ 2a^4\\plus{}2b^4 \\minus{} 1 \\equal{} n^2$, a square. $ n$ is obviously odd so we may write $ n \\equal{} 2m\\plus{}1$ and then\r\n\r\n$ 2a^4 \\plus{} 2b^4 \\minus{} 1 \\equal{} (2m\\plus{}1)^2 \\equal{} 4m^2 \\plus{} 4m \\plus{} 1$\r\n\r\nso\r\n\r\n$ a^4 \\plus{} b^4 \\equal{} 2m^2 \\plus{} 2m \\plus{} 1$\r\n\r\nbut the LHS is just $ p$ so we have\r\n\r\n$ p \\equal{} m^2 \\plus{} (m\\plus{}1)^2$.\r\n\r\nWe now have $ p$ written as a sum of $ 2$ squares in two ways: $ p \\equal{} a^4 \\plus{} b^4$ and $ p \\equal{} m^2 \\plus{} (m\\plus{}1)^2$. But it's well-known that if a prime number can be written as a sum of $ 2$ squares, then it can be so written in an essentially unique way. It follows that $ m\\equal{}a^2$ and $ m\\plus{}1\\equal{}b^2$ or vice versa, which is absurd.\r\n\r\nThe most elegant and conceptual proof that $ p\\equal{}a^2\\plus{}b^2$ has a unique solution (if it has any solutions at all) relies on simple properties of the ring $ \\mathbb{Z}[i]$. This must be written up here somewhere, but if you're not familiar with it I can provide a brief account."
}
{
"Tag": [
"abstract algebra",
"group theory",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Let $G$ be finite abelian group in which the number of solutions in $G$ of equation $x^n=e$ is at most $n$ for every positive integer $n$. Prove that $G$ must be cyclic group.",
"Solution_1": "Just count how many of each order are there (for the abelian case it's simple to show that $x^d=e$ has at least $\\gcd(d, |G|)$ solutions).\r\nBy http://www.mathlinks.ro/Forum/viewtopic.php?t=22490 the word abelian is not necessary.",
"Solution_2": "I'm new about this, Can you write the proof ZetaX?\r\n\r\nthanks",
"Solution_3": "Claim: for $d$ dividing $|G|$ there are at least $d$ solutions to $x^d=e$.\r\n\r\nThe prove for all groups is in the link.\r\n\r\nThe prove for abelian groups only:\r\n\r\nWhen $a,b$ are elements with coprime orders, then $ab$ has the order of their product (not hard to show).\r\n\r\nThus it suffices to show that for any prime power $p^n$ dividing $|G|$ there are at least $p^n$ solutions to ${x^{p^n}}=e$, since then for any $d=p_1^{v_1} \\cdot p_2^{v_2} \\cdot ... \\cdot p_m^{v_m}$ dividing $|G|$ we can take products $x=x_1 x_2 ... x_m$ where $x_i^{p_i^{v_i}} = e$ and there are enough such products (they will all be different, since when we have two such produvts $x,y$, then by the order-of-product-theorem above $xy^{-1}$ has an order $>1$ when they are different).\r\nSo it is reduced to the prime powers, where it follows from the Sylow theorem (which can be shown a lot easier for abelian groups), because we then have subgroups of any prime power dividing $|G|$, and Lagranges theorem gives us that all these elements of that subgroup are what we want.\r\n\r\nAnother way to prove this would be to use that every finite abelian group is a product of cyclic groups.\r\n\r\n\r\nSo by the above and the statement we have that there are at most and at least $d$ solutions to $x^d=e$, thus exactly $d$, for any divisor $d$ of $|G|$.\r\nNow start counting how many solutions are there for $x^{|G|}=e$ that are not solutions of 'smaller' equations.\r\nWell, when we write $y(d)$ for the number of elements with order $d$, we have that $n= \\sum_{d|n} y(d)$ for any $n$.\r\nThus the m\u00f6bius inversion formula tells us that there are $y(d) = \\sum_{n|d} \\mu (d/n) n = \\phi(d)$ elements with order $d$, and $y(|G|)$ is clearly $>0$, thus there is a generating element.",
"Solution_4": "thanks for the proof ZetaX. \r\n\r\ni don't remember that i've post this problem, until my prof gave this problem last week.\r\nBut i think there is another solution which is more elementary..\r\nthis problem appears in the early section, that is about order of group.",
"Solution_5": "Yes you are right, it can be done more directly:\r\n\r\nLet $ \\psi(d)$ be the number of elements of order $ d$. Every single $ x$ with order $ d$ gives already $ d$ different $ y$ with $ y^{d}\\equal{} e$, namely the powers of $ x$, and exactly $ \\phi(d)$ of these powers have order $ d$ again. As a result, $ \\psi(d)\\leq\\phi(d)$.\r\nNow $ n \\equal{}\\sum_{d|n}\\psi(d)\\leq\\sum_{d|n}\\phi(d) \\equal{} n$, thus we always have equality in $ \\psi(d)\\leq\\phi(d)$, especially $ \\psi(|G|) \\equal{}\\phi(|G|) > 0$, which means that the group is cyclic.",
"Solution_6": "Yes, that's what i mean, and that is also my solution after two day searching for :blush: \r\n\r\nthanks for the smart reply,"
}
{
"Tag": [
"LaTeX"
],
"Problem": "Howdy\r\nI am both new to Latex and to Mac.\r\nI have used Latex without too many problems but today i needed to use the combinations notation and discovered that I needed to install amsmath to do this.\r\nI followed the instructions from the Ams link \r\nftp://ftp.ams.org/pub/tex/amslatex/install.txt\r\ni managed to download the relevant files and stored them in my Applications folder. The compiler still doesn't recognise the \\binom command however. I have a feeling i should be careful about what folders the downloaded files should be stored but haven't a clue as to what to do next.\r\nAny help would be appreciated.\r\nMark",
"Solution_1": "I'm not familiar with LaTeX on the Mac but I can't believe any LaTeX distribution comes without amsmath installed as it is such an important part of it. \r\nWhen you put \\usepackage{amsmath} in your document does the log say it can't find the package? Do any amsmath files (such as amsmath.sty) exist on your computer, apart from the ones you've downloaded?\r\n\r\nPlease post which LaTeX distribution you have installed and then maybe others with the same distro will be able to advise.",
"Solution_2": "Problem solved...as i said i am new to Latex!...a friend spotted that i had put \\usepackage{amsmath} after the begin{doc...} command instead of before!\r\nThanks anyway.\r\nMark"
}
{
"Tag": [
"function",
"algebra unsolved",
"algebra"
],
"Problem": "Find all functions $ f: \\mathbb R\\rightarrow\\mathbb R$ satisfying $ f(x\\plus{}f(xy)\\equal{}x(1\\plus{}f(y))$ for every $ x,y\\in\\mathbb R$",
"Solution_1": "It is very easy.\r\nLet $ P(x,y)$ be the mentioned property of f.\r\nThen:\r\n$ (P(0,y))\\Rightarrow (f(f(0))\\equal{}0)$ let us have $ a\\equal{}f(0)$ then\r\n$ f(a)\\equal{}0$\r\n$ (P(1,a))\\Rightarrow (f(1)\\equal{}1)$ and now\r\n$ (P(a,1))\\Rightarrow (0\\equal{}f(a)\\equal{}f(a\\plus{}f(a))\\equal{}2a)$ so $ f(0)\\equal{}0$\r\nNow from $ P(x,0)$ we conclude that $ f(x)\\equal{}x$ :wink:"
}
{
"Tag": [],
"Problem": "Let $ABCD$ be a square with diagonals $AC$ and $BD$, and $P$ a point in one of the sides of the square. Show that the sum of the distances from P to the diagonals is constant.",
"Solution_1": "[hide=\"Fun solution\"]\nLet $AC$ be the positive $y$-axis and let $BD$ be the negative $x$-axis. WLOG $P$ is on $AB$. Then the line $AB$ is clearly the graph $y + x = \\frac{AB}{ \\sqrt{2} }$, and $x, y$ are the distances to the diagonals! [/hide]",
"Solution_2": "Let the distance from $P$ to BD and AC be $x$ and $y$ respectively and $O$ be the centre of the square. Then:\r\n\r\n$x=\\frac{AO}{AB}\\cdot PB$ and $y=\\frac{BO}{AB} \\cdot PA$. Sum up we get $x+y=BO=\\frac{AB}{\\sqrt{2}}$."
}
{
"Tag": [
"geometry",
"parallelogram",
"vector",
"invariant",
"geometry unsolved"
],
"Problem": "How do you prove that the medians of a quadrilateral bisect each other? I think I have to prove that the quadrilateral created by connecting the midpoints of the original quad is a parallelogram (aka the Varignon quadrilateral). But I can't figure out how to prove this. Any suggestions?",
"Solution_1": "Vectors!\r\n\r\nWrite each of the points of the original quadrilateral as vectors $\\vec{a},\\vec{b},\\vec{c},$ and $\\vec{d}.$\r\n\r\nThen the midpoints of the sides are $\\frac12(\\vec{a}+\\vec{b}), \\frac12(\\vec{b}+\\vec{c}), \\frac12(\\vec{c}+\\vec{d}),$ and $\\frac12(\\vec{d}+\\vec{a}).$\r\n\r\nIt turns out that we're now setup to find the midpoints of the diagonals:\r\n\r\n$\\frac12\\left[\\frac12(\\vec{a}+\\vec{b})+\\frac12(\\vec{c}+\\vec{d})\\right]= \\frac14(\\vec{a}+\\vec{b}+\\vec{c}+\\vec{d})$\r\n\r\n$\\frac12\\left[\\frac12(\\vec{b}+\\vec{c})+\\frac12(\\vec{d}+\\vec{a})\\right]= \\frac14(\\vec{a}+\\vec{b}+\\vec{c}+\\vec{d})$\r\n\r\nWe didn't even go through the proof that the smaller quadrilateral is a parallelogram, but that is also straightforward in this setup.",
"Solution_2": "Thank you Kent. I was anticipating a more \"geometric\" proof using Euclidean theorems, but now I see a different way to look at it.",
"Solution_3": "I think that we don't need to solve the basic theorems by vectors, because we will lose, the meaning of Euclidean geometry.\r\n\r\n$\\bullet$ We denote as $K,$ $L,$ $M,$ $N,$ the midpoints of the sidesegments $AB,$ $BC,$ $CD,$ $DA$ respectively, of an arbitrary quadrilateral $ABCD.$\r\n\r\nSo, we have that $KL\\parallel = \\frac{AC}{2}\\parallel = MN.$\r\n\r\nHence, the quqdrilateral $KLMN,$ is a parallelogram and the intersection point $Q,$ of it's diagonals, bisects them.\r\n\r\nSimilarly, if $P,$ $R,$ are the midpoints of $AC,$ $BD$ respectively, we have that $PL\\parallel = \\frac{AB}{2}\\parallel = NR.$\r\n\r\nThat is the point $Q,$ bisects all the medians of $ABCD.$\r\n\r\nKostas Vittas.",
"Solution_4": "One thing that the vector proof shows is that this is an affine-invariant theorem. That is, all of the points we're concerned with would retain their identities under any (invertible) affine transformation of the plane."
}
{
"Tag": [
"logarithms"
],
"Problem": "X,Y,Z are positive real numbers satisfying\r\n\r\n$\\frac{X(Y+Z-X)}_{logX}$ = $\\frac{Y(Z+X-Y)}_{logY}$ = $\\frac{Z(X+Y-Z)}_{logZ}$\r\n\r\nProve $Y^{Z}.Z^{Y}$ = $Z^{X}.X^{Z}$ =$X^{Y}.Y^{X}$",
"Solution_1": "[hide]\nLet us assume $ \\frac{x(y+z-x)}{\\log{x}}=\\frac{y(x+z-y)}{\\log{y}}=\\frac{z(x+y-z)}{\\log{z}}=k$\n$\\log{y}=\\frac{y(x+z-y)}{k}=\\frac{xy+yz-y^{2}}{k}$\n$\\log{x}=\\frac{x(y+z-x)}{k}=\\frac{xy+xz-x^{2}}{k}$\n\\begin{eqnarray*}\\log{y}-\\log{x}&=& \\log{\\frac{y}{x}}\\\\ &=& \\frac{yz-xz-y^{2}+x^{2}}{k}\\\\ &=& \\frac{(x-y)(x+y-z)}{k}\\\\ &=& \\frac{(x-y)\\log{z}}{z}\\end{eqnarray*}\n\\begin{eqnarray*}\\frac{(x-y)\\log{z}}{z}&=&\\log{\\frac{y}{x}}\\\\ (x-y)\\log{z}&=&z\\log{\\frac{y}{x}}\\\\ z^{x-y}&=&\\left(\\frac{y}{x}\\right)^{z}\\\\ \\frac{z^{x}}{z^{y}}&=&\\frac{y^{z}}{x^{z}}\\\\ z^{x}x^{z}&=&y^{z}z^{y}\\end{eqnarray*}\nand similarly for the other cases\n[/hide]"
}
{
"Tag": [
"\\/closed"
],
"Problem": "nice un :rotfl: \r\n\r\nbtw im not able to post/send pms at all wen im on my pc. i have to go on my dads laptop every time, which is irritating coz i get that opportunity on once in a while :( even rohit had the same prob on his comp. \r\nany ideas why, any1??",
"Solution_1": "i moved this here from the india forum",
"Solution_2": "Is it because no one has replied or does it not let you send one alltogether? If it's the latter, you've probably been banned from pming.",
"Solution_3": "Is there something like you have to send a min no of posts before you get PM'ing priveledges?",
"Solution_4": "I do not believe there is no requirement for pming (there might be a wait period for new users)."
}
{
"Tag": [
"inequalities"
],
"Problem": "I wrote this problem while investigating interesting integer functions. The \"f(3n)=f(2n)\" part below appears on the Mandelbrot Problem Corner.\r\n\r\nf:{0,1,2....}->{0,1,2,...}\r\ni) f(3n)=2f(n)\r\nii) f(3n+1)=f(3n)+1\r\niii) f(3n+2)=f(3n)+2, for all integers n.\r\nProve: f(3n):ge:f(2n)>f(n).\r\nAlso, when does f(3n)=f(2n), and when does f(2n) = f(n) + 1?",
"Solution_1": "grommit wrote:I wrote this problem while investigating interesting integer functions. The \"f(3n)=f(2n)\" part below appears on the Mandelbrot Problem Corner.\n\nf:{0,1,2....}->{0,1,2,...}\ni) f(3n)=2f(n)\nii) f(3n+1)=f(3n)+1\niii) f(3n+2)=f(3n)+2, for all integers n.\nProve: f(3n):ge:f(2n)>f(n).\nAlso, when does f(3n)=f(2n), and when does f(2n) = f(n) + 1?\n\n\n\nAn \"elegant\" brute-forcing...actually its not elegant at all.\n\n\n\n[hide]\n\nThis is a well-defined function. We know by i that f(0) = 0, so by the ii and iii, we get f(1) = 1 and f(2) = 2. We know f(3) = 2(1) = 2, and f(4) = 3, f(5) = 4. \n\n\n\nContinuing, we see that f(6,7,8) = 4,5,6, f(9,10,11) = 4,5,6, f(12,13,14) = 6,7,8, f(15,16,17) = 8,9,10, f(18,19,20) = 8,9,10...\n\n\n\nf(3n) thus retains a pattern of 0,2,4,4,6,8,8,10,12....\n\nNamely, f(3n) = f(3(n+3))+4, about f(n) = 4/3*n.\n\n\n\nf(2n) retains a pattern of 0,2,3,4,6,5,6,8,9,8,10,11,12,14...\n\nf(2n) = f(2(n+10)) + 10, about f(n) = n.\n\n\n\nf(n) gives a pattern of 0,1,2,2,3,4,4,5,6,4,5,6,6,7,8,8,9,10....\n\nf(n) = f(n+9) + 4, about 4/9*n.\n\n\n\nWe can trivially prove the first few, and thus f(3n):ge:f(2n)>f(n)\n\n\n\nf(3n) = f(2n) when n = 0,1,3,4, and nowhere else.\n\nf(2n) = f(n) + 1 when n = 1,2,5 and nowhere else.\n\n[/hide]",
"Solution_2": "*working*",
"Solution_3": "Ummm......masamune, your answers are false... for instance, there are infinitely many n such that f(3n) = f(2n)... not sure what you did wrong, since there don't seem to be any proofs in you post... would you mind posting your reasoning so that we can work out what is wrong with it? It is very tempting to assume that patterns here will continue in ways that they do not.\r\n\r\nHaving already seen this problem, I won't post any solutions unless someone wants me to. A very nice problem, by the way.",
"Solution_4": "Ah. The pattern doesn't continue...my bad. I'll come back later to fix this solution...",
"Solution_5": "Hmm...can we generalize anything about f(p^k*3^n), where p is a prime?",
"Solution_6": "[hide]\n\nLemma: f(n) returns n, written in base 3, read in base 2 . \n\n \n\nProof is by induction. If n = 3k + c, with c < 3, we induct on k. \n\n \n\nIf k = 0: \n\nf(0) = 2f(0) => f(0) = 0.\n\nf(1) = f(0) + 1 = 1\n\nf(2) = f(0) + 2 = 2, so the claim is satisfied for k=0.\n\n \n\nIn general,\n\nf(3(k+1) + c) = f(3(k+1)) + c = 2f(3k) + c. Since f(3k) is the base 3 representation read in base 2, we see that 3(k+1) has a base 3 representation shifted one place to the left from 3k, and that f(3(k+1)) = 2f(3k), so f(3(k+1)) has its base 2 representation shifted one place to the left of f(3k). Therefore, the lemma is proven.\n\n---\n\n\n\n\n\nLet k = a_n a_(n-1) a_(n-2) ... a_0 in base 3. Then\n\nf(k) = :Sigma: 2i a_i . \n\n\n\nLet 2k = b_(n+1) b_n ... b_0 in base 3. Then letting c_n represent potential carries in addition base 3...\n\n\n\nb_0 = 2a_0 - 3c_0\n\nb_1 = 2a_1 + c_0 - 3c_1\n\n.\n\n.\n\n.\n\nb_n = 2a_n + c_(n-1) - 3c_n\n\nb_n+1 = c_n\n\n\n\nf(2k) = :Sigma 2i b_i = 2:Sigma: 2ia_i - :Sigma: 2i c_i = 2f(k) - :Sigma: 2i c_i :le: 2f(k) = f(3k). \n\n\n\n Therefore, f(3k) :ge: f(2k), with equality iff :Sigma: 2i c_i = 0 => there are no carries in the multiplication of k by 2, in base 3 => there are only 1's and 0's in the base 3 representation of k. \n\n\n\nSince we have f(2k) :le: f(3k), we now only need to show f(2k) > f(k).\n\n\n\nf(2k) = :Sigma 2i b_i = 2 :Sigma: 2ia_i - :Sigma: 2i c_i = :Sigma: 2ia_i + :Sigma: 2i(a_i - c_i) = f(k) + :Sigma: 2i(a_i - c_i).\n\n\n\nSo, we want to show :Sigma: 2i(a_i - c_i) > 0.\n\n\n\nTaking cases: \n\na_i = 0 => c_i = 0.\n\na_i = 1 => c_i = 0 , or c_i 1 if c_(i-1) = 1\n\na_i = 2 => c_i = 1.\n\n\n\nSo a_i :ge: c_i for all i, but we need strict inequality in at least one case.\n\n\n\nIf a_i = 2 for some i we are guaranteed strict inequality. \n\n\n\nSuppose a_i = 0 or 1 for each i.\n\n\n\nThen there are no carries => c_i never equals 1 => c_i = 0 when a_i = 1. This yields strict inequality.\n\n\n\nTherefore, f(2k) > f(k) , so we have our combined inequality:\n\n\n\n f(3k) :ge: f(2k) > f(k) .\n\n\n\n\n\nNote also, that we can find when f(2k) = f(k) + 1 by this procedure. \n\n\n\nSince f(2k) = f(k) + :Sigma: 2i(a_i - c_i), we want :Sigma: 2i(a_i - c_i) = 1, or:\n\n\n\na_0 = c_0 + 1 and a_j = c_j for all j > 0. \n\n\n\nLet's take the two cases separately for a_0.\n\n\n\na_0 = 1 => c_0 = 0. Then a_j = c_j for larger j iff a_j = 0, so we get k = ...0013 = 1.\n\n\n\na_0 = 2 => c_0 = 1, so a_1 = 1 => c_1 = 1, and we can continue in a block of this fashion as far as we want. Once some a_j = 0, we can only have zeroes for any larger j. \n\n\n\n Therefore, f(2k) = f(k) + 1 when k = ...1123 (a block of 1's followed by a 2) or when k = 1. \n\n[/hide]\n\nNow that took a lot of time to write up!"
}
{
"Tag": [
"induction",
"inequalities",
"inequalities unsolved"
],
"Problem": "sorry for asking two problems concurrently.\r\n\r\nLet $a_{1},a_{2},...,a_{n}$ be positive numbers, and let $A,G,H$ be an arithmetic, geometric, and harmonic mean of that numbers.\r\n(a) If $n$ is even, show that\r\n$\\frac{A}{H}\\leq-1+2 (\\frac{A}{G})^{n}$\r\n\r\n(b) If $n$ is odd, show that\r\n$\\frac{A}{H}\\leq-\\frac{n-2}{n}+\\frac{2(n-1)}{n}(\\frac{A}{G})^{n}$",
"Solution_1": "On these types of problems, is the main way to go induction? The peculiar even/odd condition makes me think that...",
"Solution_2": "The MacLaurin inequality says that:\r\n\r\n$GM=\\sqrt[n]{d_{n}}\\leq \\sqrt[n-1]{ d_{n-1}}\\leq ... \\leq \\sqrt{d_{2}}\\leq d_{1}= AM$\r\n\r\nWhere $d_{k}$ are the aritmetic mean of the products with $k$ elements. For example with 4 variabiles $d_{2}= \\frac{1}{6}(ab+bc+cd+da+bd+ca)$.\r\n\r\nNow we have that $HM=\\frac{GM^{n}}{d_{n-1}}$ and so our inequality becomes:\r\n\r\n${AM * d_{n-1}}{GM^{n}}\\leq-1+2 \\frac{ AM^{n}}{GM^{n}}$\r\n\r\n$\\frac{AM (AM^{n-1}+AM^{n-1}-d_{n-1})}{GM^{n}}\\geq 1$\r\n\r\nNow we know that $AM^{n-1}\\geq d_{n-1}$ and so\r\n\r\n$\\frac{AM (AM^{n-1}+AM^{n-1}-d_{n-1}) }{GM^{n}}\\geq \\frac{AM *AM^{n-1}}{GM^{n}}= \\left( \\frac{AM}{GM}\\right)^{n}\\geq 1$\r\n\r\nand so the case with $n$ even is done. The case with $n$ odd is the same calculations...",
"Solution_3": "in fact, from MacLaurin it follows directly $\\frac{A}{H}\\leq (\\frac{A}{G})^{n}$ :)"
}
{
"Tag": [
"Pascal\\u0027s Triangle"
],
"Problem": "Note: There are six rows of buckets as shown on the picture.\n\n\n\n\n\n\n\nI used recursion to get the answer (really long solution)... is there a faster way? Pascal's Triangle?\n\n\n\nAnswer: [hide][/hide]",
"Solution_1": "Let $ x$ litres be poured into the top.\r\n\r\nNow, consider the problem from a different viewpoint. The buckest will distribute the water the same way no matter how they receive it as long as there is the same amount of water given to it. So we start with the amount of water the top bucket will receive, $ x$. It takes 1, and distributes the remaining halves to the 2 buckets underneath. Continue in this fashion to get a lengthy equation for the bottom buckets. Then set them equal to 1, and pwn.",
"Solution_2": "the problem is that some buckets will receive water at different rates. And at the same time another bucket will receive water at a different rate... setting up a lengthy equation would give me a headache keeping track of all the buckets",
"Solution_3": "That's the beauty of it. It doesn't matter at what rate the buckets get the water, only the total amount of water received by each bucket.",
"Solution_4": "i still don't quite understand your method.. are you assuming each bucket is 1L instead of 2 to simplify the problem?\r\n\r\nstarting with $ x$ litres from the top:\r\nthe two buckets underneath will get $ \\frac {x \\minus{} 2}{2}$ each..\r\nbucket 1&3 of row 3 will get $ \\frac {(\\frac {x \\minus{} 2}{2}) \\minus{} 2}{2}$ and bucket 2 will get $ \\frac {x \\minus{} 2}{2}$...\r\nbucket 1 of row 3 will get $ \\frac {\\frac {(\\frac {x \\minus{} 2}{2}) \\minus{} 2}{2} \\minus{} 2}{2}$ \r\n\r\ncontinuing this pattern will be a mess when it gets to the 4th row... (especially for the buckets in between)\r\n\r\nis this what you are suggesting? :blush: \r\n\r\nLooking at the arrangements of the buckets you can say for sure that bucket $ C$ and bucket $ D$ will both be the first to receive $ 2$ litres. (Pascal's method)\r\nI'm not sure if that information would help :|",
"Solution_5": "Does sombody knew how we use the pascal triangle here? I really guess we have to use them to solve it. It's the secret of the bukets beetwen. If we concidere the do the thing manualy? I will se if I found an algoritem...",
"Solution_6": "any ideas on how to solve this?"
}
{
"Tag": [
"geometry",
"geometric transformation",
"rotation",
"trigonometry",
"integration"
],
"Problem": "A non-conducting ring of mass m and radius r is lying at rest in the vertical XY plane on a smooth horizontal XZ plane. Charge +q and -q are uniformly distributed on the ring on the two sides of the vertical diameter of the ring. A constant and uniform electric field E is set up along the x direction. The ring is given a small rotation about the vertical diameter of the ring and released. Find the period of oscillation of the ring.\r\n\r\nno diagram given :noo: \r\n\r\n[hide=\"Ans\"] $ 2\\pi \\sqrt{\\frac{mr}{4qE}}$ [/hide]",
"Solution_1": "Let angle of rotation be $ \\theta$.\r\n\r\nLet $ dq$ be a small positive charge at a height $ h$ from the surface. The torque due to this charge is $ Edq\\sqrt{h^2\\plus{}2hr}\\sin\\theta\\approx Edq\\sqrt{h^2\\plus{}2hr}\\theta$\r\n\r\nNow, $ dq\\equal{}\\frac{\\lambda dh}{\\sqrt{h^2\\plus{}2hr}}$ ($ \\lambda\\equal{}\\frac{Q}{2\\pi}$) and hence, torque is $ E\\lambda\\theta dh$\r\n\r\nThus, total torque due to positive side is $ \\tau_1\\equal{}\\int_0^{2r}E\\lambda\\theta dh\\equal{}2E\\lambda\\theta r$ and total torque on the ring is $ \\tau\\equal{}4E\\lambda\\theta r\\equal{}\\minus{}\\frac{2EQ\\theta r}{\\pi}$\r\n\r\nNow, moment of inertia of the ring is $ \\tfrac{1}{2}mr^2$ and hence period of oscillation is $ T\\equal{}2\\pi\\sqrt{\\frac{\\pi mr}{4EQ}}$",
"Solution_2": "If the origin is taken as the lowest point of the ring, we have the generic elementary torque as\r\n\r\n $ d\\tau = 2\\theta \\sqrt {2yr-y^2} E {Q \\over{\\pi r}} {{r dy} \\over {\\sqrt{2yr - y^2}}}$\r\n\r\nwhich, integrated from 0 to 2r leads to $ \\tau = 4 \\theta E {Q \\over{\\pi }} r$.\r\n\r\nAs the moment of inertia with respect an diameter is $ {{mr^2}\\over 2} \\rightarrow \\omega^2 = {{4 \\pi^2} \\over T^2} = {{8EQ} \\over {\\pi m r}}$,\r\nand $ T = \\pi \\sqrt{{{\\pi mr} \\over {2EQ}}}$."
}
{
"Tag": [],
"Problem": "In the table below, you may switch the signs of all the numbers of a row, column, or parallel to one of the diagonals. In particular, you may switch the sign of each corner square. Prove that you cannot have all the entries as $ 1.$\r\n\\[ \\begin{tabular}{|c|c|c|c|}\\hline\r\n1&1&1&1\\\\\\hline\r\n1&1&1&1\\\\\\hline\r\n1&1&1&1\\\\\\hline\r\n1&\\minus{}1&1&1\\\\\\hline\\end{tabular}.\\]",
"Solution_1": "[hide=\"Solution\"]Consider the eight \"edge squares\". These are the squares around the boundary which are not the corners. Each legal move switches an even number of these squares, so there remains an odd amount of $ \\minus{} 1$'s in the edge squares. Therefore there can never be no $ \\minus{} 1$'s which which means that not all the squares can be $ 1$.[/hide]",
"Solution_2": "[quote=\"KrazyFK\"][hide=\"Solution\"]Consider the eight \"edge squares\". These are the squares around the boundary which are not the corners. Each legal move switches an even number of these squares, so there remains an odd amount of $ \\minus{} 1$'s in the [b]edge squares[/b]. Therefore there can never be no $ \\minus{} 1$'s which which means that not all the squares can be $ 1$.[/hide][/quote]\r\n\r\nCorrected a slight error",
"Solution_3": "Oh yes, thank you for spotting that. :lol:"
}
{
"Tag": [
"geometry",
"trigonometry",
"trig identities",
"Law of Cosines",
"Law of Sines"
],
"Problem": "Here's a neat geometry problem to try.\r\n\r\nAn acute triangle ABC is inscribed in a circle. Two altitudes AM and CN are extended to meet the circle at P and Q respectively. Given that PQ:AC=3:2, find cos :ang: ABC.",
"Solution_1": "[hide]Let BK be the other altitude. Let O be the orthocenter. Let QP=3, and AC=2. AngleABQ=angleQPA=angleQCA=angleABK.\n\nTHerefore QB=BO.\n\nAnglePAC=anglePQC=anglePBC=angleCBK.\n\ntherefore BP=BO\n\nand QB=BP. angleQBP=2 angleABC. Let cosABC be k, QB be x.\n\nLaw of cosine\n\nx^2+x^2-2x^2(2k^2-1)=9\n\n\n\nAlso 2/x=NC/NB=tanABC= :sqrt: (1-k^2)/k\n\n\n\nSolve this very nice system of equation, we have k=3/4 [/hide]",
"Solution_2": "A different method:\n\n\n\n[hide]Let a, b and c represent respectively the degree measures of angles CAB, ABC, and BCA. We know MAC = 90 - c, NCA = 90 - a, so in triangle HAC, angle AHC = 180 - (90 - a) - (90 - c) = 180 - b. Thus, (mAC + mPQ)/2 = 180 - b, where mXY represents the measure of the arc XY. Since mAC = 2b, mPQ = 360 - 4b. Now construct the triangles OPQ and OAC, where O is the center of the circle. The measure of the two central angles are (360 - 4b) and 2b. Applying the law of cosines and dividing, we get [tex]\\displaystyle(\\frac{AC}{PQ})^2=\\frac{1-\\cos{2b}}{1-\\cos{(360-4b)}}[/tex]. Substituting for the double-angle cosine formulae in terms of sin, we get [tex]\\displaystyle{4/9 = \\frac{\\sin^2b}{\\sin^2(2b)}=\\frac{1}{4cos^2b}}[/tex], which, given the acuteness of B, gives us cosb = 3/4.[/hide]",
"Solution_3": "Another different method:\n\n\n\nRefering to the labels in JBL's picture.\n\n\n\n[hide]\n\n[tex]\\angle{BCQ}=90^\\circ-\\angle{B}[/tex] and [tex]\\angle{BCP}=\\angle{BAP}=90^\\circ-\\angle{B}[/tex]. So [tex]\\angle{PCQ}=180^\\circ-2\\angle{B}[/tex]\n\n\n\nBy the Law of Sines:\n\n\n\n[tex]\\displaystyle\\frac{PQ}{AC}=\\displaystyle{\\frac{2R\\sin(180^\\circ-2{B})}{2R\\sin B}= 2\\cos{B}[/tex]\n\n\n\n[/hide]"
}
{
"Tag": [
"Princeton",
"college",
"calculus",
"integration",
"function",
"complex analysis",
"complex analysis theorems"
],
"Problem": "Which books do you recommend in complex analysis.\r\nPlease mention good introductory books and others. \r\n\r\n[hide=\"Some examples, but please mention others (and if possible comment these)\"]\nAnyone read Fundamentals of Complex Analysis with Applications to Engineering, Science, and Mathematics (3rd Edition) ?\nby Edward Saff , Arthur D. Snider\n\nAnyone read Basic Complex Analysis (Hardcover) by Jerrold E. Marsden , Michael J. Hoffman ? \n\n[/hide]",
"Solution_1": "i've heard from my friend that marsden's book is good. i follow complex analysis by elias stein and rami shakarchi ( vol 2 of the princeton lectures in analysis), this book is very good but its meant to give a quick introduction to the most important tools of complex analysis so it might miss some things. the excersises are excellent. another good book is complex analysis by joseph bak and donald newman.",
"Solution_2": "If you want one that is insanely cheap (it is Dover) then Complex Variables by Stephen D. Fisher is quite good. Lots of cookbook problems to get you a good feel for what to do, but also some good standard problems that require thought and teach you things. The reading is pretty straight forward. Any details left out are easy to fill in yourself.\r\n\r\nDown-side: Not the most standard presentation order. Line integrals come before Cauchy-Riemann Eqns. This didn't throw me off though.",
"Solution_3": "Thanks :) .\r\nPlease post more comments!\r\nPerhaps this forum part is not that visited\r\nas other.",
"Solution_4": "[quote=\"color\"]Which books do you recommend in complex analysis.\nPlease mention good introductory books and others. [/quote]\r\nB. V. Sabat (orig. Szabat), Introduction to complex analysis.",
"Solution_5": "The classic Ahlfor's \"Complex Analysis\", another is Stein and Shakarchi's \"Complex Analysis\". Finally, Complex Variables: Harmonic and Analytic Functions by Flanagan.",
"Solution_6": "I agreed for gka and randomizer but complete : Function Theory of Several Complex Variables by Steven G. Krantz :)",
"Solution_7": "[b]You 've definitely got to check out [color=red]Tristan Needham[/color]'s one... \n\nQuite possibly, [color=red]Roger Penrose[/color]'s praise for the book might provide you with further motivation to do so:[/b] \r\n\r\n[b][color=green]\"...By his innovative and exclusive use of the geometrical perspective,[/color] [color=red]Tristan Needham[/color] [color=green]uncovers many surprising and largely unappreciated aspects of the beauty of Complex Analysis.\"[/color][/b]"
}
{
"Tag": [
"logarithms",
"induction"
],
"Problem": "I'm not sure if this is possible, but...\r\n\r\nSolve for $x \\geq 0$: $x= \\frac{1}{x-1} + \\frac{1}{x-2} + ... +1=\\sum_{k=1}^{x-1} (\\frac{1}{x-k})$",
"Solution_1": "[quote=\"cincodemayo5590\"]I'm not sure if this is possible, but...\n\nSolve for $x \\geq 0$: $x= \\frac{1}{x-1} + \\frac{1}{x-2} + ... +1=\\sum_{k=1}^{x-1} (\\frac{1}{x-k})$[/quote]\r\n\r\nnote $x\\in\\mathbb{Z}^+$, and $x>1$ for the summation to be possible\r\n\r\nthis is taking a partial sum of the harmonic series up to $x-1$, that sum being equal to x\r\n\r\nlet $S_n=\\sum_{1}^{n} \\frac{1}{i}$ (this is your summation, just a bit clearer)\r\n\r\nwe want $S_{n-1}=n$\r\n\r\nit is pretty easy to show that $S_{n}-1<\\log_2 n$, so we want $n$ such that\r\n\r\n$n-1<\\log_2(n-1)$\r\n$2^n<2(n-1)$\r\n\r\nwhich is true for no positive integers $>1$...(easily shown by induction)\r\n\r\nthus there are no solutions"
}
{
"Tag": [
"Olimpiada de matematicas"
],
"Problem": "yo tengo una solucion pero usted derrepente tiene otra solucion; compartamos soluciones!!",
"Solution_1": "EN EL PROBLEMA 1.\r\n\r\n\r\n X= medida del angulo en C",
"Solution_2": "Estos dos problemas son variaciones de este:\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=84696"
}
{
"Tag": [
"function",
"induction",
"algebra proposed",
"algebra"
],
"Problem": "I got it from my brother and i cant do it -*-\r\n\r\nFind all function f:N->N such that\r\n1.f(1)=1\r\n2.f(2n)=2f(n)+1 for all x in N\r\n3.f(f(n))=4n-3 for all x in N\r\n....\r\n\r\nI dont have the way to overcome that f(n)=2n-1 ...",
"Solution_1": "We can prove it by induction!\r\nAssume that $ f(n)\\equal{}2n\\minus{}1$ for all $ 1\\leq n\\leq k$, we have to prove that $ f(k\\plus{}1)\\equal{}2k\\plus{}1$\r\nIn fact:\r\n-) $ k\\plus{}1\\equal{}2m$ then $ f(k\\plus{}1)\\equal{}f(2m)\\equal{}2f(m)\\plus{}1\\equal{}2(2m\\minus{}1)\\plus{}1\\equal{}4m\\minus{}1$ (Because $ m\\leq k$)\r\n-) $ k\\plus{}1\\equal{}2m\\minus{}1$ then $ f(k\\plus{}1)\\equal{}f(2m\\minus{}1)\\equal{}f(f(m))\\equal{}4m\\minus{}3$ (Because $ m\\leq k$)\r\nThen OK!"
}
{
"Tag": [
"logarithms",
"LaTeX"
],
"Problem": "Evaluate the product:\r\n\r\n$(log_{2}{3})(log_{3}{4})(log_{4}{5})(log_{5}{6})$",
"Solution_1": "[hide=\"My sorution\"]\nUse change of base formula to narrow down the equation to \\[{\\frac{\\text{log}6}{\\text{log}2}}\\]\nThen simplifying even further we get\n\\[{{\\frac{\\text{log}6}{\\text{log}2}}=\\frac{\\text{log}2+\\text{log}3}{\\text{log}2}}=1+\\frac{\\text{log}3}{\\text{log}2}=\\boxed{1+\\text{log}_{2}3}\\]\nWell that's the most I can simplify it without a calculator :wink: \n\n[/hide]",
"Solution_2": "[quote=\"math92\"]Evaluate the product:\n\n$(log_{2}{3})(log_{3}{4})(log_{4}{5})(log_{5}{6})$[/quote]\r\n\r\n*sigh* It looks like you took this from AoPS Volume 2 -_-\r\n\r\n[hide]The chain collapses to $log_{2}{6}$\n\nThis can be proven, and it is proven in AoPS Volume 2[/hide]",
"Solution_3": "[quote=\"math92\"]Evaluate the product:\n\n$(log_{2}{3})(log_{3}{4})(log_{4}{5})(log_{5}{6})$[/quote]\r\n[hide=\"solution with lazy jim log property\"]The lazy jim is a generalization to the chain rule. It says that $(\\log_{a}b)(\\log_{c}d=(\\log_{a}d)(\\log_{c}b)$\nApplying it this simplifies to $\\log_{2}6$[/hide]",
"Solution_4": "drunkenmath and bpms u guys are wrong neveroddoreven simlified it more :D",
"Solution_5": "It's hard to measure \"simplified\" and I'd definitely not say they're wrong :)\r\n\r\n[hide=\"hint\"]Try $2^{\\text{Whole thing..}}$[/hide]",
"Solution_6": "[hide]$(\\log_{2}3)(\\log_{3}4)(\\log_{4}5)(\\log_{5}6)$\n$=\\frac{\\log 3}{\\log2}\\frac{\\log 4}{\\log 3}\\frac{\\log 5}{\\log 4}\\frac{\\log 6}{\\log 5}$\n$=\\frac{\\log 6}{\\log 2}$\n$=\\frac{\\log 2+\\log3}{\\log 2}$\n$=1+\\frac{\\log3}{\\log 2}=1+\\log_{2}3$[/hide]",
"Solution_7": "[hide=\"solution\"]We switch the 3 and the 4 in the first two terms, and then switch the 5 and the 6 in the third and fourth terms and get thus:\n\n$2\\log_{4}{6}=\\log_{4}{36}=\\log_{2}{6}=1+\\log_{2}{3}$\n\nIt is impossible to simplify any more.[/hide]",
"Solution_8": "[hide][quote=\"math92\"]Evaluate the product:\n\n$(log_{2}{3})(log_{3}{4})(log_{4}{5})(log_{5}{6})$[/quote]\n\nthat equals log3/log2*log4/log 3*log5/log4*log6/log5 using change of base\ncancel out log3, log4, and log5\nand you get log6/log2\nlog 6= log2+log3\nso you get 1+log3/log2\nsimplifies into 1+log base 2 3\n(ok I fail in latex)[/hide]"
}
{
"Tag": [
"integration",
"LaTeX",
"algebra proposed",
"algebra"
],
"Problem": "If for each $x\\geq 0$ holds that $ax^2+bx+c\\geq 0$ and $a,b,c\\in R$ prove that \r\n\r\n$\\frac{a}{2008}+\\frac{b}{2007}+\\frac{c}{2006}\\geq 0$",
"Solution_1": "We are to lead $\\int_0^1 (ax^{2007}+bx^{2006}+cx^{2005})\\ dx\\geq 0.$",
"Solution_2": "This is also my solution but i am trying to find a solution without using integrals. Is it possible ?Please help",
"Solution_3": "What do you mean by saying \"and $a,b,c$\"?",
"Solution_4": "It was a latex problem. Now i fix it. Try to prove it",
"Solution_5": "[quote=\"silouan\"]This is also my solution but i am trying to find a solution without using integrals. Is it possible ?Please help[/quote]\r\nOf cours that is possible :lol:.\r\nWhen $x=0$=>$c\\geq 0$\r\nIf $a<0$=>the ineq. $ax^2+bx+c\\geq 0$ have only finity solution(contradiction)=>\r\n$a\\geq 0$.If $b>0$ the problem is done.\r\nLet $b=-d$ wher $d>0$.\r\nIf $D=b^2-4ac>0$=> $ax^2+bx+c\\geq 0$ only if $x\\leq \\frac{1}{2a}(d-\\sqrt{d^2-4ac})$\r\nor $x\\geq \\frac{1}{2a}(d+\\sqrt{d^2-4ac})$=>$\\frac{1}{2a}(d+\\sqrt{d^2-4ac})\\leq 0$(contradiction again)=>$D=b^2-4ac<0$.\r\n$\\frac{a}{2008}+\\frac{b}{2007}+\\frac{c}{2006}\\geq 0$<=>\r\n$(\\frac{a}{2008}+\\frac{c}{2006})^2\\geq(\\frac{d}{2007})^2$but \r\n$(\\frac{d}{2007})^2\\leq \\frac{4ac}{2007^2}\\leq 4\\frac{a}{2006}\\frac{c}{2008}\\leq (\\frac{a}{2008}+\\frac{c}{2006})^2$ ;)",
"Solution_6": "Thank you very much tiks nice solution .",
"Solution_7": "Thank you too Solouan for interesting problem ;)."
}
{
"Tag": [
"number theory",
"greatest common divisor"
],
"Problem": "6. Let $a, b, c, d$ be positive integers such that $ab = cd$. Prove that $a + b + c + d$ is not prime.\r\n\r\n8. Consider the sequence\r\n$x_{1} = 19$, $x_{2} = 95$, $x_{n+2} = lcm(x_{n+1}, x_{n}) + x_{n}$\r\nfor $n > 1$. Find the greatest common divisor of $x_{1995}$ and $x_{1996}$.",
"Solution_1": "[quote=\"Andreas\"]6. Let $a, b, c, d$ be positive integers such that $ab = cd$. Prove that $a + b + c + d$ is not prime.\n[/quote]\r\nPosted quite a lot of time at ML.. :) :\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=38348 for example",
"Solution_2": "I think this is $19$, furthermore this is $19$ in every cases !\r\n\r\nFirst $19 \\mid x_n$ \r\n\r\nIf $d \\mid x_n$ and $d \\mid x_{n+1}$ then $d \\mid x_{n-1}$ so $d \\mid gcd (x_1,x_2) = 19$"
}
{
"Tag": [
"logarithms",
"inequalities",
"algebra unsolved",
"algebra"
],
"Problem": "Find x $ \\in R$ of: \r\n\r\n$ 5x \\plus{} \\sqrt {6x^2 \\plus{} x^3 \\minus{} x^4}.\\log_{2}{x} > (x^2 \\minus{} x).\\log_{2}{x} \\plus{} 5 \\plus{} 5\\sqrt {6 \\plus{} x \\minus{} x^2}$",
"Solution_1": "We notice $ x\\in(0,3]$. The inequality can be rewritten as\r\n$ (x\\log_2^x\\minus{}5)(\\sqrt{6\\plus{}x\\minus{}x^2}\\minus{}(x\\minus{}1))>0$.\r\nSince $ 3\\log_2^3<5$, we have only\r\n$ \\sqrt{6\\plus{}x\\minus{}x^2} n$\nii)$ \\forall u,v \\in B,u \\plus{} v \\not\\in B$\n Proof:\nWe call a set S \"good\" if it satisfies $ \\forall u,v \\in S,u \\plus{} v \\not\\in S$\nWe use induction to solve this porblem.\nFor the base case n=1,it is easy to see.\nAssuming that it is true for $ n$,we will prove it is true for $ n \\plus{} 1$.\nLet A be a set such that $ |A| \\equal{} 3n \\plus{} 3$.Let $ A \\equal{} T \\cup \\{a_1,a_2,a_3\\}$ where $ a_1$ be the smallest number among the elements in set A and $ |T| \\equal{} 3n$.\nWe see that $ |T| \\equal{} 3n$.From the assumption above,we can choose a \"good\" subset $ B_1$ satisfying$ |B_1| > n$.Let $ B \\equal{} B_1 \\cup \\{a_1\\}$,we'll have a good subset $ B$ satisfying $ |B| > n \\plus{} 1$,$ Q.E.D$[/hide]",
"Solution_4": "Hmm...Why do you think that $ B$ is a good subset? What it has to do with that $ a_1$ is the smallest element.",
"Solution_5": "[quote=\"iura\"]For a set $ S$, let $ |S|$ denote the number of elements in $ S$. Let $ A$ be\na set of positive integers with $ |A| = 2001$. Prove that there exists a\nset $ B$ such that\n\n(i) $ B \\subseteq A$;\n(ii) $ \\dsp |B| \\ge 668$;\n(iii) for any $ u, v \\in B$ (not necessarily distinct), $ u + v \\not\\in B$.[/quote]\r\nI saw this problem in a chinese book, but with $ 2001$ changed to $ n$ and $ \\ge 668$ changed to $ > \\frac{n}{3}$ :D \r\nHere is the solution, given by a Israel mathematician N.Alon.\r\n\r\nLet the elements in $ A$ be $ a_1,a_2,...,a_n$, and choose a prime of the form $ 3k-1$ and is greater than all the elements in $ A$.\r\nThen consider $ a_i,2a_i,...,pa_i$ where $ i$ is an integer such that $ 1\\le i\\le n$. They form a complete set of residues of $ p$. Therefore there are $ k$ of them congruent to $ k,k+1,k+2,...2k-1$ modulo $ p$.\r\n\r\nLet $ x_j$ be the number of numbers among $ ja_1,ja_2,...,ja_n$ that congruent to $ k,k+1,k+2,...2k-1$ modulo $ p$. Now $ x_1+x_2+x_3+...+x_p=kn$\r\nSo there exists a $ x_t$ such that $ x_t\\ge\\frac{kn}{p}=\\frac{kn}{3k-1}\\ge\\frac{n}{3}$\r\n\r\nThen $ B={a\\in A}$ such that $ ta$ congruent to $ k,k+1,k+2,...2k-1$ modulo ${ p}$ :)",
"Solution_6": "awesome solution",
"Solution_7": "Let $ A \\equal{} \\{a_1,a_2,...,a_{2001}\\}$.Let $ p$ be a prime number such that $ p \\equiv 2 (mod 3)$ and $ p > a_i \\forall i \\equal{} 1,2,...,2001$.\r\nSuch a prime p exists by Dirichlet's Theorem,althought the result can also be easily proven directly.There is at least 1 prime congruent to 2 modulo 3(namely,2).Suppose there were only finitely many primes congruent to 2 modulo 3,and let their product be $ P$.Then $ 3P \\minus{} 1$,which is larger than $ P$ and congruent to 2 modulo 3,must have another prime divisor congruent to 2 modulo 3,contradiction.Thus,the orginal assumption was wrong,and there are infinitely many odd primes that are congruent to 2 modulo 3.Specifically,one such prime is larger than all $ a_i$.\r\nAll elements of $ S$ are distinct and non-zero modulo $ p$.\r\nCall a number $ n$ \"mediocre\" is the least positive residue of $ n$ mod $ p$ lies in $ [ \\frac {p \\plus{} 1}{3},\\frac {2p \\minus{} 1}{3}]$.For any $ 1$\u2264$ i$\u2264$ 2001$,there are exactly $ \\frac {p \\plus{} 1}{3}$ integer values of $ k \\in [1,p \\minus{} 1]$ such that $ ka_i$ is mediocre.Thus,there are $ \\frac {2001(p \\plus{} 1)}{3} \\equal{} 667(p \\plus{} 1)$ pairs of $ (k,i)$ such that $ ka_i$ is mediocre.\r\nBy Pigeonhole Principle,there exists some $ k$ for which the set $ B \\equal{} \\{a_i|ka_i$ is medicre$ \\}$ has it least $ 668$ elements.\r\nWe now claim that this $ B$ satisfies the desired properties.It suffices to show that $ k$ times the sum of any 2 elements of $ B$ cannot be mediocre and hence connot equal $ k$ times any elements of $ B$.To that end,note that $ k$ times the sum of any 2 elements of $ B$ cannot be medicre because it is congruent mod $ p$ to some number in $ [\\frac{2(p\\plus{}1)}{3},\\frac{2(2p\\minus{}1)}{3}]$ or, eqivalently,to some number in $ [0,\\frac{(p\\minus{}1)}{3}] \\cup [\\frac{(2p\\plus{}2)}{3},p\\minus{}1]$,which is a set containing no mediocre numbers.Thus,the set $ B$ satisfies the desired properties.",
"Solution_8": "It is a special case of a theorem due to Paul.Erdos about sum-free set ( not have elements $u$ and $v$ such that $u+v$ belong also to it.)\nevery set of positive integers A always have a sub set S with $|S|>|A|/3$ which is sum free.you can see it in many combinatoric book.\nThe same as VMO 2006 pro 5 :rotfl: . :oops:\nand probabilistic method will work properly.",
"Solution_9": "This solution seems different from everyone else's (though it uses the same $(\\tfrac{1}{3},\\tfrac{2}{3})$ interval modulo $1$ idea).\n\nLet $n=2001$, and note that $668=\\tfrac{n}{3}+1$. For a positive real number $r$, define $S_r=\\{rx \\colon x \\in \\mathbb{R}_{>0},\\tfrac{1}{3}\\frac{n}{3}\\left(\\frac{m}{Nm+\\frac{1}{3}}-\\frac{m}{(N+1)m-\\frac{1}{3}}\\right),\\]\nwhich shows that there exists a real number $r \\in [\\tfrac{m}{(N+1)m-\\frac{1}{3}},\\tfrac{m}{Nm+\\frac{1}{3}}]$ such that if $B=S_r \\cap A$, then $|B|>\\tfrac{n}{3}$.\n\nIt suffices to prove that\n\\[\\int_\\frac{m}{(N+1)m-\\frac{1}{3}}^\\frac{m}{Nm+\\frac{1}{3}} |S_r \\cap \\{k\\}| \\,\\mathrm{d}r>\\frac{1}{3}\\left(\\frac{m}{Nm+\\frac{1}{3}}-\\frac{m}{(N+1)m-\\frac{1}{3}}\\right)\\]\nfor all $k \\in A$. But we have\n\\begin{align*}\n\\int_\\frac{m}{(N+1)m-\\frac{1}{3}}^\\frac{m}{Nm+\\frac{1}{3}} |S_r \\cap \\{k\\}| \\,\\mathrm{d}r&=\\frac{k}{Nk+\\frac{1}{3}}-\\frac{k}{Nk+\\frac{2}{3}}+\\frac{k}{Nk+\\frac{4}{3}}-\\frac{k}{Nk+\\frac{5}{3}}+\\cdots+\\frac{k}{(N+1)k-\\frac{2}{3}}-\\frac{k}{(N+1)k-\\frac{1}{3}} \\\\\n&\\ge \\frac{1}{3}\\left(\\frac{k}{Nk+\\frac{1}{3}}-\\frac{k}{Nk+\\frac{2}{3}}+\\frac{k}{Nk+\\frac{2}{3}}-\\frac{k}{Nk+\\frac{3}{3}}+\\cdots+\\frac{k}{(N+1)k}-\\frac{k}{(N+1)k+\\frac{1}{3}}\\right) \\\\\n&=\\frac{1}{3}\\left(\\frac{k}{Nk+\\frac{1}{3}}-\\frac{k}{(N+1)k+\\frac{1}{3}}\\right).\n\\end{align*}\nThus, it suffices to prove that\n\\begin{align*}\n\\frac{k}{Nk+\\frac{1}{3}}-\\frac{k}{(N+1)k+\\frac{1}{3}}&>\\frac{m}{Nm+\\frac{1}{3}}-\\frac{m}{(N+1)m-\\frac{1}{3}} \\\\\n\\Longleftrightarrow \\frac{k^2}{(Nk+\\frac{1}{3})((N+1)k+\\frac{1}{3})}&>\\frac{m^2-\\frac{2}{3}m}{(Nm+\\frac{1}{3})((N+1)m-\\frac{1}{3})}.\n\\end{align*}\nIndeed, notice that the left-hand side tends to $\\tfrac{1}{N(N+1)}$ while the right-hand side tends to $\\tfrac{1-\\frac{2}{3m}}{N(N+1)}$ as $N$ goes to infinity, as desired. $\\square$\n\n[b][color=#9a00ff]Remark:[/color][/b] This solution can be motivated, despite having a bunch of random constructions that appear seemingly out of nowhere. At first, I tried to make an argument using $\\nu_2$: if there are more than $\\tfrac{n}{3}$ odd numbers in $A$, choose those; otherwise, look at the subset of even numbers and try to choose something. This argument isn't strong enough, but it uses a prototypical version of $S_r$: the set $\\{rx \\colon x \\in \\mathbb{N},x \\equiv 1 \\pmod{2}\\}$ where $r$ is a power of $2$. Eventually, I realized two things: we could set $r$ to be any real number, and we could relax $x \\equiv 1 \\pmod{2}$ to $x \\approx 1 \\pmod{2}$. This gave me the version of $S_r$, the set of all positive real numbers less than $\\tfrac{1}{3}r$ away from the nearest odd multiple of $r$, which I used to solve the problem. I later scaled the definition of $S_r$ by $2$ for writeup purposes.\n\nThe biggest conceptual jump in this problem for me was seriously thinking about using small $r$, as you might not expect small $r$ to give anything useful, and indeed, most of the solutions in this thread use a large prime $p$, which can be thought of as a large $r$ except you can transform $A$ modulo $p$ with multiplication. My first try was with $r \\in [\\tfrac{1}{N+1},\\tfrac{1}{N}]$ (using the definition of $S_r$ used in this writeup), but this wasn't quite enough because it gave a density slightly less than $\\tfrac{1}{3}$. After flailing with different constructions, I realized that you could cut off very small bits of the interval $[\\tfrac{1}{N+1},\\tfrac{1}{N}]$ to create $[\\tfrac{m}{(N+1)m-\\frac{1}{3}},\\tfrac{m}{Nm+\\frac{1}{3}}]$ while not changing the sum. I was reluctant to try this because I thought it would almost certainly be insufficient, but it turned out to be enough."
}
{
"Tag": [],
"Problem": "In cate feluri putem acoperi un dreptunhi 1Xn cu k dominouri.\r\nObs: un domino reprezinta un dreptunghi 1X2.",
"Solution_1": "[hide=\"parere\"]\nhmm...daca nu ma insel eu, raspunsul este $\\binom{n-k}{k}$ (evident problema nu are sens pentru $2k>n$).\nconsider patratelele dreptunghilui ca fiind un sir de $k$ $1$-uri si $n-k$ $0$-uri, unde $1$ reprezinta patratelul din stanga al unui domino.\natunci, din ipoteza, nu exista doi de $1$ alaturati. acum, din sir, tai cele $k$ zerouri situate imediat in dreapta unui $1$, si se obtine un sir de $n-k$ cifre, dintre care $k$ sunt $1$.\nreciproc, dintr`un sir de $n-k$ cifre, din care $k$ sunt $1$, adaug cate un $0$ dupa fiecare $1$ si obtin un sir \"bun\" de $n$ cifre.\nacum, dintr`un sir de $n-k$ cifre unde am $1$ pe pozitiile $x_{1},x_{2},\\ldots,x_{k}$, sirul de $n$ cifre obtinut va avea $1$ pe pozitiile $x_{1},x_{2}+1,x_{3}+2,\\ldots,x_{k}+k-1$. \ndeci corespondenta obtinuta este bijectiva, asadar numarul de siruri \"bune\" de lungime $n$ este chiar numarul sirurilor cu $k$ de $1$ de lungime $n-k$, si de aici vine raspunsul.\n[/hide]",
"Solution_2": "La solutia ta nu vad nimic sa fie gresit...insa pe un caz particular nu imi da ....de ex : fie n=8 si k=3...dupa cum spui tu ar trebui sa dea 10....fie a,b,c cele 3 dominouri...fixam primele 2 dominouri( adica pe pozitiile 1 si 2 asezam 2 dominouri a)...pentru pozitia 3 avem 3 modalitati de asezare (a,b sau c),dar pentru fiecare dintre acestea mai avem 3 modalitati in care putem sa asezam dominourile pentru pozitia 4....deci 9 asezari...pana acum ( am considerat fixe doar primele doua) ...",
"Solution_3": "hm..uite\r\n\r\n$\\begin{tabular}{|c|c|c|c|c|c|c|c|c|}\\hline & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 \\\\\\hline 1 & A & A & B & B & C & C & & \\\\\\hline 2 & A & A & B & B & & C & C & \\\\\\hline 3 & A & A & B & B & & & C & C \\\\\\hline 4 & A & A & & B & B & C & C & \\\\\\hline 5 & A & A & & B & B & & C & C \\\\\\hline 6 & A & A & & & B & B & C & C \\\\\\hline 7 & & A & A & B & B & C & C & \\\\\\hline 8 & & A & A & B & B & & C & C \\\\\\hline 9 & & A & A & & B & B & C & C \\\\\\hline 10& & & A & A & B & B & C & C \\\\\\hline \\end{tabular}$"
}
{
"Tag": [
"analytic geometry",
"graphing lines",
"slope"
],
"Problem": "i was wondering if anyone could help me out with some problems on messenger i am having troubles with? mainly slope problems, linear equations, and inequalities.\r\n\r\ni have all messengers but ICQ, AIM =eazee ryd3r(or iamalexwayne)\r\nMSN=amentopringles@hotmail.com\r\nyahoo=fuzzy000wuzzy00\r\n\r\nthanks :)",
"Solution_1": "Welcome!\r\n\r\nI moved your post from the Olympiad section because it doesn't sound that you need help with olmypiad stuff. I would recommend taking a look on the [url=http://www.mathlinks.ro/Classes/AoPS_C_About.php]AoPS Classes curriculum[/url] and see if any of the intro classes help you with this. \r\n\r\nOr try to mingle in the forums, by posting problems on the forum (read the descriptions first) and trying to solve problems from your level.",
"Solution_2": "[quote=\"Valentin Vornicu\"]Welcome!\n\nI moved your post from the Olympiad section because it doesn't sound that you need help with olmypiad stuff. I would recommend taking a look on the [url=http://www.mathlinks.ro/Classes/AoPS_C_About.php]AoPS Classes curriculum[/url] and see if any of the intro classes help you with this. \n\nOr try to mingle in the forums, by posting problems on the forum (read the descriptions first) and trying to solve problems from your level.[/quote]\r\n\r\ni'm sorry, i'm new here, i don't even know what olympiad means, i was just searching math forum on google in hopes someone who enjoys or is knowledgeable about algebra would help me out. truthfully i just want someone to work some problems out for me.. and show me how they did some parts i didnt understand, since i cant do that on exams with the teacher of my homeschool course. i didnt want to post the exact problems on the board itself, i was hoping for private, direct communication via messenger, so was just advertising for a helper basically.\r\n\r\nits algrebra 2, but this lesson is basically a review of algebra one, i imagine its very easy for some of the people on here, and would just like some help to speed the class up.",
"Solution_3": "You are encouraged to post your problems in the appropriate forums: either [url=http://www.mathlinks.ro/Forum/index.php?f=298]middle school forum[/url] if you're a middle schooler (or equivalent of it) or [url=http://www.mathlinks.ro/Forum/index.php?f=214]high school forum[/url]. \r\n\r\nFor private lessons I think you can check out [url=http://www.mathlinks.ro/Forum/viewtopic.php?t=23348]Richard Rusczyk's offer[/url] or any of the [url=http://www.mathlinks.ro/Classes/AoPS_C_About.php]AoPS classes[/url] I mentioned.",
"Solution_4": "i'm gonna post an example here, and maybe you can help or redirect me to where i would best be helped? not looking for private lessons really, just someone in private direct communication(via instant message) to help me with who knows this stuff algrebra/algrebra 2 stuff. \r\n\r\nwat kind of system uv equationz is 2x-3y=5 and -2x+5y=-1\r\na. k(c)onsistent and e(i)ndependent\r\nb. Konsistent and Dependent\r\nc. InKonsistent \r\nd. Inkonsistent and depindent\r\n\r\nsorry for the intentional typos, wasdone for seurity purposes, still readable.",
"Solution_5": ":?: Security purposes :?: \r\n\r\nPlease explain\r\n\r\nI think that you should find a tutor near you. None of the Art of Problem Solving classes would help you with this. \r\n\r\nI would recommend using the Middle School or Getting Started forums to ask for help ( I know you said that you don't want to use the boards but I still recommend it...). As for someone working problems out for you do you not have a textbook? Every textbook works out problems. There are probably sites online that solve problems.",
"Solution_6": "Are you trying to get someone to help you do a take-home test?",
"Solution_7": "[quote=\"tokenadult\"]Are you trying to get someone to help you do a take-home test?[/quote]\r\npretty much. :blush: \r\ni've already did a good % of it, i'd just like some people to confirm the answers i got and help with the ones i couldnt get.",
"Solution_8": "so what forum would be appropiate for \r\n\r\nwat kind of system uv equationz is 2x-3y=5 and -2x+5y=-1 \r\na. k(c)onsistent and e(i)ndependent \r\nb. Konsistent and Dependent \r\nc. InKonsistent \r\nd. Inkonsistent and depindent \r\n\r\nand\r\n\r\nwat is the slope intercept form of the equation of a line that goes through (1, 4) and is perepndicular to a line whose equation is y=2/3x+5 ?\r\n\r\na. y=2/3x+5\r\nb. y=-3/2x+7\r\nc. y=-3/2x+11/2\r\nd.y=-3/2x+5/2\r\n\r\nthose 3/2's are fractions just in case ya didnt know.\r\n\r\nalso another one i had trouble with:\r\nwhat is the standard form of an equation for a line that passes through (-1, -2) and has slope -3?\r\n\r\na. 3x+y=-5\r\nb. 3x+y=-1\r\nc. 3x+y=-1\r\nc.3x+y=5\r\nd. 3x+y=-7\r\n\r\nand 2 more but ill have to make a picture of graphs to show the problem. surely someone can help me with these, i bet these are cake to some of you on this site.",
"Solution_9": "Come on...we're not doing your homework/take home tests!!!\r\n\r\nAnd plus, I already told you which forums to use."
}
{
"Tag": [
"number theory unsolved",
"number theory"
],
"Problem": "Let p; q and r be prime numbers. It is given that p divides qr - 1, q divides rp - 1, and r divides pq - 1.\r\nDetermine all possible values of pqr.",
"Solution_1": "We will use the following fact: $a \\mid x, b \\mid y \\Rightarrow ab \\mid xy$.\r\n\r\nSince $p \\mid qr-1, q \\mid pr-1, r \\mid pq-1$ we have $pqr \\mid (pq-1)(qr-1)(pr-1)$. But now expanding and substracting multiples of $pqr$ we get $pqr \\mid pq+qr+pr-1$. Suppose WLOG $p \\leq q \\leq r$. Then $pqr \\leq pq+qr+pr-1 < 3qr$, so $p<3$. Then $p=2$. Now we have $2 \\mid qr-1$ so $qr$ is odd, which means $21) a(0)=-1,a(1)=1\r\nfind a(n)\r\nthanks!! :lol:",
"Solution_1": "[quote=\"admire9898\"]the problem is:\n\na(n)=2a(n-1)+3a(n-2)+3^n\n(n>1) a(0)=-1,a(1)=1\nfind a(n)\nthanks!! :lol:[/quote]\r\n\r\nLet $ a_n \\equal{} b_n \\plus{} \\frac {3^{n \\plus{} 1}}{4}n$. The sequence becomes $ b_n \\equal{} 2b_{n \\minus{} 1} \\plus{} 3b_{n \\minus{} 2}$ whose characteristic equation is $ x^2\\minus{}2x\\minus{}3\\equal{}0$, or also $ (x\\minus{}3)(x\\plus{}1)\\equal{}0$\r\n\r\nSo $ b_n \\equal{} \\alpha 3^n \\plus{} \\beta( \\minus{} 1)^n$\r\n\r\nStarting conditions $ b_0 \\equal{} \\minus{} 1$ and $ b_1 \\equal{} \\minus{} \\frac 54$ imply $ \\alpha \\equal{} \\minus{} \\frac {9}{16}$ and $ \\beta \\equal{} \\minus{} \\frac {7}{16}$\r\n\r\nAnd so $ a_n \\equal{} \\minus{} \\frac {9}{16}3^n \\minus{} \\frac {7}{16}( \\minus{} 1)^n \\plus{} \\frac {3^{n \\plus{} 1}}{4}n$"
}
{
"Tag": [
"group theory",
"abstract algebra",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Let $ G$ be a group and $ H$ be a subgroup of $ G$. Suppose that $ aH \\equal{} bH$ if and only if $ Ha \\equal{} Hb$, where $ a, b\\in G$. Show that $ H$ is a normal subgroup of $ G$.",
"Solution_1": "$ aH\\equal{}bH$ $ \\Leftrightarrow$ $ Ha\\equal{}Hb$ is equivalent to $ b^{\\minus{}1}a \\in H$ $ \\Leftrightarrow$ $ ab^{\\minus{}1} \\in H$.\r\n\r\n$ \\forall h \\in H$ and $ \\forall a \\in G$ there is $ b \\in G$ such that $ ab^{\\minus{}1}\\equal{}h$. \r\n$ ab^{\\minus{}1} \\in H$ $ \\Leftrightarrow$ $ b^{\\minus{}1}a \\equal{}a^{\\minus{}1}(ab^{\\minus{}1})a\\equal{}a^{\\minus{}1}ha \\in H$ , therefore $ H \\triangleleft G$ ."
}
{
"Tag": [
"geometry"
],
"Problem": "$\\angle BPT= \\angle TPD$\r\n\r\nProve that $B,Z,L$ are collinear.",
"Solution_1": "[b]Lemma[/b]. [color=blue]agar shekle $(1)$ ra dashte bahim .angah:$\\angle BAD=\\angle DAC$[/color](ba enekas be markaze $A$ mitooonid sabe konid)\r\n\r\n\r\n\r\n[color=red](rahehal shekle aghaye madani)[/color]\r\nhal ba borhan kholf mitavanid sabet konid dayere mohitye mosalase$ZTQ$ bar khatol markazein va do dayere momas ast.\r\n\r\nbaraye inke sabet konim $B,Z,L$ hamkhatand.bayad sabet konim ke cheharzeliye $BZQL$ mohati ke ba enekas be markaze $P$ hal mishe.\r\n\r\nmibakhsid ke hal nadashtam ziad toozih bedam.",
"Solution_2": "baayad saabet konim chehar zelie $BZQD$ mohaatie. pas baayad saabet konim ke agar $S$ markaze tajaanose daakhelie do daayer baashad; aangaah $Q,Z,S$ ham khatand.\r\ndar mosalase $\\triangle NTM$baraaye noghaate $Q,Z,S$ menelaaoos minevisim; tavajoh konid:\r\n$\\frac{MS}{SN}=\\frac{R}{r}$\r\n$TZ*TM=TO^2-R^2$\r\n$TQ*TN=-TO'^2+r^2$\r\n$MZ*MT=2R^2$\r\n$NQ*NT=2r^2$\r\n\r\nage dorost mohaasebe konid; hamechi montaghel mishe be khatol markazeyn; va kaafie tool haaye rooye khatol markazeyn ro mohaasebe konid ;)"
}
{
"Tag": [
"group theory",
"abstract algebra",
"superior algebra",
"superior algebra solved"
],
"Problem": "Since this one seems really easy (and since sometimes I do some pretty obvious mistakes :(), I just want to make sure I got it right:\r\n\r\n$G$ is a group generated by two elements of order $2$. Show that $G$ contains a subgroup of index $2$.",
"Solution_1": "Is $G$ finite?",
"Solution_2": "I don't see why it needs to be finite- index 2 makes perfectly good sense for infinite groups.",
"Solution_3": "I think something like this should work (regardless of the order of the group $G$):\r\n\r\nLet $H$ be the subset of $G$ (obviously a subgroup) consisting of those words in $a,b$ (the mentioned generators) with an even number of letters. A word having an even number of letters is equivalent to its reduced form, with alternating $a$'s and $b$'s having an even number of leters. \r\n\r\nWhat we should prove now is that no two words, one in $H$ and one not in $H$, can ever be equal. It's easy to see that this means proving that no reduced word with an odd number of letters, like $abab\\ldots ba$ can be equal to the identity. This is easy: assuming WLOG that $abab\\ldots ba=1$, multiply this relation both to the left and to the right with $a$ to get $bab\\ldots ab=1$, then multiply this both to the left and right with $b$ to get an even smaller word on the left side and so on, until we get either $a=1$ or $b=1$, which contradicts the fact that the orders of $a,b$ are both $2$.",
"Solution_4": "I specialize in short sentences that make people doubt their own \"proofs\" :D",
"Solution_5": "Thanks, I got the same Grobber ( equivalently the subgroup $$ generated by the product $ab$, which will also contain $(ab)^{-1}=ba$ )."
}
{
"Tag": [
"symmetry"
],
"Problem": "Source: Ireland 1993\r\n1. The real numbers $\\alpha$ and $\\beta$ satisfy the equations \r\n$\\alpha^{3}-3\\alpha^{2}+5\\alpha-17=0$,\r\n$\\beta^{3}-3\\beta^{2}+5\\beta+11=0$.\r\nCompute $\\alpha+\\beta$.",
"Solution_1": "any cubic of the form $x^{3}-3x^{2}+5x+c$, where $c$ is constant, is symmetric about its inflection point, which is at $x=1$. Pick $c=-3$ (chosen because it's the average of $-17$ and $11$. It has a real root at $x=1$, and so the sum of the two values we want is $1+1=2$.",
"Solution_2": "is this the only way to do it? sorry i dont understand your solution. :blush:",
"Solution_3": "i dont know that it's the only solution, but it's the only one i found. The idea to use cubic equation's symmetry about the inflection point is pretty useful though.",
"Solution_4": "You can do it by factoring.\r\n\r\nSee, $x^{3}-3x^{2}+5x = (x-1)^{3}+2x+1$. So, adding the two equalities we get $(\\alpha-1)^{3}+2\\alpha+1+(\\beta-1)^{3}+2\\beta+1 = 17-11 = 6 \\iff (\\alpha-1)^{3}+(\\beta-1)^{3}+2(\\alpha+\\beta-2) = 0$. Factoring $(\\alpha-1)^{3}+(\\beta-1)^{3}$ we obtain $\\alpha+\\beta-2$ as a factor, so the equation factors and we're close home."
}
{
"Tag": [
"AMC",
"AIME"
],
"Problem": "A sequence of real numbers x_n is defined recursively as follows: x_0, x_1 are arbitrary positive real numbers, and \r\nx_(n+2)=[1+x_(n+1)]/(x_n).\r\n\r\nFind x_1998.",
"Solution_1": "Nice problem. Where did it come from? It looks so familiar...",
"Solution_2": "Do you mean find x_1998 in terms of x_0 and x_1, or is there a numerical value no matter what numbers x_0 and x_1 are?",
"Solution_3": "Im pretty sure its in terms of x_0 and x_1.",
"Solution_4": "It's in terms of x_1 and x_0.",
"Solution_5": "[color=cyan]Definitely in terms of x_0 and x_1. As for the other bit, it might be an old AIME question. I don't really know, though. Sorry.[/color]",
"Solution_6": "#12 on AIME 2002 was similar but not identical. If you have a copy of the AIME 2002, please post #12 here. I forget what it says exactly.",
"Solution_7": "Let me see what i can do . . .",
"Solution_8": "Is it #12 at http://www.kalva.demon.co.uk/aime/aime02.html ?",
"Solution_9": "Yeah, that will do. It's in different words, but the problem is the same.",
"Solution_10": "So how about we do this problem??\r\nAny takers??\r\n\r\nJoel, Simon, u guys wait awhile...",
"Solution_11": "I shall stand aside as well...not because Im as smart as Simon and Joel, but because I want to take the 2002 AIME as a real test.",
"Solution_12": "but this question isn't the same one on the AIME",
"Solution_13": "Oh, you were refering to your original question?",
"Solution_14": "yes",
"Solution_15": "Well, at least Ill have the solutions to the questions after I finish the AIME. 8-)",
"Solution_16": "I solved this problem on a piece of paper. I'm waiting for others to solve it. I'm already too old to take the AIME again.",
"Solution_17": "C'mon\r\nsomeones gotta answer dis\r\n\r\nno fun if simon or I post the solution...",
"Solution_18": "Well, you could probably calculate x_2, x_3, x_4, and x_5, and a simple pattern will emerge. I am too lazy to do right now though, I have summer reading to finish. 8-)",
"Solution_19": "yea\r\nsigh",
"Solution_20": "Why don't you try it? Maybe something nice will happen. Maybe it won't. I shall say no more (for now).",
"Solution_21": "yea\r\ndon't be lazy like me\r\nthats how u do bad on the AIME\r\n\r\nremember, u need that 150 so ripply can see ur not-so-handsome picture on the perfect scores site",
"Solution_22": "[color=cyan]ouch, that was harsh Long . . .\n\nLOL[/color]",
"Solution_23": "I believe it repeats with period 5.",
"Solution_24": "Ok, so will someone actually post the ENTIRE solution?????",
"Solution_25": "I agree. If you solve a problem, post the full solution. If you're giving a hint, that's ok too, but if you're a solver for a particular problem, post the full solution."
}
{
"Tag": [
"analytic geometry",
"number theory",
"relatively prime"
],
"Problem": "In a rectangular coordinate system, a lattice point $ (a,b)$ is defined as a point $ (a,b)$ whose coordinates $ a$ and $ b$ are integers. If a dot is placed in the diagram on every lattice point whose coordinates are relatively prime, how many dots will there be?\n[asy]draw((2,2)--(2,5)--(5,5));\ndraw((3,2)--(3,4)--(5,4));\ndraw((4,2)--(4,3)--(5,3));\ndraw((2,3)--(4,3)--(4,5));\ndraw((2,4)--(3,4)--(3,5));\ndraw((2,2)--(5,2)--(5,5));\nlabel(\"$(2,2)$\",(2,2),SW);\nlabel(\"$(2,5)$\",(2,5),NW);\nlabel(\"$(5,5)$\",(5,5),NE);\nlabel(\"$(5,2)$\",(5,2),SE);[/asy]",
"Solution_1": "[asy]draw((2,2)--(2,5)--(5,5));\ndraw((3,2)--(3,4)--(5,4));\ndraw((4,2)--(4,3)--(5,3));\ndraw((2,3)--(4,3)--(4,5));\ndraw((2,4)--(3,4)--(3,5));\ndraw((2,2)--(5,2)--(5,5));\nlabel(\"$(2,2)$\",(2,2),SW);\nlabel(\"$(2,5)$\",(2,5),NW);\nlabel(\"$(5,5)$\",(5,5),NE);\nlabel(\"$(5,2)$\",(5,2),SE);\n\nlabel(\"$\\times$\",(2,2));\nlabel(\"$\\bigcirc$\",(2,3));\nlabel(\"$\\times$\",(2,4));\nlabel(\"$\\bigcirc$\",(2,5));\nlabel(\"$\\bigcirc$\",(3,2));\nlabel(\"$\\times$\",(3,3));\nlabel(\"$\\bigcirc$\",(3,4));\nlabel(\"$\\bigcirc$\",(3,5));\nlabel(\"$\\times$\",(4,2));\nlabel(\"$\\bigcirc$\",(4,3));\nlabel(\"$\\times$\",(4,4));\nlabel(\"$\\bigcirc$\",(4,5));\nlabel(\"$\\bigcirc$\",(5,2));\nlabel(\"$\\bigcirc$\",(5,3));\nlabel(\"$\\bigcirc$\",(5,4));\nlabel(\"$\\times$\",(5,5));[/asy]\r\n\r\nThus, $ \\boxed{10}$."
}
{
"Tag": [
"complex numbers",
"algebra open",
"algebra"
],
"Problem": "I have a question.\r\n\r\nTo prove with complex numbers that a triangle ABC is equilateral it's sufficient to show that a,b,c are distinct roots of a complex number.\r\n\r\n\r\n where A(a), B(b), C(c) .",
"Solution_1": "[quote=\"tarta\"] a,b,c are distinct roots of a complex number.\n[/quote]\r\n\r\nWhat is this? i.e? :maybe:",
"Solution_2": "Yes , if a,b,c are the roots of the equiation $z^{3}=r(cos\\phi+i\\cdot sin\\phi)$ $\\rightarrow$ A,B,C are points on the circle C(o,r) and ABC is equilateral."
}
{
"Tag": [
"Asymptote",
"LaTeX"
],
"Problem": "I am using Asymptote in MiKTex + TeXnicCenter, details can be found the next pages: \r\n \r\nhttp://www.mathlinks.ro/Wiki/index.php/Asymptote:_Getting_Started/Windows/Interactive_Mode \r\n \r\nhttp://www.mathlinks.ro/Wiki/index.php/Asymptote:_Advanced_Configuration \r\n \r\n \r\n \r\nSome time ago, I wrote the asymptote code in TeXnicCenter, and save the file as \r\nfilename.asy, then the pdf document with the image show up automatically. But now , on contrary, the pdf didn't open up automatically, though the pdf with the asympyote picture can generate. \r\n \r\nI didn't konw what's wrong. How to get the pdf show up automatically ? \r\nMany thanks",
"Solution_1": "I use windows at the moment and created a batch file. I know this is a late post, but I figure its as good a place as any to share.\nThe shell variables should be pretty self explanatory.\n\n1) Copy the code into a text file. \n2) Change the ASYMPTOTE_EXE and PDFLATEX_EXE to represent the location of your executables. \n3) Change the file extension from *.txt to *.bat.\n4) Use customize in TeXnicCenter to run the batch file (directions in file).\n\n[code]\nrem Purpose: This batch file is for creating pdf output from TeXnicCenter using Asymptote.\nrem Input: Takes a *.tex latex file that presumably has an Asymptote section in it.\nrem Output: outputs a *.pdf file in the same directory as the input file.\nrem Modification History: Created on 4/22/2010 \nrem\nrem ------------------------------------\nrem SETUP INSTRUCTIONS FOR USE WITH TEXNICCENTER\nrem In TeXniCCenter menubar, Tools->Customize, Tools tab, New Command (box looking icon)\nrem \tCommand: point to this batch file e.g. \"C:\\your_favorite_file.bat\"\nrem \tArguments: the *.tex file. Use \"%tc\"\nrem\tInitial directory: directory the *.tex file was in. Use \"%dc\"\nrem ------------------------------------\n\nrem set the location of your asymptote executable\nset ASYMPTOTE_EXE=\"C:\\Program Files (x86)\\Asymptote\\asy.exe\"\necho ASYMPTOTE_EXE: %ASYMPTOTE_EXE%\n\nrem set the location of your pdflatex executable\nset PDFLATEX_EXE=\"C:\\Program Files (x86)\\MiKTeX 2.8\\miktex\\bin\\pdflatex.exe\"\necho PDFLATEX_EXE: %PDFLATEX_EXE%\n\nrem convert the input file to long name\nset TEX_FILE_IN=%~f1%\necho TEX_FILE_IN: %TEX_FILE_IN%\n\nrem get the drivename\nset DRIVENAME=%~d1%\n\nrem get the basename\nset BASENAME=%~n1%\n\nrem get the directory name (aka path)\nset DIRNAME=%~p1%\n\nrem set the asymptote filename that will be generated by pdflatex\nset ASYMPTOTE_FILE_GENERATED=%DRIVENAME%%DIRNAME%%BASENAME%.asy\necho ASYMPTOTE_FILE_GENERATED: %ASYMPTOTE_FILE_GENERATED%\n\nrem generate *.asy file\n%PDFLATEX_EXE% \"%TEX_FILE_IN%\"\n\nrem generate pdf from *.asy file\n%ASYMPTOTE_EXE% \"%ASYMPTOTE_FILE_GENERATED%\"\n\nrem generate pdf from *.tex file and now includes Asymptote info\n%PDFLATEX_EXE% \"%TEX_FILE_IN%\"\n\nrem cleanup shell variables\nset \"TEX_FILE_IN=\"\nset \"DRIVENAME=\"\nset \"BASENAME=\"\nset \"DIRNAME=\"\nset \"PDFLATEX_EXE=\"\nset \"ASYMPTOTE_EXE=\"\nset \"ASYMPTOTE_FILE_GENERATED=\"\n\nrem uncomment last line to troubleshoot\nrem pause[/code]"
}
{
"Tag": [
"real analysis",
"real analysis unsolved"
],
"Problem": "Let $ (f_n) \\in L_p(X)$ be a sequence that converges almost everywhere to $ f$ and there exist $ g\\in L_p(X)$ such that $ |f_n| \\le g a.e.$, for all $ n \\in \\mathbb{N}$.\r\n\r\nProve that $ (f_n)$ converges almost uniformly to $ f$.\r\n\r\nI let $ E \\subseteq X$ such that $ m(E) \\equal{} 0.$\r\nSo, for every $ x\\in X\\cap E^c$,and for all $ \\epsilon >0,$ there exist an $ N(\\epsilon,x)\\in \\mathbb{N}$ such that for all $ n>N(\\epsilon,x),$ $ |f_n(x)\\minus{}f(x)| < \\epsilon$\r\n\r\nIn order to prove almost uniform convergence,I try to construct $ (f_n)$ so that $ f_n \\rightarrow f$uniformly on $ X\\cap E_\\delta$ where $ m(E_\\delta)< \\delta$ for $ \\delta >0$.\r\n\r\nBut how can I get rid of the dependence of the natural number on x?(so that it will become a uniform convergence on $ X\\cap E_\\delta)$",
"Solution_1": "Hint: The set where $ g>\\epsilon$ has finite measure, and on the set where $ g\\leq\\epsilon$, how big can $ |f_n\\minus{}f|$ be?"
}
{
"Tag": [
"number theory unsolved",
"number theory"
],
"Problem": "It's given a set $A$ of infinitely many distinct positive integers. Prove that there exists a prime $p>C$ ($C$ is an arbitrary constant) which devides $a+b$ where $a$ and $b$ are different elements of $A.$",
"Solution_1": "Where this problem are from?\r\nIt seems very nice! I don't have nice results, at this moment.\r\n\r\nYou know any generalization?",
"Solution_2": "Maybe this problem relation for sumsets theory, I thinks we can use following theorem \r\n\r\n[b]Theorem (Alon, Nathanson, Ruzsa)[/b]. Let $\\mathcal{A},\\mathcal{B}$ be nonempty subsets of $\\mathbb{Z}_{p}$ with $|\\mathcal{A}|\\not =|\\mathcal{B}|$, and let $\\mathcal{C}: =\\{a+b|a\\in\\mathcal{A},b\\in\\mathcal{B}, a\\not =b\\}$. Then $|\\mathcal{C}|\\geq \\min (p,|\\mathcal{A}|+|\\mathcal{B}|-2)$.",
"Solution_3": "I think that in this case you cannot find the prime ...you can only prove that there is some because we can choose the elements of $A$ to give some recidue mod prime and so no sum would be devicible by this prime.",
"Solution_4": "\u0417\u0434\u0440\u0430\u0432\u0435\u0439 :) ,\r\n\r\nYes, Alon's theorem stated above has no relevance to the problem. This problem (and many similar ones) are instantly killed by a deep theorem in diophantine approximations, the finiteness of the nondegenerate solution to an [i]$S$-unit equation[/i]. This theorem states that if $S = \\{p_{1},\\ldots,p_{r}\\}$ is any given finite set of primes and $n$ is fixed, then an equation $x_{1}+\\cdots+x_{n}= 1$ has only finitely many solutions $x_{i}$ in rational numbers of the shape $\\pm p_{1}^{k_{1}}\\cdots p_{r}^{k_{r}}$, with $k_{j}\\in \\mathbb{Z}$ (that is to say, only primes in $S$ appear in the numerators and denominators of the $x_{i}$), and such that no proper subsum of $x_{1}+\\cdots+x_{n}$ vanishes. To instantly kill your problem using this [extraordinarily deep!] result, assume to the contrary that all prime factors of the numbers $a+b$, $a,b \\in A, a \\neq b$, form a finite set $S$, and consider all various quadruples $a,b,c,d$ of pairwise distinct elements in $A$. Then by assumption, $X : = a+b, Y: = c+d, Z: = a+c, T : = b+d$ are all \"$S$-units\" (have all their prime factors in $S$), and moreover satisfy $X+Y = Z+T = a+b+c+d$, whereas $X/T+Y/T-Z/T = 1$, and all $X/T, Y/T, Z/T$ are $S$-units, and moreover it is self-evident that no subsum of $X/T+Y/T-Z/T$ vanishes. Applying the above-stated theorem with $n=3$, we conclude that $X/T, Y/T, Z/T$ can take on only finitely many values; but $X/T = (a+b)/(b+d)$ and $a,b,d \\in A$ were chosen perfectly arbitrary distinct elements of $A$, whence $A$ is finite, contradiction.\r\n\r\nBut of course, this problem doesn't require one to recourse to such a formidable result as the $3$-variable $S$-unit equation (which was only proven in the 1970s, I believe, and after previous futile efforts!) :) In fact, there is a simpler solution -- but I wouldn't like to spoil the fun, so I won't write it... yet :)",
"Solution_5": "This is a bit ugly but it seems to work.\r\n\r\nSuppose we have a counterexample: every element of $\\{a+b | a\\neq b\\in A\\}$ is divisible only by the primes in $S=\\{2,p_{1},\\ldots,p_{k}\\}$. Let $S$ be minimal among all such counterexamples.\r\n\r\nWe must have $k>0$. Otherwise, fixing $a,b\\in A$, we would have arbitrarily large pairs of powers of 2 that differ by the constant $b-a$, which is absurd.\r\n\r\nSuppose $p\\in S$ is an odd prime. For $i=0,\\ldots,p-1$, let $A_{i}=\\{a\\in A | a\\equiv i \\mod p\\}$. If $A_{i}$ is infinite, with $i\\neq 0$, then every sum of two elements of $A_{i}$ is $\\equiv 2i \\mod p$, hence divisible only by primes in $S\\setminus\\{p\\}$, contradicting minimality of $S$. Thus all but finitely many elements of $A$ are divisible by $p$. Passing to this cofinite set, and dividing out by $p$, we may similarly conclude that all but finitely many elements of $A$ are divisible by $p^{2}$. In fact, for all $n>0$, all but finitely many elements of $A$ are divisible by $Q^{n}$, where $Q=p_{1}\\ldots p_{k}$ is the product of the odd primes in $S$.\r\n\r\nDividing $A$ by a power of 2 if necessary, fix $x\\in A$ odd, and write $x=p_{1}^{\\alpha_{1}}\\ldots p_{k}^{\\alpha_{k}}u$, where $(u,Q)=1$. Fix $n>\\max\\{\\alpha_{i}\\}$.\r\n\r\nCase I: There are infinitely many even elements of $A$. By the above, we may pick $y=2Q^{n}v \\in A$. Now $x+y$ is odd, so by assumption, we can write\r\n$x+y = p_{1}^{\\alpha_{1}}\\ldots p_{k}^{\\alpha_{k}}u+2Q^{n}v = p_{1}^{\\beta_{1}}\\ldots p_{k}^{\\beta_{k}}$\r\nBut $\\alpha_{i}3$ there exist integers $x,y,k$ with $0<2kp$ has no sense so $x^2+y^2<2p^2$. And $k=0$ can't be OK because there aren't any squares whose sum is 3.\r\n\r\nNaphthalin",
"Solution_2": "Yes, and this was posted before:\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=76830 post #4\r\n\r\nTo show it for the prime fields in another way:\r\nLemma: every residue $\\mod p$ is sum of two squares.\r\n\r\nProof: Look at the numbers of type $x^2+1 \\mod p$.\r\nIf all of them would be squares, we would get that $1,2,3,...$ all are squares solving our problem (but this can only happen for $p=2$ indeed).\r\nSo there is at least one non-square of type $s=x^2+1 \\mod p$ otherwise.\r\nBut when $t,s$ are non-squares $\\mod p$, then $ts^{-1} \\mod p$ is a square. Using this for any given nonsquare $t$ and the $s$ above we get an $y$ with $y^2 \\equiv ts^{-1} \\mod p$, thus $t \\equiv tss^{-1} \\equiv y^2 (x^2+1) = (xy)^2+y^2 \\mod p$.\r\n\r\nMore generally we have:\r\n\r\n- Every prime residue $\\mod p^n$ is sum of two squares as long as $p \\neq 2$.\r\nAlso the residues $ap^k$ with $a$ coprime to $p$ and $k$ even and $p$ has no sense so $x^2+y^2<2p^2$. And $k=0$ can't be OK because there aren't any squares whose sum is 3.\n\nNaphthalin[/quote]\r\nYou are not right. You proved, that there is $k<2p$, but not $2k3$ we can see that $k>0$(neither $k\\leq -1$ nor $k=0$($x^2+y^2\\not=3$)."
}
{
"Tag": [
"trigonometry",
"function",
"real analysis",
"real analysis unsolved"
],
"Problem": "Is there some way to show that\r\n\r\nfor x = $ k\\pi/2\\sqrt2$, k a (edit: nonzero) integer,\r\n\r\ncos(x) is never of the form $ a/2^{b}$ with a,b integers?\r\n\r\nTo generalize - given that cos(x) is rational, can we say anything about x?\r\n\r\nI've been playing with convergence arguments on the sum / product formulas for the function, but I'm kind of stuck...\r\n\r\nThanks very much for your help.",
"Solution_1": "Take $ k=0$. Then $ \\cos 0=1$. Then take $ a=4,b=2$.",
"Solution_2": "Well, the Gelfond-Schneider theorem says that $ \\alpha^\\beta$ is transcendental if $ \\alpha$ is algebraic and different from $ 0$, $ 1$; and $ \\beta$ is algebraic and irrational. This takes care of your question (just take $ \\alpha=e^{i\\pi/4}$, $ \\beta=\\sqrt 2$) but it is not an easy theorem. On the other hand, I doubt some nice elementary approach is possible (as usual, I'll be excited to be proved wrong on this account :))."
}
{
"Tag": [
"probability",
"probability and stats"
],
"Problem": "30% of bottles produced in a certain plant are defective. If a bottle is defective, the probability that an inspector will notice and remove it from the filling line is 0.9. If the bottle is not defective, the probability that the inspector will think it is defective and remove it from the filling line is 0.2.\r\nIf a bottle is removed, what is the probability that it is defective?",
"Solution_1": "since there is a .9 probability that a bottle will be removed a worker, which is defective, that means 90 out of 100 bottles will be thought to be defective and removed. Also, since there is a .2 probability that there is a normal bottle that might be removed by a worker, that means 20 bottles out of 100 will be removed, but is not defective. So the probability that it is defective is 70%."
}
{
"Tag": [
"real analysis",
"real analysis theorems"
],
"Problem": "let $ f(x,y,z)\\equal{}xy^2\\plus{}yz^2\\plus{}zx^2$\r\n\r\ngiven $ c>0$ (const.) find\r\n\\[ \\sup_{x,y,z>0\\land x\\plus{}y\\plus{}z\\equal{}c}f(x,y,z)\\]\r\nyou mustn't use Lagrange multipliers",
"Solution_1": "any idea for this excercise ?",
"Solution_2": "http://www.artofproblemsolving.com/Forum/viewtopic.php?t=267499"
}
{
"Tag": [
"function",
"calculus",
"integration",
"trigonometry",
"algebra",
"polynomial",
"calculus computations"
],
"Problem": "Hi! I just fished a math project and got the grade on it but the one question I got wrong is bothering me I must know the answer can anyone help me? here is that question. \r\n[img]http://img.photobucket.com/albums/v705/pantha2/bipartisan-problem.gif[/img]\r\nAlso Could you show me how you got that answer? Thank you.",
"Solution_1": "I am assuming $y$ is a function of $t.$ This is an example of a first order linear ODE, and the method of solving it is a standard one that can be found in most Calculus textbooks.\r\n\r\nThe general form of such an ODE is, $\\frac{dy}{dt}+P(t)\\cdot y = Q(t),$ where $P(t), Q(t)$ and $y$ are functions of $t.$ \r\n\r\nFirst, we need to find an integrating factor (I.F.), which in the general case is $e^{\\int P(t)\\, dt}.$ In this problem, the I.F. is $e^{\\int-\\frac12 \\, dt}= e^{-\\frac{t}{2}}.$ We multiply both sides of the equation $\\frac{dy}{dt}-\\frac12 y = 2\\cos t$ with the above I.F to get\r\n\r\n$e^{-\\frac{t}{2}}\\cdot \\frac{dy}{dt}-y \\cdot \\frac12 e^{-\\frac{t}{2}}= 2\\cos t \\cdot e^{-\\frac{t}{2}}$\r\n\r\n$\\Rightarrow \\frac{d}{dt}\\left( y\\cdot e^{-\\frac{t}{2}}\\right) = 2\\cos t \\cdot e^{-\\frac{t}{2}}$\r\n\r\n$\\Rightarrow \\int d(y\\cdot e^{-\\frac{t}{2}}) = \\int 2\\cos t \\cdot e^{-\\frac{t}{2}}\\, dt$\r\n\r\n$\\Rightarrow y\\cdot e^{-\\frac{t}{2}}= 2\\int \\cos t \\cdot e^{-\\frac{t}{2}}\\, dt+C$\r\n\r\nJust use integration by parts to evaluate the integral on the RHS of the equation above. I will that as an exercise for you.",
"Solution_2": "OMGosh! Thank you! That helps a lot.",
"Solution_3": "Another method is to break the problem into its homogenous and particular equations. The homogeneous equation is $y'-\\frac{1}{2}y = 0$, which is straight-forward to solve. Then you just need one particular solution to the equation. If you just guess $y = a\\cos t+b\\sin t$ you get $y' =-a \\sin t+b\\cos t$ and so $y'-\\frac{y}{2}= (-a-\\frac{b}{2})\\sin t+(b-\\frac{a}{2})\\cos t = 2\\cos t$ so $-a-\\frac{b}{2}= 0$ and $b-\\frac{a}{2}= 2$, an system which is easy enough to solve. Then add the two solutions together to get the solution to the original equation.",
"Solution_4": "We spend more energy classifying problems in a differential equations course than in most other courses I know. Why? Because a succesful classification leads us to a method.\r\n\r\nFieryHydra classified this as a first order linear ODE. That leads to the method of integrating factors.\r\n\r\nJBL classified this as a constant coefficient linear ODE, nonhomogeneous, but with the nonhomegenity lying within the \"exponential-polynomial\" class of functions generated by sums of products of polynomials, exponentials, sines, and cosines. That classification leads to the method of undetermined coefficients.\r\n\r\nThe useful classifications are not disjoint, and this particular problem does lie within the appropriate intersection. JBL's classification is usually seen in the context of second order or higher equations, but there's no reason not to include first order equations as part of the same classification, as appropriate."
}
{
"Tag": [
"modular arithmetic",
"number theory proposed",
"number theory"
],
"Problem": "Let $ x,y\\in N$ satisfy :\r\n$ x|y^{x\\minus{}1}\\plus{}1$ where $ x>1$\r\nProve that $ y\\equiv 1 (\\mod 2)$",
"Solution_1": "[quote=\"TTsphn\"]Let $ x,y\\in N$ satisfy :\n$ x|y^{x \\minus{} 1} \\plus{} 1$ where $ x > 1$\nProve that $ y\\equiv 1 (\\mod 2)$[/quote]\r\n\r\nIf $ y$ is even, then $ x$ is odd and all prime divisors of $ x$ are equal to 1 modulus $ 2^1$\r\n\r\nAssume now, that all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for some positive integer $ p$.\r\nThen $ x\\equal{}1\\pmod{2^p}$ and then $ x\\minus{}1\\equal{}k2^p$ for some k.\r\nThen, for any prime divisor $ q2^p\\plus{}1$ of x, we have $ \\minus{}1\\equal{}A^{2^p}\\pmod{q2^p\\plus{}1}$ where $ A\\equal{}y^k$\r\nRaising this equality at power $ q$, we have $ (\\minus{}1)^q\\equal{}A^{q2^p}\\pmod{q2^p\\plus{}1}$; But, since $ q2^p\\plus{}1$ divides x, it does not divide A and so $ A^{q2^p}\\equal{}1\\pmod{q2^p\\plus{}1}$ (fermat) and so $ q$ is even.\r\n\r\nSo, if all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for some positive integer $ p$, all prime divisors of $ x$ are equal to 1 modulus $ 2^{p\\plus{}1}$.\r\n\r\nAnd, since $ y$ even implies all prime divisors of $ x$ are equal to 1 modulus $ 2^1$, all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for any positive integer $ p$. And this is clearly impossible if $ x\\neq 1$\r\n\r\nSo $ x>1$ implies $ y$ odd.",
"Solution_2": "[quote=\"pco\"][quote=\"TTsphn\"]Let $ x,y\\in N$ satisfy :\n$ x|y^{x \\minus{} 1} \\plus{} 1$ where $ x > 1$\nProve that $ y\\equiv 1 (\\mod 2)$[/quote]\n\nIf $ y$ is even, then $ x$ is odd and all prime divisors of $ x$ are equal to 1 modulus $ 2^1$\n\nAssume now, that all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for some positive integer $ p$.\nThen $ x \\equal{} 1\\pmod{2^p}$ and then $ x \\minus{} 1 \\equal{} k2^p$ for some k.\nThen, for any prime divisor $ q2^p \\plus{} 1$ of x, we have $ \\minus{} 1 \\equal{} A^{2^p}\\pmod{q2^p \\plus{} 1}$ where $ A \\equal{} y^k$\nRaising this equality at power $ q$, we have $ ( \\minus{} 1)^q \\equal{} A^{q2^p}\\pmod{q2^p \\plus{} 1}$; But, since $ q2^p \\plus{} 1$ divides x, it does not divide A and so $ A^{q2^p} \\equal{} 1\\pmod{q2^p \\plus{} 1}$ (fermat) and so $ q$ is even.\n\nSo, if all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for some positive integer $ p$, all prime divisors of $ x$ are equal to 1 modulus $ 2^{p \\plus{} 1}$.\n\nAnd, since $ y$ even implies all prime divisors of $ x$ are equal to 1 modulus $ 2^1$, all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for any positive integer $ p$. And this is clearly impossible if $ x\\neq 1$\n\nSo $ x > 1$ implies $ y$ odd.[/quote]\r\nGood solution.I have other method but I will post after you solve same problem: \r\nProve that $ x\\equiv 0 (\\mod 2)$\r\nI think here is an hard problem : Does $ (x,y)$ infinite?",
"Solution_3": "[quote=\"TTsphn\"]\nProve that $ x\\equiv 0 (\\mod 2)$\nI think here is an hard problem : Does $ (x,y)$ infinite?[/quote]\r\n\r\nLet $ p>2$ be an odd prime. Then $ (2p\\minus{}1)^{2p\\minus{}1}\\plus{}1\\equal{}(\\minus{}1)^{2p\\minus{}1}\\plus{}1\\equal{}0\\pmod{2p}$\r\n\r\nAnd so $ x\\equal{}2p$ and $ y\\equal{}2p\\minus{}1$ are a solution to the equation $ x|y^{x\\minus{}1}\\plus{}1$\r\n\r\nAnd there are an infinity of such pairs $ (x,y)$",
"Solution_4": "Good solution. But can you prove $ x\\equiv 0(\\mod 2)$",
"Solution_5": "[quote=\"TTsphn\"]Good solution. But can you prove $ x\\equiv 0(\\mod 2)$[/quote]\r\nHello TTsphn,\r\n\r\nIt's in my first post in this thread :\r\n\r\n$ x$ odd implies all prime divisors of $ x$ are equal to 1 modulus $ 2^p$ for any positive integer p.\r\n\r\nSo $ x$ odd implies $ x\\equal{}1$\r\n\r\nSo $ x>1$ implies $ x$ even (and so $ y$ odd, which was the target in the first post).",
"Solution_6": "Hello pco. Here is my solution.\r\nCall p is a prime divisor of n such that $ v_2(p\\minus{}1)$ is min. \r\nImply that $ v_2(p\\minus{}1)\\leq v_2(n\\minus{}1)$\r\nBut $ p|\\gcd(2^{p\\minus{}1}\\minus{}1,2^{2(n\\minus{}1)}\\minus{}1)\\equal{}2^{\\gcd(p\\minus{}1,2(n\\minus{}1)}\\minus{}1|2^{n\\minus{}1}\\minus{}1$\r\nSo $ p|2\\Longleftrightarrow p\\equal{}2$\r\nSo $ x\\equiv 0(\\mod 2)$\r\nFrom this $ y\\equiv 1(\\mod 2)$",
"Solution_7": "[quote=\"TTsphn\"]Hello pco. Here is my solution.\nCall p is a prime divisor of n such that $ v_2(p \\minus{} 1)$ is min. \nImply that $ v_2(p \\minus{} 1)\\leq v_2(n \\minus{} 1)$\nBut $ p|\\gcd(2^{p \\minus{} 1} \\minus{} 1,2^{2(n \\minus{} 1)} \\minus{} 1) \\equal{} 2^{\\gcd(p \\minus{} 1,2(n \\minus{} 1)} \\minus{} 1|2^{n \\minus{} 1} \\minus{} 1$\nSo $ p|2\\Longleftrightarrow p \\equal{} 2$\nSo $ x\\equiv 0(\\mod 2)$\nFrom this $ y\\equiv 1(\\mod 2)$[/quote]\r\n\r\nEuhhh, I'm sorry.\r\n\r\nI don't understand why $ p|\\gcd(2^{p \\minus{} 1} \\minus{} 1,2^{2(n \\minus{} 1)} \\minus{} 1)$. Sure, $ p|2^{p \\minus{} 1} \\minus{} 1$ but there is no reason for $ p|2^{2(n \\minus{} 1)} \\minus{} 1$.\r\n\r\nFor example, with $ n\\equal{}33$ and $ p\\equal{}11$, $ p\\equal{}11$ divides $ 2^{p\\minus{}1}\\minus{}1\\equal{}1023\\equal{}11\\times 93$ but $ p$ does not divide $ 2^{2(n\\minus{}1)}\\minus{}1\\equal{}2^{64}\\minus{}1\\equal{}4\\pmod{11}$",
"Solution_8": "I am sorry .Replace 2 by $ y$",
"Solution_9": "Here is real hard problem . \r\nFor each $ y\\in N$ if there exist infinite $ x$ such that $ x|y^{x\\minus{}1}\\plus{}1$\r\nFor the case $ y\\equiv 0(\\mod 2)$ then it has no solution $ x>1$\r\nBut in other case?",
"Solution_10": "It is not hard. Let $ x|y^{x-1}+1$ and x>1 is odd, $ p|x$ is prime divisor. $ x-1$ is even, therefore period y mod p - $ T{}_y{}_p$ must be multiple 4 ($ 4|T{}_y{}_p$). If for any prime divisor $ p=1\\mod 2^k$, we get $ 2^k|x-1$, therefore for period $ T{}_y{}_p$ must be $ 2^{k+1}|T{}_y{}_p\\Longrightarrow p=1\\mod 2^{k+1}$. Therefore no odd solution x>1.\r\nIf $ y$ is even $ 2|y^{x-1}+1$ for $ x>1$, therefore no solution x>1.\r\nIf y>1 is odd, we consider odd primes $ p_i$, suth that $ m_p=\\frac{T{}_y{}_p}{2}$ is odd and we can find solution $ x= 2^kp_1...p_m$. It is solution if and only if $ m{}_{p{}_i}|2^{k}p_1...p_m-1$.\r\nFor any prime p, suth that $ m_p$ is odd $ x=2^kp$ solution if $ 2^k|y+1$ and $ m_p|2^k-1$."
}
{
"Tag": [
"ARML",
"search"
],
"Problem": "lol, i don't even know how to begin.\r\n\r\nThere are several values for a prime, p, with the property that any five-digit multiple of p remains a multiple of p under \"cyclic permutation.\" Once such value is 41 (for example, since 50635 is a multiple of 41, so are 55063, 35506, 63550, and 6355); another such value of p is 3. Computer the value of p greater than 41.",
"Solution_1": "Posted [url=http://www.artofproblemsolving.com/Forum/viewtopic.php?search_id=132837513&t=128972]here[/url].",
"Solution_2": "oops, sorry about that :blush:"
}
{
"Tag": [
"function",
"inequalities",
"AMC",
"USA(J)MO",
"USAMO",
"floor function",
"algebra unsolved"
],
"Problem": "Prove that [nx] => [x]+[2x]/2+...+[nx]/n where n is a natural number and x any real number",
"Solution_1": "I think should be $ x>0$\r\nLet \r\n$ f(x)\\equal{}[nx]\\minus{}(\\frac{[x]}{1}\\plus{}\\frac{[2x]}{2}\\plus{}...\\plus{}\\frac{[nx]}{n}$\r\n$ f(x\\plus{}1)\\equal{}f(x)$\r\nSo enough to prove that $ f(x)\\geq 0 ,\\forall x\\in [0,1)$\r\nWe will prove the equalities:\r\n$ [x\\plus{}\\frac{n\\minus{}k}{n}]\\geq \\frac{[kx]}{k},\\forall 00 doesn't hold when k=5, n=9, x=4/9",
"Solution_3": "This is an old USAMO problem (1981/5?). A generalization was proposed as APMO 1999/2 (?), which was the following (paraphrased): If $ \\{a_i\\}$ is a sequence of real numbers satisfying $ a_i\\plus{}a_j\\geq a_{i\\plus{}j}\\ \\forall i,j\\in [1,n]\\cap\\mathbb{N}$, then $ \\sum_{i\\equal{}1}^n \\frac{a_i}{i}\\geq a_n$. Taking $ a_i\\equal{}\\minus{}\\lfloor ix\\rfloor$, the result is obtained. The generalization can be solved using a simple induction.",
"Solution_4": "comes easily from Hermite identity",
"Solution_5": "Can you explain that please?",
"Solution_6": "Can someone write the solution using Hermite's Id. ( manuscript is OK )!?\ntnx a lot"
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "Let $ a,b,c\\geq 0$, s.t. $ a^3\\plus{}b^3\\plus{}c^3\\equal{}3$.Prove that:\r\n$ a^4b^3\\plus{}b^4c^3\\plus{}c^4a^3\\leq 3$",
"Solution_1": "We have: $ \\sum\\ a^6b^3.\\sum\\ a^3b^3. \\sum\\ a^3b^3 \\geq\\ (\\sum\\ a^4b^3)^3$\r\nWe only need to prove: $ \\sum\\ a^6b^3.\\sum\\ a^3b^3. \\sum\\ a^3b^3 \\leq\\ 27$\r\nIt can rewrite: Given $ a, b, c$ satisfy $ a \\plus{} b \\plus{} c \\equal{} 3.$\r\nProve that: $ (a^2b \\plus{} b^2 \\plus{} c^2a)(ab \\plus{} bc \\plus{} ca)^2 \\leq\\ 27$\r\nIt's obviously trues because: $ (a^2b \\plus{} b^2c \\plus{} c^2a)(ab \\plus{} bc \\plus{} ca) \\leq\\ 9$\r\nand $ ab \\plus{} bc \\plus{} ca \\leq\\ 3$\r\n :)"
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "Find all pairs of integers $ (x,y)$ satisfying $ 1\\plus{}1996x\\plus{}1998y\\equal{}xy.$",
"Solution_1": "$ 1\\plus{}1996x\\plus{}1998y\\equal{}xy$\r\n$ \\iff (x\\minus{}1998)(y\\minus{}1996)\\equal{}(1997\\plus{}1)(1997\\minus{}1)\\minus{}1\\equal{}1997^2\\plus{}1\\minus{}1\\equal{}1997^2$\r\n\r\nThus,\r\n$ (x, y)$\r\n$ \\equal{}(1998\\minus{}1997^2, 1996\\minus{}1) \\ \\ \\text{i.e.} \\ \\ (\\minus{}3986011, 1995)$\r\n$ \\equal{}(1998\\minus{}1997, 1996\\minus{}1997) \\ \\ \\text{i.e.} \\ \\ (1, \\minus{}1)$\r\n$ \\equal{}(1998\\minus{}1, 1996\\minus{}1997^2) \\ \\ \\text{i.e.} \\ \\ (1997, \\minus{}3986013)$\r\n$ \\equal{}(1998\\plus{}1, 1996\\plus{}1997^2) \\ \\ \\text{i.e.} \\ \\ (1999, 3990005)$\r\n$ \\equal{}(1998\\plus{}1997, 1996\\plus{}1997) \\ \\ \\text{i.e.} \\ \\ (3995, 3993)$\r\n$ \\equal{}(1998\\plus{}1997^2, 1996\\plus{}1) \\ \\ \\text{i.e.} \\ \\ (3990007, 1997)$"
}
{
"Tag": [
"probability"
],
"Problem": "Marcus and Al are playing Rock/Paper/Scissors. The first person to win 10 times does not have to do the dishes. The score is now 9 to 7 in favor of Marcus. If ties are ignored and not counted, what is the probability Marcus will not have to do the dishes? Express your answers as a common fraction.",
"Solution_1": "[hide]i got 1/2 + 1/4 + 1/8, so 7/8. there is a 1/2 chance of him winning the first game, a 1/4 of losing the first but winning the second, and a 1/8 of losing the first 2 but winning the last one.[/hide]",
"Solution_2": "[hide]Well, the prob that he loses is 1/2 * 1/2 * 1/2, because to lose he has to lose 3 games in a row. \n\nSo the prob he wins is 7/8.[/hide]",
"Solution_3": "[hide]The probability of him winning one game is $\\frac12$. The probability of him losing then winning is $\\frac12\\cdot\\frac12=\\frac14$.\nThe probability of him losing two then winning is $\\frac18$. $\\frac12+\\frac14+\\frac18=\\frac78$.[/hide]"
}
{
"Tag": [
"number theory",
"least common multiple",
"inequalities",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Find \\[ \\left [\\mathbb{Q}(\\sqrt [3] {2},\\sqrt{1+\\sqrt{3}}): \\mathbb{Q}(\\sqrt [3] {2})\\right]; \\]\r\n\r\n \\[ \\left [\\mathbb{Q}(\\sqrt{2},\\sqrt{3},\\sqrt [3] {5}): \\mathbb{Q}(\\sqrt{2})\\right]; \\]\r\n\r\n \\[ \\left[\\mathbb{Q}(\\sqrt [3] {2},\\sqrt{3},\\sqrt{5}): \\mathbb{Q}(\\sqrt{3}+\\sqrt{5})\\right]. \\]",
"Solution_1": "$4$\r\n\r\n$6$\r\n\r\n$2$ (note that $\\frac{2}{\\sqrt{3}+\\sqrt{5}}=-\\sqrt{3}+\\sqrt{5}$, thus $\\sqrt{3}, \\sqrt{5}$ are contained in the right one)\r\n\r\nAll by using $\\mathrm{lcm}([L: K],[M: K]) \\leq [LM : K] \\leq [L: K] \\cdot [M: K]$.",
"Solution_2": "I have posted another one. ;)",
"Solution_3": "I have added another one. :P",
"Solution_4": "But, it would be better if you could post full solutions. I have some friends who are watching the topic and they don't understand all you have said. In fact, they asked me to post this kind of exercices. :)",
"Solution_5": "The inequality $\\mathrm{lcm}([L: K],[M: K]) \\leq [LM : K] \\leq [L: K] \\cdot [M: K]$ comes from:\r\n- $[LM : K]$ must be a multiple of $[L: K]$ and $[M: K]$, thus the LHS.\r\n- when $a_1 , ... , a_{[L: K]}$ is a basis of $L|K$ and $b_1 , ... , b_{[M: K]}$ for $M|K$, then $LM|K$ is generated by the products $a_i b_j$, thus the RHS.\r\n\r\nTo most ugly part indeed is to show that some of the degrees, e.g. that one of $\\sqrt{1+\\sqrt{3}}$ is $4$ and not $2$, which can be done by calculation:\r\nassume that $\\sqrt{1+\\sqrt{3}}$ has just degree $2$, then $\\sqrt{1+\\sqrt{3}}$ must be a square $a^2$ in $\\mathbb{Q}[\\sqrt{3}]$. Then also $\\overline{a}^2 = \\overline{1+\\sqrt{3}} = 1-\\sqrt{3}$ (overline=conjugation). But $\\overline{a}$ is real, thus $\\overline{a}^2 \\geq 0$, which contradicts $1-\\sqrt{3}<0$.\r\n\r\nOthers are similar, something that works everytime is just ugly calculation that often gets simpler by conjugates, norms etc."
}
{
"Tag": [],
"Problem": "http://www.sporcle.com/games/world.php\r\n\r\nPost your results! \r\n\r\nMine:\r\n[hide=\"Spoiler\"]\n[img]http://i42.tinypic.com/5vskfd.png[/img]\nI MISSED EGYPT. Argh.\n[/hide]",
"Solution_1": "wow i got 99...",
"Solution_2": "Argh.\r\n\r\nMissed Hungary, Honduras, Cote d'Ivoire, Republic of the Congo, and Boliva...for a total of 190/195. ARGH.",
"Solution_3": "I got to about 60 before I resorted to [url]http://www.youtube.com/watch?v=IDtdQ8bTvRc[/url].",
"Solution_4": "I got 190...because I can't spell; I memorized all the countries in 5th grade, but I just can't spell some of them properly. :blush: I did pretty well on the fifty states \"typing test,\" though. (Yes, since it's so easy, it's practically a typing test.) I typed fifty states in 94 seconds, which is 1.88 seconds per state. \r\n\r\nBTW...animaniacs = very unreliable source. According to them, the following are countries: \r\n\"Republic Dominican\" (should be Dominican Republic)\r\nCaribbean (region with many countries)\r\nGreenland (owned by Denmark)\r\nPuerto Rico (owned by United States)\r\nBermuda (owned by United Kingdom\r\n\"Tobago\" (should be Trinidad and Tobago)\r\nSan Juan (capital of Puerto Rico)\r\nFrench Guyana (possibly still a country when Animaniacs made the song, but is now \"reclaimed\" property of France)\r\nGuam (owned by United States)\r\nCzechoslovakia (split into Czech Republic and Slovakia)\r\nScotland (one of the divisions of United Kingdom)\r\n\"Both Yemens\" (I believe that Animaniacs isn't so old that the Yemens haven't joined yet)\r\nEngland (one of the divisions of United Kingdom)\r\nBurma (the proper name is Myanmar)\r\nKampuchea (the proper name is Cambodia)\r\nAsia (a continent; wow, did they actually think people won't notice?)\r\nKorea (Korea is split into North Korea and South Korea)\r\nTibet (owned by China)\r\n\"New Guinea\" (should be Papua New Guinea; New Guinea [or the alternate Indonesian name of Irian Jaya] refers to the island)\r\nBorneo,Sumatra (islands, part of Indonesia, can't believe they actually put that)\r\nOkay, they said \"the Spanish Sahara is gone\"...I guess I'll assume that they know it's gone...\r\n\"Algier\" (should be Algeria)\r\nDahomey (ancient empire centered in Benin)\r\nZaire (forgivable; when Animaniacs came out it was probably still called Zaire; now it's the Democratic Republic of Congo)\r\nMahore (owned by France)\r\n\"Cayman\" (if they're referring to \"Cayman Islands,\" that's owned by the United Kingdom, if they're referring to something else, they're referring to something inexistant)\r\nHong Kong (still owned by China)\r\nAbu Dhabi (capital of the United Arab Emirates)\r\nYugoslavia (split into Slovenia, Croatia, Bosnia and Herzegovina, Serbia, Montenegro, Kosovo, and Macedonia)\r\nCrete (island belonging to Greece)\r\nTransylvania (regoin on Romania\r\nPalestine (still considered part of Israel)\r\n\r\nThat's a lot; do you think they thought they'd get away with all that crap? Most of us would definitely be able to identify most of that by like 5th grade. They also left a whole bunch out. Seriously, bad, bad, bad. They also did a similar thing with making up chemical compounds in an extremely blatant fashion in the Ice Cream Song; seriously, who would fall for \"deoxylite trisilicon\"?",
"Solution_5": "66. I suck at Geography.",
"Solution_6": "OMG... I got 108.",
"Solution_7": "58. I feel sad.",
"Solution_8": "Saddest Quote Ever: \"It took them 40 minutes to locate Canada on the map.\"",
"Solution_9": "[quote=\"r15s11z55y89w21\"]\"Both Yemens\" (I believe that Animaniacs isn't so old that the Yemens haven't joined yet)\n\nZaire (forgivable; when Animaniacs came out it was probably still called Zaire; now it's the Democratic Republic of Congo)\n\nHong Kong (still owned by China)\n\nYugoslavia (split into Slovenia, Croatia, Bosnia and Herzegovina, Serbia, Montenegro, Kosovo, and Macedonia)[/quote]\r\n\r\nAirdate of said animaniacs song: 1993\r\n\r\nThe two Yemens joined cerca 1990, so that one is a mistake.\r\n\r\nZaire still existed up until 1997. \r\n\r\nHong Kong was still a dependent territory of the UK until its return to China in 1997. Even then, it's largely autonomous.\r\n\r\nFederal Republic of Yugoslavia existed up until 2003 (albeit only established in 1992).\r\n\r\nOtherwise... yeah. XD\r\n\r\nEpically failed the 50 states typing test (07:22).",
"Solution_10": "I got all of the Africa countries though cause I can draw a mental map of all 53.",
"Solution_11": "I got $ 88$.\r\n\r\nFor U.S states, I got $ 2: 42$. :lol:",
"Solution_12": "I got 106.\r\n\r\nI was going to type dominican republic and then it put it as Dominica. That is one bad thing about it.",
"Solution_13": "112, I am really slow at typing. :blush:",
"Solution_14": "88. I missed like everything in africa except the really obvious ones :oops:",
"Solution_15": "46...for some reason it wouldn't let me do pretty easy countries, like eqypt. :(",
"Solution_16": "That's because it's spelled Egypt.",
"Solution_17": "132... failed all of the western african countries and the pacific islands...",
"Solution_18": "It took me 2 minutes 4 seconds to name all the states. :D",
"Solution_19": "101 countries...got only 14 from Africa and like 3 from Oceania/Austrailia...also I can't remember all the -istan countries :P.",
"Solution_20": "Pakistan\r\nAfghanistan\r\nUzbekistan\r\nTurkmenistan\r\nKazakhstan\r\nKyrgyzstan\r\nI know I'm forgetting one...",
"Solution_21": "Tajikastan (something tells me I misspelled that)",
"Solution_22": "I got 125 countries. I used to go on geography sprees, thinking I could bang out every single country in about 5 continents but the rest threw me off. Need more practice, especially in Central America and the Pacific nations.",
"Solution_23": "[quote=\"r15s11z55y89w21\"]Tajikastan[/quote]\r\nTHAT'S it!\r\nAnd it's Tajikistan",
"Solution_24": "122.. I don't know if it's good... OR BAD O:",
"Solution_25": "109. I can't believe I missed Belgium :wallbash_red:",
"Solution_26": "113. But I forgot a bunch, including:\r\nNorth Korea\r\nSouth Korea\r\nKenya\r\nBelgium\r\nGuatemala\r\nPhillippines\r\nMacedonia\r\nand a bunch of other obvious ones........................ :censored:",
"Solution_27": "138\r\nI used to know all the countries in 4th grade, but I forgot some of the African countries. :o :o :o",
"Solution_28": "44\r\n\r\nDang it. I used to know much more than that. :(",
"Solution_29": "I improved to 154.\r\n\r\nThe trick is to tackle everything continent by continent.............."
}
{
"Tag": [],
"Problem": "Here's the ques\r\n\r\nThe major product will be (2)\r\n\r\nThe question is why??\r\n\r\nplz answer fast",
"Solution_1": "Here allylic addition of bromine has taken place in 2 which stabilises 2 :)",
"Solution_2": "Not sure but, Since 2 is a conjugate system Resonance comes into play thereby stabilizing the system. In 1 the double bond is quite away from the site of reaction and is less stable as campared to 2. :maybe:",
"Solution_3": "[quote=\"murgi\"]Here allylic addition of bromine has taken place in 2 which stabilises 2 :)[/quote]\r\nif u see carefully,it is vinylic addition rather than allylic\r\n\r\n\r\n@pulkit-Do u mean the resonance betw lp of Br and double bond in 2 :) \r\n\r\n-acc to resonance,dounle bond will be in resonance with intermediate carbocation only in 1 :P",
"Solution_4": "an edit... (not my mistake-my teacher wrote it wrong :mad: )\r\nNow addition does takes place at allylic carbon\r\nBut still carbocation intermediate of 1 is more stabilised by resonance.... :?:",
"Solution_5": "I think that alkene in first is a terminal alkene thereby is more reactive than 2, Hence less stable.\r\n\r\nBut I also really feel that 1 will be more stable due to tertiary carbocation,more hyperconjugative pairs :fool: \r\n\r\nGuess I would have to wait till my tutor teaches me alkene :blush:",
"Solution_6": "The most stable ressonance structure was the first one on my mechanism, not the second. However, one reason the major product is the 1-bromo is that its double bond is more stable, since it is more substituted.",
"Solution_7": "But still how is the third figure in your post the most stable one? :o \r\n\r\nThe carbocation in carcul's second figure is a tertiary one and should be most stable. :(",
"Solution_8": "carcul sir, we were able to guess out this mechanism...ques is just that why the primary c+ is more stable resonance structure :( :)",
"Solution_9": "I actually think, this reaction is done in present of peroxide. First addition then elimination gives you product 2 which is also the major product.",
"Solution_10": "[quote=\"Conjurer\"]I actually think, this reaction is done in present of peroxide. First addition then elimination gives you product 2 which is also the major product.[/quote]\r\nI agree................",
"Solution_11": "Peroxide is completely absent,i m sure :( :maybe:",
"Solution_12": "got the answer...Themodyanamically stable product :mad: :maybe: but how??? :(",
"Solution_13": "More substituted double bond?\r\n\r\nNice question btw.",
"Solution_14": "Case of Kinetics versus thermodynamics...\r\n\r\nWould you like to go for a kinetically stable product (which proceed through a more stable intermediate) or would you like to go for a thermodynamically stable product (which in a majority of the cases, means a more stable product)?\r\n\r\nThe choice is yours... :P \r\n\r\nAnd Conjurer, what makes you think of peroxide?",
"Solution_15": "I THINK THE QUESTION IS WRONG. THE DOUBLE BOND IN THE OPTION 2 WILL SHIFT TOWARDS LEFT FOR ONE BECAUSE OF HYPERCONJUGATION OF POSITIVE CHARGE. :D :D :D ",
"Solution_16": "It is true that the double bond in choice 2 is misplaced, and the question is basically asking if it is better to 1) have a more stable positive charge on the tertiary center or 2) whether it is better to have a double bond with more substituents. Kinetically, the first product will likely be favored (as hell_ever noted above) since the positive charge is in the more stable place, while thermodynamically, the second will be favored. This is because the final products do not have a positive charge anywhere, so the main factor is the number of substituents on the double bond",
"Solution_17": "Not only in thermo aspect but you could determine stability based on arylic and vinylic position of $Br$"
}
{
"Tag": [
"Mafia",
"search"
],
"Problem": "Hi. I'll like to see if there wich are the differences between the mafia games that are played in your country, for example, here in Mexico (and in most of latinamerica):\r\nThe only roles are civilians, mafia (also called assasins), inspector and angel (mostly called doctor). \r\n\r\nThere are other games like one called Vampires, th diferences are:\r\n1) Instead of mafia, there is a Vampire who instead of killing other people, makes them vampires (one person who has another role, might be also converted, in which case, will have both roles).\r\n2) Inspector and angel roles are similars (searchs for vampires and heals from being converted to a vampire)\r\n3) There is another character, The \"santo\" who select a person each turn, if that person is a vampire (or is turned into a vampire in that turn) dies.\r\n\r\nAlso, Does anybody knows were the mafia started? I've heared that it was in Argentina, but i'm not sure",
"Solution_1": "For United States\r\nwell if u re not playing online, and ure just playing with cards, there are usually citizens, mafia, inspector, and doctor...\r\nSometimes there are serial killers and magicians"
}
{
"Tag": [
"factorial",
"function",
"calculus",
"integration",
"algebra",
"functional equation",
"calculus computations"
],
"Problem": "What is the process for proving the relationship of the gamma function and the factorial, starting with the factorial?",
"Solution_1": "Why would you start with the factorial? The thing to be proven is about the Gamma function, and it is usually done by integration by parts. Once you've established the functional equation $ \\Gamma(z \\plus{} 1) \\equal{} z \\Gamma (z)$, you're pretty much done.",
"Solution_2": "show that 'at integers', the gamma function (shifted by one) and the factorial satisfy the same recursion, and are equal at $ 0$. depending on your definition, the recursion may actually define the factorial.",
"Solution_3": "How do you prove that the Gamma function is the only (log convex) function with this property?",
"Solution_4": "[quote=\"0714446459923\"]How do you prove that the Gamma function is the only (log convex) function with this property?[/quote]\r\nThat is the so-called [url=http://en.wikipedia.org/wiki/Bohr-Mollerup_theorem]Bohr-Mollerup Theorem[/url] (proofs can be found using links on that page)."
}
{
"Tag": [
"number theory unsolved",
"number theory"
],
"Problem": "Find all natural numbers $n$ such that\r\n\r\n(a) $n$ is not a power of two\r\n\r\n(b) $2^n+1=p^m$ for some prime number $p$ and natural number $m$",
"Solution_1": "Only $2^3+1=3^2$ works.\r\n\r\nThis is what is needed for http://www.mathlinks.ro/Forum/viewtopic.php?t=82146 and can e.g. be proven by seeing that $p=2^k+1$, setting in and looking at the $p$-adic valuations."
}
{
"Tag": [
"abstract algebra",
"Ring Theory",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Let $ M$ be the $ \\mathbb{Z}[i]$-module generated by the elements $ v_1$, $ v_2$ such that $ (1\\plus{}i)v_1\\plus{}(2\\minus{}i)v_2\\equal{}0$ and $ 3v_1\\plus{}5iv_2\\equal{}0$. Find an integer $ r \\geq 0$ and a torsion $ \\mathbb{Z}[i]$-module $ T$ such that $ M \\cong \\mathbb{Z}[i]^r \\times T$.",
"Solution_1": "Are you familiar with how to compute [url=http://en.wikipedia.org/wiki/Smith_normal_form]Smith normal forms[/url] over a generic PID?",
"Solution_2": "Yes I am. However, I do not know how to use it in this problem. How do I use that here? If not that way, if you have any other recommendations for solving the problem, that would be great too.",
"Solution_3": "What is the Smith normal form of $ \\left[ \\begin{array}{cc} 1 \\plus{} i & 2 \\minus{} i \\\\ 3 & 5i \\end{array} \\right]$? What does that tell you about a good choice for $ T$?"
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Let x,y,z>0,s.t.x+y+z=1,and k is a specified nonnegative integer\r\nprove that:\r\n$ \\frac{x^{k\\plus{}2}}{{x^{k\\plus{}1}}\\plus{}{y^k}\\plus{}{z^k}}\\plus{}\\frac {y^{k\\plus{}2}}{{y^{k\\plus{}1}}\\plus{}{x^k}\\plus{}{z^k}}\\plus{}\\frac {z^{k\\plus{}2}}{{z^{k\\plus{}1}}\\plus{}{x^k}\\plus{}{y^k}}\\geq \\frac {1}{7}$",
"Solution_1": "posted before, It's serbia 2007\r\nsee here : http://www.mathlinks.ro/viewtopic.php?p=860869#p860869"
}
{
"Tag": [
"geometry",
"3D geometry",
"sphere",
"projective geometry",
"advanced fields",
"advanced fields unsolved"
],
"Problem": "Hello,\r\n\r\nI am looking for explicit formula for canonical (Riemannian) metric on the real projective space $ \\mathbb{RP}^n$. Could anyone give me some references or put this formula here?\r\n\r\nThere is well known Fubini-Study (Kahler) metric for complex projective space and I could not find something similar for real space.",
"Solution_1": "The same as the sphere since the covering map is a local isometry."
}
{
"Tag": [
"inequalities"
],
"Problem": "If a < b, show that a < (a + b)/2 < b. The number (a + b)/2 is called the ARITHMETIC MEAN of a and b.",
"Solution_1": "[quote=\"Interval\"]If a < b, show that a < (a + b)/2 < b. The number (a + b)/2 is called the ARITHMETIC MEAN of a and b.[/quote]\r\n\r\nOK, it is easy to show that the midpoint of two numbers is betweeen the two numbers.\r\n\r\nSo, this should be in Getting Started.",
"Solution_2": "Just multiply through by 2...yeah someone should move this.",
"Solution_3": "Can you provide me with the steps for proving this question.",
"Solution_4": "[quote=\"Interval\"]Can you provide me with the steps for proving this question.[/quote]Multiply through by 2:\r\n\r\n$2a < a+b < 2b$\r\n\r\nThen look at the two inequalities seperately:\r\n\r\n$2a < a+b \\Rightarrow a 0$.\r\n\r\nLastly, consider the function $ g(k) \\equal{} max(f(1),f(2),f(3),...,f(k))$. Then $ g(1) < 0$, $ g(m) \\ge f(m) > 0$, and since $ g(x \\plus{} 1)$ is always either 0 or 1 greater than $ g(x)$, $ g(x)$ must hit every integer between $ f(1)$ and $ f(m)$ at some integer k between 1 and m.\r\n\r\nThis implies there exists an $ f(k)$ s.t. $ f(k)\\equal{}0$, so we're done."
}
{
"Tag": [
"absolute value",
"wilkinson catalyst",
"P2 catalyst",
"birch reduction"
],
"Problem": "Set $A$ consists of $m$ consecutive integers whose sum is $2m$, and set $B$ consists of $2m$ consecutive integers whose sum is $m$. The absolute value of the difference between the greatest element of $A$ and the greatest element of $B$ is $99$. Find $m$.",
"Solution_1": "[hide=\"solution\"]\nlet: $A=\\{a+1,...,a+m\\}, B=\\{b+1,...,b+2m\\}$\n$S_a=ma+\\frac{m(m+1)}{2}=2m\\Rightarrow 2a+m+1=4(1)$\n$S_b=2mb+\\frac{2m(2m+1)}{2}=m\\Rightarrow 2b+2m=0(2)$\nalso we have: $b+m-a=99(3)$\n$\\Rightarrow (1,2) a=\\frac{3-m}{2}, b=-m$\n\nplugging these into (3) we get:\n$-3+m=198\\Rightarrow \\boxed{m=201}$[/hide]",
"Solution_2": "I got different answer... I checked it and seems correct..\r\n[hide]\nTelling from the numbers, some of them might be negative. So let $1-k$ be the first number that the consecutive integers start on. \nWe have sums to be:\n$\\frac{m^2+m}{2}-mk=2m$ and $\\frac{4m^2+2m}{2}-m(k+99)$\nSolving gives $\\boxed{m=65}$.[/hide]",
"Solution_3": "I got the same results as [b]amirhtlusa[/b].\r\n\r\n$S_A=S_m=\\frac{a_1+a_m}{2}\\cdot m = \\frac{a_1+a_1+m-1}{2}\\cdot m = 2m \\Rightarrow \\boxed{2a_1+m=5}$ [1]\r\n\r\n$S_B=S_{2m}=\\frac{b_1+b_{2m}}{2}\\cdot 2m = \\frac{b_1+b_1+2m-1}{2}\\cdot 2m = m \\Rightarrow \\boxed{b_1+m=1}$ [2]\r\n\r\n$|a_m-b_{2m}|=99 \\Rightarrow |a_1+m-1-(b_1+2m-1)|=99 \\Rightarrow \\boxed{m+b_1-a_1=99}$ [3]\r\n\r\nFrom [b][1][/b], [b][2][/b] and [b][3][/b], we have that:\r\n\r\n$\\boxed{a_1=-98; b_1=-200; m=201}$",
"Solution_4": "[quote=\"jimhu\"]I got different answer... I checked it and seems correct..\n[hide]\nTelling from the numbers, some of them might be negative. So let $1-k$ be the first number that the consecutive integers start on. \nWe have sums to be:\n$\\frac{m^2+m}{2}-mk=2m$ and $\\frac{4m^2+2m}{2}-m(k+99)$\nSolving gives $\\boxed{m=65}$.[/hide][/quote]\r\n\r\nBy the way, each set starts with a different number :!:",
"Solution_5": "[quote=\"Ezbakhe Yassin\"]I got the same results as [b]amirhtlusa[/b].\n\n$S_A=S_m=\\frac{a_1+a_m}{2}\\cdot m = \\frac{a_1+a_1+m-1}{2}\\cdot m = 2m \\Rightarrow \\boxed{2a_1+m=5}$ [1]\n\n$S_B=S_{2m}=\\frac{b_1+b_{2m}}{2}\\cdot 2m = \\frac{b_1+b_1+2m-1}{2}\\cdot 2m = m \\Rightarrow \\boxed{b_1+m=1}$ [2]\n\n$|a_m-b_{2m}|=99 \\Rightarrow |a_1+m-1-(b_1+2m-1)|=99 \\Rightarrow \\boxed{m+b_1-a_1=99}$ [3]\n\nFrom [b][1][/b], [b][2][/b] and [b][3][/b], we have that:\n\n$\\boxed{a_1=-98; b_1=-200; m=201}$[/quote]]\r\n\r\nHmm.. Strange. If we test my answer, we have 65 consecutive integers starting at -30. \r\nAnd then 130 consecutive integers starting at -129 .The sums should be: \r\n$\\frac{65^2+65}{2}-(65)(31)=130=2m$\r\n$\\frac{130^2+130}{2}-(65)(31+99)=130=65=m$[/hide]",
"Solution_6": "Could anyone try to solve it using another method?",
"Solution_7": "i just checked the answer with Kalva, 201 is correct",
"Solution_8": "It's the difference between the greatest elements that we're looking for, not the least elements. ;)",
"Solution_9": "[quote=\"JesusFreak197\"]It's the difference between the greatest elements that we're looking for, not the least elements. ;)[/quote]\r\n\r\nYes, but difference between least terms is still 99, therefore I used to my advantage...",
"Solution_10": "No, because each sequence has a different number of terms. ;)",
"Solution_11": "Sorry for the revive, but I wanted to post a cleaner solution.\n[hide=\"Work with the means.\"]\nThe mean of Set $A$ is $\\frac{2m}{m}=2$, so this is also the median. There are $\\frac{m-1}{2}$ elements greater than the median, so the greatest element in this set is $\\frac{m+3}{2}$.\nThe mean of Set $B$ is $\\frac{1}{2}$, so the middle two elements are $0$ and $1$. There are $m$ elements after $0$, so $m$ is the greatest element in set $B$. Since $m$ is positive, $m >\\frac{m+3}{2}$, and $m-\\frac{m+3}{2}=99$. Solving, we get $m=201$.\n[/hide]",
"Solution_12": "Wait, what is the source of this problem? I think I saw it before on AIME.",
"Solution_13": "2004 AIME 1, Problem 2.",
"Solution_14": "[quote=donot]Sorry for the revive, but I wanted to post a cleaner solution.\n[hide=\"Work with the means.\"]\nThe mean of Set $A$ is $\\frac{2m}{m}=2$, so this is also the median. There are $\\frac{m-1}{2}$ elements greater than the median, so the greatest element in this set is $\\frac{m+3}{2}$.\nThe mean of Set $B$ is $\\frac{1}{2}$, so the middle two elements are $0$ and $1$. There are $m$ elements after $0$, so $m$ is the greatest element in set $B$. Since $m$ is positive, $m >\\frac{m+3}{2}$, and $m-\\frac{m+3}{2}=99$. Solving, we get $m=201$.\n[/hide][/quote]\n\n\npls dont laugh if my doubt is trivial\n\nwhy m>(m+3)/2 ,it is thou not alwys true??\n\n",
"Solution_15": "\n[hide=@#13][url=aops.com/community/user/174289][b]daniellionyang[/b][/url] \u00b7 Jul 27, 2016, 7:47 PM [url=aops.com/community/p6725701](view)[/url][color=transparent]helo[/color]\nWait, what is the source of this problem? I think I saw it before on AIME.\n\n-----------\n[color=#5b7083][aops]x[/aops] 13[color=transparent]hellloolo[/color] [aops]Y[/aops] 0 [color=transparent]hellloolo[/color] [/hide]\nPosting this for contest collections.",
"Solution_16": "Let the first sequence be ${a+1, a+2, \\ldots a+m}$ and let the second sequence be ${b+1, b+2 \\ldots b+2m}$. Then $2a+m+1=4, 2b+2m+1=1$, so $2a+m=3, 2b+2m=0$. We also have $(b+2m)-(a+m)=99$, so $b+m-a=99$. Solving gives $m=\\boxed{201}$.",
"Solution_17": "What I did was set the smallest numbers of set A and B to\n$a-m+1$ and $b-2m+1$, respectively.\n\nThen I got the equations\n\n$b-2m=0$\n$2a-m=3$\n$b-a=99$\nWhich gives me an answer of $m=67$, but somehow im a third off? Where did I go wrong?"
}
{
"Tag": [
"LaTeX",
"quadratics",
"algebra unsolved",
"algebra"
],
"Problem": "Find all real x,y,z that \\left\\{\\begin{array}{c}x+y+zx=\\frac12\\\\ \\\\ y+z+xy=\\frac12\\\\ \\\\ z+x+yz=\\frac12\\end{array}\\right.",
"Solution_1": "You mean $ \\{\\begin{array}{c}x+y+zx=\\frac12\\\\ \\\\ y+z+xy=\\frac12\\\\ \\\\ z+x+yz=\\frac12\\end{array}$.",
"Solution_2": "place your latex codes between dollar signs",
"Solution_3": "tnx very much about your help",
"Solution_4": "We can easily show that all are positive. Subtract the eq in pairs to get: x-z=x(y-z) & y-x=y(z-x). Divide the two eq. & simplify to get: (y/z)+(x/y)+(z/x)=3. Apply A.M. G.M. to get x=y=z. Solve any eq. by the quadratic formula. Sorry for the inconvenience but my latex is not working.",
"Solution_5": "[quote=\"amitsaha\"]We can easily show that all are positive[/quote]\r\nMaybe easily, but wrong : $ (x,y,z)=(-1,2,\\frac 12)$ is a solution, for example.\r\n\r\n$ \\{\\begin{array}{c}x + y + zx = \\frac12 \\\\\r\n \\\\\r\ny + z + xy = \\frac12 \\\\\r\n \\\\\r\nz + x + yz = \\frac12\\end{array}$\r\n\r\n\r\nFirst equation gives $ y=\\frac 12-x-xz$\r\n\r\nPlugging this in the third equation, we get $ x(1-z-z^2)=\\frac{1-3z}2$ and so $ 1-z-z^2\\ne 0$ and $ x=\\frac{1-3z}{2(1-z-z^2)}$\r\n\r\nPlugging these values in the second equation, we get $ 4z^5+2z^4-20z^3-5z^2+11z-2=0$ which may be written $ (z-2)(z+1)(2z-1)(2z^2+4z-1)=0$\r\n\r\nAnd so $ 5$ solutions :\r\n$ (x,y,z)=(-1,2,\\frac 12)$\r\n\r\n$ (x,y,z)=(\\frac 12,-1,2)$\r\n\r\n$ (x,y,z)=(2,\\frac 12,-1)$\r\n\r\n$ (x,y,z)=(-1-\\sqrt{\\frac 32},-1-\\sqrt{\\frac 32},$ $ -1-\\sqrt{\\frac 32})$\r\n\r\n$ (x,y,z)=(-1+\\sqrt{\\frac 32},-1+\\sqrt{\\frac 32},$ $ -1+\\sqrt{\\frac 32})$",
"Solution_6": "tnx pco\r\n\u062e\u06cc\u0644\u06cc \u062a\u0648\u067e \u0628\u0648\u062f\r\nIt's in persian which is mean very very nice"
}
{
"Tag": [],
"Problem": "$ c\\in \\mathbb{R}$ and $ c \\equal{} \\frac {\\sqrt {a \\minus{} 2b} \\minus{} 3b}{4a \\minus{} \\sqrt {2b \\minus{} a}}$\r\n$ c \\plus{} \\frac {a}{b} \\equal{} ?$\r\n\r\n[hide]use that a square root in this equation must be real[/hide]",
"Solution_1": "[hide]$ a\\minus{}2b\\geq 0$ and $ 2b\\minus{}a \\geq 0$, thus $ a\\minus{}2b\\equal{}2b\\minus{}a\\equal{}0 \\implies a/b\\equal{}2$\n\nThen $ c \\equal{} \\minus{} \\frac{3b}{4a} \\equal{} \\minus{} \\frac{3}{8}$\n\nThen $ c\\plus{}a/b\\equal{}13/8$[/hide]",
"Solution_2": "What if $ a,b$ are not neccessarily reals? Then how do you know that both $ \\sqrt{2b\\minus{}a}$ and $ \\sqrt{a\\minus{}2b}$ must be reals?",
"Solution_3": "If a-2b and b-2a are real numbers, then a,b are real",
"Solution_4": "I don't see why $ a\\minus{}2b$ and $ b\\minus{}2a$ must be reals.",
"Solution_5": "[quote=\"cosinator\"]I don't see why $ a \\minus{} 2b$ and $ b \\minus{} 2a$ must be reals.[/quote]\r\nThe OP said \"use that a square root in this equation must be real \", but it was hidden. (not sure why...)",
"Solution_6": "[b]Altheman [/b]read it wrong, it was a-2b and 2b-a. And yes, there's no good reason why they should have to be real, except that it's given (check [b]queen's [/b]hidden content). :wink: \r\n\r\nBut what [b]Altheman[/b] said, you can say that let x = Im(a) and y=Im(b). And in that case, you know that x-2y = 0 and y-2x=0. Solving the system gives (x,y)=(0,0), so a and b must be real.",
"Solution_7": "Zuton force, I'm not convinced. You are assuming that the result of the square roots must be real, but in reality just the expression must be real.\r\n\r\nThinking... $ c \\in \\mathbb{R} \\implies arg(\\sqrt{a\\minus{}2b}\\minus{}3b) \\equal{} arg(4a\\minus{}\\sqrt{2b\\minus{}a}) \\implies \\sqrt{a\\minus{}2b}\\minus{}3b \\equal{} k(4a\\minus{}\\sqrt{2b\\minus{}a})$ for some $ k \\in \\mathbb(R)$. Hence we have $ 4ak\\plus{}3b \\equal{} \\sqrt{a\\minus{}2b}\\plus{}k\\sqrt{2b\\minus{}a}$.\r\n\r\nI don't know where to go from there, but thinking about it now, $ k\\equal{}c$...\r\n\r\nWait... did the problem specify that the square roots must give reals????",
"Solution_8": "we dont want a complex value here so that is why we say $ c\\in R$",
"Solution_9": "I know $ c$ is real, what I'm asking is does that imply that $ \\sqrt{2b\\minus{}a}$ and $ \\sqrt{a\\minus{}2b}$ are necessarily also reals, or could imaginary parts end up canceling later?",
"Solution_10": "[hide]Since $ c \\in \\mathbb{R}$, $ a\\minus{}2b \\ge 0$ and $ 2b\\minus{}a \\ge 0 \\implies a\\minus{}2b \\le 0$, which implies that $ \\displaystyle a\\minus{}2b\\equal{}2b\\minus{}a\\equal{}0 \\implies \\frac{a}{b}\\equal{}2$. Then:\n\\[ c\\plus{}\\frac{a}{b}\\equal{}\\frac{\\minus{}3b}{4a}\\plus{}\\frac{a}{b}\\equal{}\\minus{}\\frac{3}{4} \\cdot \\frac{b}{a} \\plus{}\\frac{a}{b}\\equal{}\\minus{}\\frac{3}{8}\\plus{}2\\equal{}\\boxed{\\frac{13}{8}}\\][/hide]",
"Solution_11": "123456789, I would completely agree if we were given $ a$ and $ b$ are real. However, isn't it possible for $ c$ to be real if $ a$ and $ b$ are complex? Couldn't the imaginary parts cancel in the expression?",
"Solution_12": "It says that $ c$ is a real number, so if the denominator has an imaginary term, the numerator won't, and it won't cancel out.",
"Solution_13": "are you assuming that $ a$ and $ b$ are real? Because your explanation makes it sound like you are.",
"Solution_14": "[quote=\"facis\"]$ 4ak \\plus{} 3b \\equal{} \\sqrt {a \\minus{} 2b} \\plus{} k\\sqrt {2b \\minus{} a}$[/quote]\r\n\r\n$ 4ak \\plus{} 3b \\equal{} (1 \\plus{} ki)\\sqrt {a \\minus{} 2b}$\r\n$ (4ak\\plus{}3b)^2\\equal{}(a\\minus{}2b)(1\\plus{}ki)^2$\r\n\r\nhmmm......",
"Solution_15": "how can you cancel complex numbers when only one of the root becomes complex?",
"Solution_16": "[quote=\"queen\"]how can you cancel complex numbers when only one of the root becomes complex?[/quote]\r\n\r\nHow do you know only one of the roots is complex?\r\n\r\nFor real $ a$ and $ b$, it's obvious because $ \\sqrt{a\\minus{}2b}\\equal{}\\sqrt{\\minus{}(2b\\minus{}a)}\\equal{}i\\sqrt{2b\\minus{}a}$. But the problem doesn't say $ a$ and $ b$ have to be real, only that $ c$ is real. For example, if $ \\sqrt{a\\minus{}2b}\\equal{}1\\plus{}i$, $ \\sqrt{2b\\minus{}a}\\equal{}i\\minus{}1$, and they're both complex.\r\n\r\nThe question is whether you can find a pair of complex values of $ a$ and $ b$ that will make both the numerator and denominator complex, in such a way that they cancel out and produce a real quotient."
}
{
"Tag": [
"rotation"
],
"Problem": "how many of the 26 capital letters are symmetrical with respect to a point?",
"Solution_1": "I might be misunderstanding the question but...\r\n\r\n[hide=\"my \"not sure\" answer\"] 26(26-3) = 598\n\numm.. try not to trust me.. the wording was hard for me to understand...[/hide]",
"Solution_2": "Hint: The answer is less than 26\r\n\r\n\r\nI don't believe there is a formula to evaluate this.. you either have to know if off hand, or consider each letter one by one.",
"Solution_3": "[hide]\nUhh, I get 7, I'm not sure though.[/hide]",
"Solution_4": "[hide]Let's just type the alphabet and see.... ABCDEFGHIJKLMNOPQRSTUVWXYZ\n\nOK. We have A, B, C, D, E, H, I, K, M, O, T, U, V, W, X, and Y. \n\n(That is if there's any place that you can fold the figure with a line of symmetry. otherwise, I don't know.)\n\nIf so, then I get 16. [/hide]",
"Solution_5": "wat does symmetric w.r.t a point mean??",
"Solution_6": "with respect to a point",
"Solution_7": "I think he means what does symmetric with respect to a point actually mean. Does it mean its the same if rotated 180 degrees about a point? If so, then I see I, O, and X. ah, and H, N, S, and Z? I'll say 7.",
"Solution_8": "if point djshowdown2's result is right\r\nif line elementsofmathkid's result is right"
}
{
"Tag": [
"algebra",
"binomial theorem",
"number theory proposed",
"number theory"
],
"Problem": "Fined all $ k,n \\in N$, that $ k^{n}+2^{n}+1\\mid k^{n+1}+2^{n+1}+1$",
"Solution_1": "I'm assuming $ \\mathbb{N}= \\{1, 2, \\ldots\\}$ and $ \\mathbb{N}_{0}= \\mathbb{N}\\cup \\{0\\}$. In the first place, recall that $ \\gcd(2^{a}+1, 2^{b}+1) = 1$, whenever $ a, b \\in \\mathbb{N}_{0}$ and $ \\gcd(a,b) = 1$. Then let $ (n,k) \\in \\mathbb{N}^{\\,2}$ such that\r\n\\[ (k^{n}+2^{n}+1) \\mid (k^{n+1}+2^{n+1}+1)\\quad\\qquad (1) \\]\r\nSo actually, taking $ k = 1$ would entail $ (2^{n-1}+1) \\mid (2^{n}+1)$, whereas settling $ k = 2$ would imply $ (2^{n+1}+1) \\mid (2^{n+2}+1)$ (absurd in both cases). On the other hand, when fixing $ n = 1$, it's clear that (1) leads to have $ (k+3) \\mid (6k+4)$, i.e. $ (k+3) \\mid 14$, which is possible iff $ k \\in \\mathbb{N}$ and $ (k+3) \\in \\{1, 2, 7, 14\\}$, viz iff $ k = 4$ or $ k = 11$. Hence from here on going we can admit $ k \\ge 3$ and $ n \\ge 2$. Now $ k^{n+1}+2^{n+1}+1 = k\\cdot (k^{n}+2^{n}+1)-2^{n}(k-2)-k+1$. Thus from (1)\r\n\\[ \\frac{2^{n}(k-2)+k-1}{k^{n}+2^{n}+1}\\in\\mathbb{N}\\quad\\Longrightarrow\\quad k^{n}\\le 2^{n}(k-3)+k-2\\quad\\qquad (2) \\]\r\nthat is impossible whether $ k \\ge 3$ and $ n \\ge 2$, since then (through Newton's binomial theorem)\r\n\\[ k^{n}\\ge (k-1)^{n}+2n (k-1)^{n-1}+2n(n+1) (k-1)^{n-2}> 2^{n}\\cdot (k-3)+k-2. \\]\r\nThus there are no other solutions, apart from those we have already detected. :wink:",
"Solution_2": "Great!\r\nSo there are only two solutions: $ (k,n)=((4,1),(11,1))$.",
"Solution_3": ":) Indeed, I did a few typos, that I'm just going to fix."
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "I have found an interesting one by running test on computer program and get:\r\nIf $a,b,c$ is nonnegative and $a+b+c=3$ then\r\n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq ab+bc+ca$\r\n But I don't know how to prove it. Can anyone do it?",
"Solution_1": "[EDIT: This was nonsense.]",
"Solution_2": "If I am right that it is a weak one :lol: \r\nin fact we can prove \r\n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq 3$\r\nWlog $a \\ge b \\ge c$\r\nwe have\r\n$3-bc \\ge 3-ac \\ge 3-ba$\r\n$\\frac 1{a+1} \\ge \\frac 1{b+1} \\ge \\frac 1{c+1}$\r\nso we get:$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\ge \\frac 1 3(9-ab-bc-ca)(\\frac 1{a+1}+\\frac 1{b+1} +\\frac 1{c+1})$\r\nit is easy to prove $9-ab-bc-ca \\ge 6$\r\nand $\\frac 1{a+1}+\\frac 1{b+1} +\\frac 1{c+1} \\ge \\frac 3 2$(caughy)\r\n :)",
"Solution_3": "[quote=\"zhaobin\"]If I am right that it is a weak one :lol: \nin fact we can prove \n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq 3$\nWlog $a \\ge b \\ge c$\nwe have\n$\\frac 1{a+1} \\ge \\frac 1{b+1} \\ge \\frac 1{c+1}$\n\n :)[/quote]\r\n\r\nYes, you use Trebusep but if $a>b>c$ then $\\frac 1{a+1} \\ge \\frac 1{b+1} \\ge \\frac 1{c+1}$ is wrong ! Right? :D Then you prove the weaker inequality :D",
"Solution_4": "Oops... :(",
"Solution_5": "ok,Here goes my solution again(for your orginal one),hope it is right.\r\nbut unfortunatly it is ugly.\r\nlet $x=a^2+b^2+c^2,y=ab+bc+ca$\r\nand notice $f(t)=\\frac 1 {1+t}$ is convex $t>0$\r\nso we have \r\n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq (9-y) \\frac 1 {1+\\frac{9-3abc}{9-y}}$\r\nso we need to prove:\r\n$(9-y)^2 \\ge y(9-y+9-3abc)$\r\nbecause $-3abc \\le a^2+b^2+c^2-2ab-2bc-2ca=x-2y$(schur and a+b+c=3)\r\nand $9-y=x+y$\r\nwe need to prove \r\n$(x+y)^2 \\ge y(x+y+x+2y+x-2y)$\r\nequal to\r\n$(x+y)^2 \\ge y(3x+y)$\r\n$(x+2y)x \\ge 3xy$\r\n$3 \\ge y$\r\nand it is proved :)",
"Solution_6": "Strange. I couldn't belived that it can be prove by AM-GM. Indeed, I have to use Schur one times and expand... for very long time. \r\n I will check your sols again. But I hope you edit your sols first because it is difficult to read. Anyway, thank you for your effort! :lol:",
"Solution_7": "[quote=\"zhaobin\"]ok,Here goes my solution again(for your orginal one),hope it is right.\nbut unfortunatly it is ugly.\nlet $x=a^2+b^2+c^2,y=ab+bc+ca$\nand notice $f(t)=\\frac 1 {1+t}$ is convex $t>0$\nso we have \n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq (9-y) \\frac 1 {1+\\frac{9-3abc}{9-y}}$\n[/quote]\r\nOk! I have checked your sols. already. I can prove the rest of yours but I don't understand how can you use Jense here. Can u solve me where is your \"t\" in your sols and how can u obtain that form. Thank u!",
"Solution_8": "[quote=\"Gibbenergy\"][quote=\"zhaobin\"]ok,Here goes my solution again(for your orginal one),hope it is right.\nbut unfortunatly it is ugly.\nlet $x=a^2+b^2+c^2,y=ab+bc+ca$\nand notice $f(t)=\\frac 1 {1+t}$ is convex $t>0$\nso we have \n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq (9-y) \\frac 1 {1+\\frac{9-3abc}{9-y}}$\n[/quote]\nOk! I have checked your sols. already. I can prove the rest of yours but I don't understand how can you use Jense here. Can u solve me where is your \"t\" in your sols and how can u obtain that form. Thank u![/quote]\r\nok,because $f(t)=\\frac 1 {1+t}$ is convex $t>0$\r\nwe have \r\n$\\frac{3-bc}{9-y} f(a)+ \\frac{3-ca}{9-y} f(b)+ \\frac{3-ab}{9-y} f(c) \\ge f(\\frac{a(3-bc)}{9-y}+\\frac{b(3-ca)}{9-y}+\\frac{c(3-ab)}{9-y})$\r\nthis is because $\\frac{3-bc}{9-y}+ \\frac{3-ca}{9-y}+\\frac{3-ab}{9-y}=1$\r\nthus we get:\r\n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq (9-y) \\frac 1 {1+\\frac{9-3abc}{9-y}}$\r\nHope it is clear now.\r\nthanks if you like my solution :P",
"Solution_9": "[quote=\"zhaobin\"]\nok,because $f(t)=\\frac 1 {1+t}$ is convex $t>0$\nwe have \n$\\frac{3-bc}{9-y} f(a)+ \\frac{3-ca}{9-y} f(b)+ \\frac{3-ab}{9-y} f(c) \\ge f(\\frac{a(3-bc)}{9-y}+\\frac{b(3-ca)}{9-y}+\\frac{c(3-ab)}{9-y})$\nthis is because $\\frac{3-bc}{9-y}+ \\frac{3-ca}{9-y}+\\frac{3-ab}{9-y}=1$\nthus we get:\n$\\frac{3-bc}{a+1}+\\frac{3-ac}{b+1}+\\frac{3-ba}{c+1} \\geq (9-y) \\frac 1 {1+\\frac{9-3abc}{9-y}}$\nHope it is clear now.\nthanks if you like my solution :P[/quote]\r\nOk. It is look very neat and logic. I think you are successful now. And you are amazing. Your sols is better than mine a lots. Anyway, congragulation!",
"Solution_10": "I have just found a different solution to this beautiful inequality.\r\n\r\n[color=blue][b]Problem.[/b] Let a, b, c be three positive reals satisfying a + b + c = 3. Prove that\n\n$\\frac{3-bc}{a+1}+\\frac{3-ca}{b+1}+\\frac{3-ab}{c+1}\\geq bc+ca+ab$.[/color]\r\n\r\n[i]Solution.[/i] Using the notation $\\sum$ for cyclic sums, this inequality rewrites as $\\sum\\frac{3-bc}{a+1}\\geq bc+ca+ab$. Now, we will homogenize this inequality by noting that\r\n\r\n$\\frac{3-bc}{a+1}=\\frac{3^2-3bc}{3a+3}=\\frac{\\left(a+b+c\\right)^2-3bc}{3a+\\left(a+b+c\\right)}=\\frac{a^2+b^2+c^2+2ca+2ab-bc}{4a+b+c}$\r\n\r\nand $bc+ca+ab=3\\frac{bc+ca+ab}{3}=3\\frac{bc+ca+ab}{a+b+c}$, so the inequality in question takes the homogeneous form\r\n\r\n$\\sum\\frac{a^2+b^2+c^2+2ca+2ab-bc}{4a+b+c}\\geq 3\\frac{bc+ca+ab}{a+b+c}$.\r\n\r\nNow, since $b^2+c^2\\geq 2bc$, we have\r\n\r\n$a^2+b^2+c^2+2ca+2ab-bc\\geq a^2+2bc+2ca+2ab-bc=a^2+2ca+2ab+bc$\r\n$=a\\left(a+b+c\\right)+\\left(bc+ca+ab\\right)$,\r\n\r\nso that it remains to prove the (stronger) inequality\r\n\r\n$\\sum\\frac{a\\left(a+b+c\\right)+\\left(bc+ca+ab\\right)}{4a+b+c}\\geq 3\\frac{bc+ca+ab}{a+b+c}$.\r\n\r\nThis rewrites as\r\n\r\n$\\sum\\frac{a\\left(a+b+c\\right)}{4a+b+c}+\\sum\\frac{bc+ca+ab}{4a+b+c}\\geq 3\\frac{bc+ca+ab}{a+b+c}$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)\\sum\\frac{a}{4a+b+c}+\\left(bc+ca+ab\\right)\\sum\\frac{1}{4a+b+c}$\r\n$\\geq \\left(bc+ca+ab\\right)\\cdot\\frac{3}{a+b+c}$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)\\sum\\frac{a}{4a+b+c}\\geq\\left(bc+ca+ab\\right)\\left(\\frac{3}{a+b+c}-\\sum\\frac{1}{4a+b+c}\\right)$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)\\sum\\frac{a}{4a+b+c}\\geq\\left(bc+ca+ab\\right)\\sum\\left(\\frac{1}{a+b+c}-\\frac{1}{4a+b+c}\\right)$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)\\sum\\frac{a}{4a+b+c}\\geq\\left(bc+ca+ab\\right)\\sum\\frac{3a}{\\left(a+b+c\\right)\\left(4a+b+c\\right)}$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)\\sum\\frac{a}{4a+b+c}\\geq\\frac{3\\left(bc+ca+ab\\right)}{a+b+c}\\sum\\frac{a}{4a+b+c}$\r\n$\\Longleftrightarrow\\ \\ \\ \\ \\ a+b+c\\geq\\frac{3\\left(bc+ca+ab\\right)}{a+b+c}\\ \\ \\ \\ \\ \\Longleftrightarrow\\ \\ \\ \\ \\ \\left(a+b+c\\right)^2\\geq 3\\left(bc+ca+ab\\right)$.\r\n\r\nBut this inequality is obvious from\r\n\r\n$\\left(a+b+c\\right)^2-3\\left(bc+ca+ab\\right)=\\frac12\\left(\\left(b-c\\right)^2+\\left(c-a\\right)^2+\\left(a-b\\right)^2\\right)\\geq 0$.\r\n\r\nThus, the problem is solved.\r\n\r\n Darij"
}
{
"Tag": [
"algebra",
"polynomial",
"Vieta"
],
"Problem": "P(X) is a monic polynomial of degree 4 with integer co-efficients and constant term 4. One zero is $ \\sqrt2$, another zero is rational and the sum of the zeroes is positive. Factorise P(x) fully over $ R$.\r\n\r\nTwo of the roots of $ 2x^{4}\\minus{}5x^{3}\\plus{}cx^{2}\\plus{}dx\\minus{}18\\equal{}0$ are reciprocals, while the remaining roots are opposites of each other. Find the four roots of the equation and evaluate c and d\r\n\r\n[hide=\"my soln. for qn 2 but don't understand a certain step\"]\nLet the roots be $ \\alpha.\\frac{1}{\\alpha},\\beta,\\minus{}\\beta$\nproduct of the roots , simplifying etc,\n$ a\\equal{}\\frac{1}{2},2$ and $ \\b\\equal{}\\pm3.$ I can write this in factored form i.e. $ (2x\\minus{}1)(x\\minus{}2)...$ expand and equate similar degree but the book has another way\n\nusing sum and product 2 at a time, $ 1\\minus{}\\beta^{2}\\equal{}\\minus{}8\\equal{}\\Rightarrow c\\equal{}\\minus{}16$\nand sum of product 3 at a time,=$ \\frac{\\minus{}45}{2}\\equal{}\\equal{}\\Rightarrow d\\equal{}\\minus{}45$\n\nWhy are the multiplying by 2[/hide]",
"Solution_1": "[hide=\"hint for #1\"]if the polynomial has integer coefficients and one of its roots is rational we can conclude that $ \\minus{}\\sqrt{2}$ is a root. Giving us the factorized form of $ (x\\minus{}\\sqrt{2})(x\\plus{}\\sqrt{2})(x^{2}\\plus{}cx\\minus{}2)$ [/hide]",
"Solution_2": "but $ \\minus{}\\sqrt2$ is not rational :S\r\n\r\ndid you take $ \\minus{}\\sqrt2$ as a zero because the qn says integer coeffecients are rational? \r\n\r\nthanks for the hint appreciate it\r\n\r\nany takers on number 2?",
"Solution_3": "[quote=\"bos1234\"]but $ \\minus{}\\sqrt2$ is not rational :S\n\ndid you take $ \\minus{}\\sqrt2$ as a zero because the qn says integer coeffecients are rational? [/quote]\r\n\r\nYes. This is the [b]conjugate root theorem[/b] (for algebraic roots over rational coefficients). Specifically, given rationals $ a, b, c$ such that $ a\\plus{}b\\sqrt{c}$ is a root, $ a\\minus{}b\\sqrt{c}$ is also a root. I give a proof of this in my blog.",
"Solution_4": "I'm not sure I understand where you have trouble with on #2... is it just this last part?\r\n\r\n[quote=\"bos1234\"]\nand sum of product 3 at a time,=$ \\frac{\\minus{}45}{2}\\equal{} \\equal{}\\Rightarrow d \\equal{}\\minus{}45$\n\nWhy are the multiplying by 2[/quote]\r\n\r\nWell they multiplied it by 2 because the polynomial isnt monic -- the first coefficient is a $ 2$. So by vieta's, $ \\frac{\\minus{}d}{2}\\equal{}\\frac{\\minus{}45}{2}$, so $ d\\equal{}45$, supposedly. Im not sure why the book has it negative, though (unless I did something wrong?)",
"Solution_5": "[quote=\"t0rajir0u\"][quote=\"bos1234\"]but $ \\minus{}\\sqrt2$ is not rational :S\n\ndid you take $ \\minus{}\\sqrt2$ as a zero because the qn says integer coeffecients are rational? [/quote]\n\nYes. This is the [b]conjugate root theorem[/b] (for algebraic roots over rational coefficients). Specifically, given rationals $ a, b, c$ such that $ a\\plus{}b\\sqrt{c}$ is a root, $ a\\minus{}b\\sqrt{c}$ is also a root. I give a proof of this in my blog.[/quote]\r\n\r\n@bos1234\r\n\r\nThe 2001 HSC Extension 2 paper has a question based on the Conjugate-Root Theorem (for algebraic roots over rational coefficients).",
"Solution_6": "ok thanks\r\nill check it out"
}
{
"Tag": [
"calculus",
"integration",
"combinatorics unsolved",
"combinatorics"
],
"Problem": "Given a positive integer $ n$, calculate: $ \\sum_{k \\equal{} 1}^n\\left( \\minus{} 1\\right)^k\\left(n \\plus{} k\\right)!\\left(n \\minus{} k\\right)!$.",
"Solution_1": "[quote=\"keira_khtn\"]Given a positive integer $ n$, calculate: $ \\sum_{k \\equal{} 1}^n\\left( \\minus{} 1\\right)^k\\left(n \\plus{} k\\right)!\\left(n \\minus{} k\\right)!$.[/quote]\r\nDenote $ S_n: \\equal{} \\sum\\limits_{k \\equal{} 1}^n( \\minus{} 1)^k(n \\plus{} k)!(n \\minus{} k)!\\;$ and\r\n\\[ \\begin{array}{|c|} \\hline L_n: \\equal{} (n!)^2 \\plus{} S_n \\equal{} \\sum\\limits_{k \\equal{} 0}^n( \\minus{} 1)^k(n \\plus{} k)!(n \\minus{} k)! \\; \\; \\; ,\\; \\; \\; \\; R_n: \\equal{} \\frac {1}{2}\\left((n!)^2 \\plus{} ( \\minus{} 1)^n\\frac {(2n \\plus{} 1)!}{n \\plus{} 1}\\right) \\\\\r\n\\hline\\end{array}\\; .\r\n\\]\r\n[color=red][b] Prove [/b][/color] $ \\begin{array}{|c|}\\hline L_n \\equal{} R_n \\\\\r\n\\hline \\end{array}\\; .$\r\nHowever $ L_0 \\equal{} R_0 \\equal{} 1\\; ,\\; L_1 \\equal{} R_1 \\equal{} \\minus{} 1 \\; ,\\; L_2 \\equal{} R_2 \\equal{} 22\\; ,... .$ Find a recurrence satisfied by both sequences $ (L_n) \\; , \\; (R_n)\\; .$ \r\nOr, use WZ-method (Wilf-Zeilberger $ \\approx$ (1990) , in fact [b]Markoff -method $ \\approx$ (1890) [/b] .\r\nUsing Integral Calculus it's possible to find [b]a proof[/b] in following way: Denote by $ \\Gamma$ the Gamma , and by $ B(\\cdot ,\\cdot )$ the Beta functions. We give\r\n\\[ \\begin{array}{c} (n \\minus{} k)!(n \\plus{} k)! \\equal{} \\Gamma(n \\minus{} k \\plus{} 1)\\Gamma(n \\plus{} k \\plus{} 1) \\equal{} B(n \\minus{} k \\plus{} 1,n \\plus{} k \\plus{} 1)(2n \\plus{} 1)! \\equal{} \\\\\r\n\\equal{} (2n \\plus{} 1)! \\int\\limits_{0}^{1}t^{n \\minus{} k}(1 \\minus{} t)^{n \\plus{} k}\\; dt \\equal{} ( \\minus{} 1)^k (2n \\plus{} 1)! \\int\\limits_{0}^{1}t^n(1 \\minus{} t)^n\\left(\\frac {t \\minus{} 1}{t}\\right)^k \\; dt\\; . \\end{array}\r\n\\]\r\nTherefore , using equality $ B(p,q) \\equal{} \\int\\limits_{0}^{1}t^{p \\minus{} 1}(1 \\minus{} t)^{q \\minus{} 1}\\; dt\\; ,\\; ( p > 0,q > 0),$ one finds\r\n\\[ \\begin{array}{c} S_n: \\equal{} \\sum\\limits_{k \\equal{} 1}^{n}( \\minus{} 1)^k(n \\minus{} k)!(n \\plus{} k)! \\equal{} \\sum\\limits_{k \\equal{} 0}^{n}( \\minus{} 1)^k(n \\minus{} k)!(n \\plus{} k)! \\minus{} (n!)^2 \\equal{} \\\\\r\n\\equal{} (2n \\plus{} 1)!\\int\\limits_{0}^{1}t^n(1 \\minus{} t)^n \\sum\\limits_{k \\equal{} 0}^{n}\\left(\\frac {t \\minus{} 1}{t}\\right)^k \\; dt \\minus{} (n!)^2 \\equal{} \\\\\r\n\\equal{} (2n \\plus{} 1)! \\int\\limits_{0}^{1}(1 \\minus{} t)^n \\left( t^{n \\plus{} 1} \\plus{} ( \\minus{} 1)^n (1 \\minus{} t)^{n \\plus{} 1}\\right)\\; dt \\minus{} (n!)^2 \\equal{} \\\\\r\n\\equal{} \\displaystyle \\frac { n!(n \\plus{} 1)! \\plus{} ( \\minus{} 1)^n(2n \\plus{} 1)! }{2n \\plus{} 2} \\minus{} (n!)^2 \\equal{} \\displaystyle \\frac { ( \\minus{} 1)^n(2n \\plus{} 1)! \\minus{} n!(n \\plus{} 1)!}{2n \\plus{} 2}\\; , \\end{array}\r\n\\]\r\nwhich completes the solution."
}
{
"Tag": [
"MATHCOUNTS"
],
"Problem": "A bucket is put under two faucets. If one faucet is turned on alone, the bucket will be filled in 3 minutes; if the other is turned on, the bucket will be filled in 2 minutes. If both are turned on, how many seconds will it take to fill the bucket?",
"Solution_1": "[hide=\"Click here to reveal hidden content\"]\n\nOne faucet can fill the 1/3 of the bucket, and the other fills 1/2 of the bucket in a minute. Both on can fill 5/6 of the bucket in a minute, so 6/5 minutes can fill the bucket. That is [b]1 minute, 12 seconds.[/b][/hide]",
"Solution_2": "[quote=\"steve123456\"][hide=\"Click here to reveal hidden content\"]\n\nOne faucet can fill the 1/3 of the bucket, and the other fills 1/2 of the bucket in a minute. Both on can fill 5/6 of the bucket in a minute, so 6/5 minutes can fill the bucket. That is [b]1 minute, 12 seconds.[/b][/hide][/quote]\r\n\r\nJust a clarification - the question asked for how many seconds, so the answer is actually 72 seconds. Again, though a minor point, it can cost a full question on competitions like Mathcounts, where they're looking for answers in a specific form. :)"
}
{
"Tag": [
"function",
"linear algebra",
"matrix",
"vector",
"linear algebra unsolved"
],
"Problem": "Let $x_{i}$ and $x'_{i}$ in $\\mathbb{R}^{n}$ for $i=1,...,k\\leq n$ have $\\parallel x_{i}-x_{j}\\parallel \\geq \\parallel x'_{i}-x'_{j}\\parallel$ then exists a continous family of k-tuples $x_{i}^{t}$,such that $x_{i}^{0}=x_{i}$ and $x_{i}^{1}=x'_{i}$ and $\\parallel x_{i}^{t}-x_{j}^{t}\\parallel$ is decreasing in $t$.",
"Solution_1": "I don't understand your problem. Does $x_{n}^{t}$ mean $(x_{n})^{t}$ ?",
"Solution_2": "I'd write those as $x_{i}(t)$ myself: we want a continuous function $x(t)=\\begin{bmatrix}x_{1}(t)\\\\ x_{2}(t)\\\\ \\vdots\\\\ x_{n}(t)\\end{bmatrix}$ such that $x(0)=x$ and $x(1)=x'$, with the additional property that $\\|x_{i}(t)-x_{j}(t)\\|$ is a decreasing function for each $i,j$.\r\n\r\nAt least, that's what I think the problem is- the notation is very confusing. $x,x',x(t)$ are vectors of vectors."
}
{
"Tag": [
"calculus",
"logarithms",
"derivative",
"function"
],
"Problem": "1. Which is larger: [tex]\\sqrt{n+1}^\\sqrt{n}\r\n[/tex] or [tex]\\sqrt{n}^\\sqrt{n+1}[/tex] for n>8",
"Solution_1": "[hide]a[/hide]",
"Solution_2": "Forgot to Mention Without a Calculator",
"Solution_3": "I think you ought to try to prove it with calculus. Showing that it works for n=9 doesn't prove it works for n=10 or indeed any other value.",
"Solution_4": "Calc stands for calculus, not calculator.... That did seems suspiciously easy. I'm going to leave now. My apologies.",
"Solution_5": "[tex]\\displaystyle \\sqrt{n+1}^{\\sqrt{n}}\\quad ?\\quad\\sqrt{n}^{\\sqrt{n+1}}[/tex]\r\n\r\n[tex]\\displaystyle \\sqrt{n} \\ln \\sqrt{n+1}\\quad? \\quad\\sqrt{n+1} \\ln \\sqrt{n}[/tex]\r\n\r\n[tex]\\displaystyle \\frac{\\ln \\sqrt{n+1}}{\\sqrt{n+1}} \\quad?\\quad \\frac{\\ln \\sqrt{n}}{\\sqrt{n}}[/tex]\r\n\r\n[tex]\\displaystyle f(a)=\\frac{\\ln a}{a}[/tex]\r\n\r\n[tex]\\displaystyle f'(a)=\\frac{1-\\ln a}{a^2}[/tex]\r\n\r\n[tex]f'(a)[/tex] is negative, i.e. [tex]f(a)[/tex] is decreasing, when [tex]a>e[/tex]. Since [tex]n \\ge 8[/tex], [tex]\\sqrt{n} > e[/tex], so [tex]f(a)[/tex] is decreasing, so ? is [tex]\\le[/tex].",
"Solution_6": "[quote=\"zabelman\"][tex]\\displaystyle \\sqrt{n+1}^{\\sqrt{n}}\\quad ?\\quad\\sqrt{n}^{\\sqrt{n+1}}[/tex]\n\n[tex]\\displaystyle \\sqrt{n} \\ln \\sqrt{n+1}\\quad? \\quad\\sqrt{n+1} \\ln \\sqrt{n}[/tex]\n\n[tex]\\displaystyle \\frac{\\ln \\sqrt{n+1}}{\\sqrt{n+1}} \\quad?\\quad \\frac{\\ln \\sqrt{n}}{\\sqrt{n}}[/tex]\n\n[tex]\\displaystyle f(a)=\\frac{\\ln a}{a}[/tex]\n\n[tex]\\displaystyle f'(a)=\\frac{1-\\ln a}{a^2}[/tex]\n\n[tex]f'(a)[/tex] is negative, i.e. [tex]f(a)[/tex] is decreasing, when [tex]a>e[/tex]. Since [tex]n \\ge 8[/tex], [tex]\\sqrt{n} > e[/tex], so [tex]f(a)[/tex] is decreasing, so ? is [tex]\\le[/tex].[/quote]\r\n\r\nI think you left more to be desired with this... You have shown that they are BOTH decreasing, but that doesn't mean that one is greater or equal to the other... and at what rate are both of them decreasing, does one decrease faster/more slowly than the other?\r\n\r\nAn also, you made the substitution [tex]\\displaystyle{a = \\sqrt{n}}[/tex] and [tex]\\displaystyle{a = \\sqrt{n+1}}[/tex], which is fine, but you must also remember to apply [tex]\\displaystyle{\\; da}[/tex] at each differentiation, because it will change your result...",
"Solution_7": "I was saying that since f(a) is decreasing, [tex]\\sqrt{n}<\\sqrt{n+1}[/tex] implies [tex]f(\\sqrt{n})>f(\\sqrt{n+1})[/tex], which is the basic meaning of \"decreasing.\"",
"Solution_8": "And, correct me if I'm wrong, f'(a) means the derivative of f(a) with respect to a. You would be multiplying by da/da which is 1.",
"Solution_9": "zableman, you are correct, i'm sorry. I misunderstood what you wrote.\r\n\r\njuggle, no....\r\n\r\nhe made a substitution g(n) = n + 1\r\n\r\nso the derivative of f(g(x)) must use the chain rule. It is the same for f(a) where a represents an equation.",
"Solution_10": "Yeah, I guess f'(a) does not necessarily mean the derivative of f(a) with respect to a. Still, let's assume we are differentiating with respect to a nonetheless.\r\n\r\nI think we agree up to the last sentence. You are saying because we are using n, we need to differentiate with respect to n. But, I think zabelman used f(a) for an explanatory cause so we can look at the original piece with more insight. He said if a>e, then f(a) is decreasing. If a = sqrt(n), then a is in fact always greater than e.\r\n\r\nIn other words, zabelman showed that the function ln(a)/a is strictly decreasing for a>e. Which means that when we start plugging in sufficiently large values for a, the function begins to decrease. In our case, we start plugging numbers at sqrt(8) and then escalate to no bound. Since it is always greater than e, it is always decreasing.\r\n\r\nSorry about the sqrt(x) thing - I get confused easily and avoid complications."
}
{
"Tag": [
"probability"
],
"Problem": "On a dice we have the following numbers written: 0,1,2,3,4,5. We throw this dice until the sum of the points that appered is bigger than 12. Which sum has the highest probability that we can obtain from all the other's?\r\na)13\r\nb)14\r\nc)15 or\r\nd)16.\r\n\r\n10x",
"Solution_1": "Not sure to understand exactly the terms of the problem. A try nevertheless :\r\n\r\n\r\nLet's denote E(n,p) the event : we have 'n' so far and we dice a 'p'\r\nLet's denote E(n) the event : we have n \r\n\r\n\r\nThen :\r\n\r\n E(13) = E(12,1) U E(11,2) U E(10,3) U E(9,4) U E(8,5)\r\n E(14) = E(12,2) U E(11,3) U E(10,4) U E(9,5)\r\n E(15) = E(12,3) U E(11,4) U E(10,5)\r\n E(16) = E(12,4) U E(11,5)\r\n E(17) = E(12,5)\r\n\r\nNow all the E(p,i) i=0..5 are equiprobable so 13 is the more probable"
}
{
"Tag": [
"modular arithmetic",
"number theory solved",
"number theory"
],
"Problem": "a) Given any positive integer $n$, show that there exist distint positive integers $x$ and $y$ such that $x + j$ divides $y + j$ for $j = 1 , 2, 3, \\ldots, n$;\r\n\r\nb) If for some positive integers $x$ and $y$, $x+j$ divides $y+j$ for all positive integers $j$, prove that $x = y$.",
"Solution_1": "a) Choose any positive integer $x$ and $y=(n+x)!+x.$\r\n\r\nb) For each positive integer $j$, the number $\\dfrac{y+j} {x+j}$ is a positive integer. Since the limit is $1$ when $j$ goes to $+ \\infty$, it follows that there exists $n$ such that $1 \\leq \\dfrac{y+n} {x+n} < 2.$ Therefore $\\dfrac{y+n} {x+n} =1$ and $x=y.$\r\n\r\nPierre.",
"Solution_2": "Could you explain a bit more on your b solution?",
"Solution_3": "Hmm, what's not clear\u00bf\r\n\r\nHe shows that the quotient $q_{n}= \\frac{y+n}{x+n}$ gets smaller than $2$ at some point, and $q_{n}$ clearly is always a positive integer (by assumption), thus for big $n$, $q_{n}$ will be $1$.",
"Solution_4": "[quote=\"ZetaX\"]Hmm, what's not clear\u00bf\n\nHe shows that the quotient $q_{n}= \\frac{y+n}{x+n}$ gets smaller than $2$ at some point, and $q_{n}$ clearly is always a positive integer (by assumption), thus for big $n$, $q_{n}$ will be $1$.[/quote]\r\nThank you, I couldn't make all of those connections at once...",
"Solution_5": "a)\r\n$x+j|y+j$ <=> $x+j|y-x$ for $j=1,2,3,..,n$\r\nwe can take $y-x=\\prod_{j=1}^{n}(x+j)$\r\n$y=\\prod_{j=1}^{n}(x+j)+x=(x+n)!+x$\r\n\r\nb)\r\nfor all positive integers $j$, $x+j|y+j$\r\n<=> for all positive integers $j$, $x+j|y-x$\r\n$j=-x+xa$ => $xa|y-x$ => $a|y-x$ for all positive integers $a$\r\n=> $y-x=0$\r\n=> $x=y$",
"Solution_6": "I have a different solution for the $2^{nd}$ part.\n We know that there are infinitely many primes.\n We have here that , \\[\\ x+j\\equiv0\\pmod{y+j} \n \\ \\forall \\ j\\in \\mathbb{N} ............ [1] \\]\n \n Now we can say , \\[ \\exists m \\in \\mathbb{N} \\] such that x+m is a prime. \n so, y+m=1 or y+m=x+m ..... from [1] \n But \\[ \\ y+m \\neq 1 \\] as y and m are +ve integers. \n So, y+m = x+m.\n \\[ \\Longrightarrow x=y.... \\]\n\\[ QED. \\]",
"Solution_7": "[quote=mynamearzo]I have a different solution for the $2^{nd}$ part.\n We know that there are infinitely many primes.\n We have here that , \\[\\ x+j\\equiv0\\pmod{y+j} \n \\ \\forall \\ j\\in \\mathbb{N} ............ [1] \\]\n \n Now we can say , \\[ \\exists m \\in \\mathbb{N} \\] such that x+m is a prime. \n so, y+m=1 or y+m=x+m ..... from [1] \n But \\[ \\ y+m \\neq 1 \\] as y and m are +ve integers. \n So, y+m = x+m.\n \\[ \\Longrightarrow x=y.... \\]\n\\[ QED. \\][/quote]\n\nMan, I didn't get your solution at all! On what basis you made a case that $y+m=1$?",
"Solution_8": "[quote=Rushil]a) Given any positive integer $n$, show that there exist distint positive integers $x$ and $y$ such that $x + j$ divides $y + j$ for $j = 1 , 2, 3, \\ldots, n$;\n\nb) If for some positive integers $x$ and $y$, $x+j$ divides $y+j$ for all positive integers $j$, prove that $x = y$.[/quote]\n\nIt has been some time since I bumped this thread. \n[hide=Solution for Part b]\nFor the divisibility, obviously $x \\le y$\n$x+j \\mid y-x+x+j \\implies x+j \\mid x-y$.\nNow pick some $j$ such that $|x-y| \\le x+j$. This forces $x-y=0$ and we are done! $\\blacksquare$\n[/hide]"
}
{
"Tag": [
"geometry",
"circumcircle",
"trigonometry",
"perpendicular bisector",
"geometry unsolved"
],
"Problem": "D is a point on the minor arc AC of the circumcircle of triangle ABC with center O. Take point P on [AB] such that angleADP=angleOBC and take a point Q on [BC] such that angleCDQ=angleOBA . Prove that angleDPQ=angleDOC.",
"Solution_1": "Sorry, i forgot to say that ABC is an acute angled triangle",
"Solution_2": "A sketch:\r\n\r\nShow first that if $A',C'$ are the intersections between the altitudes from $A,C$ of $ABC$ respectively and the circle $(O)$, then $DP,DQ$ pass through $C',A'$ respectively. Then show that $PQ$ always passes through $H$, the orthocenter of $ABC$. \r\n\r\nWhat we have to show is that $\\angle DPH=\\angle DOC$. Since $AB$ is the perpendicular bisector of $HC'$, we have $\\angle PDH=2\\angle DC'C=\\angle DOC$, Q.E.D.",
"Solution_3": "Nice problem.\r\nLet $E=PD\\cap{AO}, F=QD\\cap{CO}$. Simple angle chasing shows that $OEDF$ is cyclic, thus $\\angle{DOC}=\\angle{DEF}$, so we have to prove that $\\angle{DEF}=\\angle{DPQ}$ or that $PQ$ and $EF$ are parallel or that $\\frac{QF}{QD}=\\frac{PE}{PD}$. From triangles $APE$ and $APD$ we obtain: $\\frac{PE}{PD}=\\frac{AE\\sin{OAB}}{AD\\sin{BAD}}=\\frac{\\sin{OCB}\\cdot{\\sin{OAB}}}{\\sin{DEO}\\cdot{\\sin{BAD}}}$ . Similarly we obtain : $\\frac{QF}{QD}=\\frac{\\sin{OCB}\\cdot{\\sin{OAB}}}{\\sin{DFO}\\cdot{\\sin{BCD}}}$. But $DEOF$ and $ABCD$ are cyclic, so $\\sin{BAD}=\\sin{BCD}$ and $\\sin{DEO}=\\sin{DFO}\\Rightarrow{\\frac{QF}{QD}=\\frac{PE}{PD}}$. QED"
}
{
"Tag": [
"LaTeX"
],
"Problem": "Can someone please help me! I've accidentally deleted the wrong passage and then asked it to view on pdf form. I then thought that everything was ok, since the number of pages seemed correct and closed Texnic center, but that isn't so! Is there any way which I can get what I deleted back? Please tell me this is so? I can't believe what I've done! (I've deleted quite a bit you see and need it for tomorrow!)\r\n\r\nHelp! :( :(",
"Solution_1": "If Undo (Ctrl-Z) doesn't work, then there's no way you can get that back unless you had a backup of the file ...",
"Solution_2": "I can't do ctrl-z since I closed the program and then have now started it again to realise what I've done! Doens't it save a log of everything we do? :( \r\n\r\n(sorry, I'm pretty freaked out!) :(",
"Solution_3": "No, if you don't have a backup you're pretty much screwed. :|",
"Solution_4": "Crap! Thanks anyway. I guess I better get started again. :o \r\n\r\nServes me right anyway, I should have been more careful. :(",
"Solution_5": "For anyone interested, I have managed to rescue myself. (I had saved the document as a pdf file earlier, and so am just going to print the deleted pages from there!) Phew. lesson of the day- before editing save a txt file of the tex!",
"Solution_6": "[quote=\"beans\"]For anyone interested, I have managed to rescue myself. (I had saved the document as a pdf file earlier, and so am just going to print the deleted pages from there!) Phew. lesson of the day- before editing save a txt file of the tex![/quote]TeXnic center automatically makes backups of your file once every 10 minutes (by default, you can lower that down from the options page to 1 minute I think).",
"Solution_7": "Where does it save these backup files to?",
"Solution_8": "They just overwrite previous saves (this is a good reason to save as soon as you open the document, and to save under a new name [b][i]before[/b][/i] making any changes that you don't want saved in the old file."
}
{
"Tag": [
"induction",
"logarithms",
"number theory unsolved",
"number theory"
],
"Problem": "let n is a natural number\r\nProve that\r\n2>1/n+1 +...+ 1/3n+1",
"Solution_1": "[hide=\"Solution\"]$ \\frac {1}{n\\plus{}2} < \\frac {1}{n\\plus{}1}$\n$ \\frac {1}{n\\plus{}3} < \\frac {1}{n\\plus{}1}$\n...\n$ \\frac {1}{3n\\plus{}1} < \\frac {1}{n\\plus{}1}$=>\n\n$ \\frac {1}{n\\plus{}1} \\plus{}\\frac {1}{n\\plus{}2} \\plus{} ... \\plus{}\\frac {1}{3n\\plus{}1}$ < $ (2n\\plus{}1)(\\frac {1}{n\\plus{}1})$ < $ 2$\n\nThe end!\n\nThis can be proven by mathematical induction also.[/hide]",
"Solution_2": "We can make the rather stronger claim that the upper bound is $ \\ln 3 \\approx 1.0986$, which is quite a bit tighter."
}
{
"Tag": [
"inequalities proposed",
"inequalities"
],
"Problem": "Let : \r\n\\[ a \\geq b \\geq c \\geq 1 \\]\r\n\\[ x \\geq y \\geq z \\geq 1 \\]\r\nProve that :\r\n\\[ a^x + b^y + c^z \\geq 3(\\frac{a+b+c}{3})^\\frac{x+y+z}{3} \\]",
"Solution_1": "By applying Power-Mean on the RHS, we see that it's sufficient to prove\r\n\\[\r\na^x + b^y + c^z \\geq a^{\\frac{x+y+z}{3}} + b^{\\frac{x+y+z}{3}} + c^{\\frac{x+y+z}{3}}\r\n\\]\r\nThat should be obvious. Or you can just apply a simple smoothing argument on the exponents.",
"Solution_2": "\\[ a^x + b^y + c^z \\geq a^\\frac{x+y+z}{3} + b^\\frac{x+y+z}{3} + c^\\frac{x+y+z}{3} \\]\r\n You mean this equality is obvious . Please show me this .",
"Solution_3": "Okay\r\n\r\n[b]Lemma:[/b] Suppose that $a\\geq b \\geq 1$ and $p \\geq q \\geq x \\geq 0$, then $a^{p+x} + b^{q-x} \\geq a^p + b^q$.\r\n\r\n[b]Proof:[/b] We have \\[ a^{p+x} + b^{q-x} \\geq a^p \\cdot b^x + b^{q-x} \\cdot 1 \\geq a^p + b^q \\] (we used the Rearrangement Inequality). $\\square$\r\n\r\nSo we can just apply a smoothing argument on the LHS to make the exponents equal while not increasing its value. (I'm trying to think of a better way to this)\r\n\r\nMore generally, I think that if $a_1 \\geq a_2 \\geq \\cdots \\geq a_n \\geq 1$ and the sequence $x_1 \\geq x_2 \\geq \\cdots \\geq x_n \\geq 0$ majorizes the sequence $y_1 \\geq y_2 \\geq \\cdots \\geq y_n \\geq 0$, then\r\n\\[\r\na_1^{x_1} + \\cdots + a_n^{x_n} \\geq a_1^{y_1} + \\cdots + a_n^{y_n}\r\n\\]",
"Solution_4": "Here is a different route to prove this which resembles the proof of Chebycheff.\r\n\r\n[b]Claim:[/b]: $\\displaystyle a^x+b^y+c^z \\geq \\frac{a^x+a^y+a^z}{3} + \\frac{b^x+b^y+b^z}{3} + \\frac{c^x+c^y+c^z}{3}$ (which makes a nice problem itself :) :) )\r\n\r\n\r\n[b]Lemma[/b]: If $x\\geq y > 0$ and $u\\geq v \\geq 1$, then $u^x-u^y \\geq v^x-v^y$.\r\n\r\n[b]Proof:[/b] $u^x-u^y = u^y(u^{x-y} - 1) \\geq v^y(v^{x-y} - 1) = v^x-v^y$.\r\n\r\n\r\nBy the Lemma, we have:\r\n\r\n(1) $a^x+b^y+c^z\\geq a^y+b^z+c^x$ since $(a^x-a^y) + (b^y-b^z) \\geq (c^x-c^y) + (c^y-c^z) = c^x-c^z$.\r\n\r\n(2) $a^x+b^y+c^z\\geq a^z+b^x+c^y$ since $a^x-a^z = (a^x-a^y)+ (a^y-a^z) \\geq (b^x-b^y) + (c^y-c^z) $.\r\n\r\n(3) $a^x+b^y+c^z\\geq a^x+b^y+c^z$ is obviously true.\r\n\r\nSo the claim is true by adding (1), (2) and (3).\r\n\r\nFrom the claim, we could go either AM-GM first then Power Mean or the reverse.\r\n\r\nAM-GM first leads to billzhao's claim above.\r\n\r\nBy AM-GM, $\\displaystyle \\frac{a^x+a^y+a^z}{3} + \\frac{b^x+b^y+b^z}{3} + \\frac{c^x+c^y+c^z}{3} \\geq a^{\\frac{x+y+z}{3}} + b^{\\frac{x+y+z}{3}} + c^{\\frac{x+y+z}{3}}$.\r\n\r\nPower Mean first,\r\n\r\n$\\displaystyle \\frac{a^x+b^x+c^x}{3} + \\frac{a^y+b^y+c^y}{3} + \\frac{a^z+b^z+c^z}{3} $\r\n$\\displaystyle \\geq \\left(\\frac{a+b+c}{3}\\right)^x + \\left(\\frac{a+b+c}{3}\\right)^y + \\left(\\frac{a+b+c}{3}\\right)^z$ then apply AM-GM to get the desired result.",
"Solution_5": "$Claim$:Let $a_{n}$ and $b_{n}$ are two decreasing sequences such that:$a_{i} >=1$ and $b_{i}>=1$ for all 1<=i<=n.\r\nThen we have:$\\sum a_{i}^{b_{i}}$>=$\\sum a_{i}^{\\sum b_{i}/n}$.\r\n$Proof:$\r\nWell,I have another solution based on this lemma (Prove by Abel's identity)\r\n$Lemma$:If $x_1,...,x_{n}$ and $y_1,...,y_{n}$ are two positive sequences satisfied:\r\nFor all 1<=j<=n:\r\n$\\prod x_{i}$>=$\\prod y_{i}$ with i=1 to j.\r\n$y_1>=y_2>=...>=y_{n}$.Then:\r\n$x_1+...+x_{n}$>=$y_1+...+y_{n}$\r\n\r\nThen we apply this lemma :$x_{i}=a_{i}^{b_{i}}$ and $y_{i}=a_{i}^{(\\sum b_{i})/n}$ (the condition of x_{i} and y_{i} to the lemma can be easily to prove).Then we have this claim. :)"
}
{
"Tag": [
"geometry proposed",
"geometry"
],
"Problem": "Dear everyone.\r\n\r\nFor four points A,B,C,D\r\nNine-point-circles of triangles ABC,BCD,CDA,DAB are concurrent.\r\n\r\nMay I ask you a favor? thanks.",
"Solution_1": "view topic: http://www.mathlinks.ro/viewtopic.php?t=291949"
}
{
"Tag": [
"calculus",
"integration",
"algebra",
"function",
"domain",
"Ring Theory",
"superior algebra"
],
"Problem": "Hello together,\r\n\r\nIm searching for an integral domain $ R$ with this property:\r\n\r\nThere are $ a,p \\in R\\setminus \\{0\\}$ with $ p$ prim and $ a$ arbitrarily, so that $ \\max\\{n\\in \\mathbb{N}| p^n \\text{ devides } a\\}$ does not exist, in other words, such that $ p$ is an infinity divisor of $ a$.\r\n\r\nOr, if there is none, how to proof? The only condition to $ R$ is to be an integral domain, if $ R$ is not integral dom, you can clearly find examples.\r\n\r\nThank you.\r\nNatroc",
"Solution_1": "Consider the ring $ R \\equal{} K[x, \\frac {z}{x}, \\frac {z}{x^2}, ... , \\frac {z}{x^n}, ...] \\subset K(x,z)$. It surely is an integral domain as it is a subset of a field. Also, $ x$ is prime ($ x$ divides some element iff it divides the $ z$-free part, so primality of $ x$ is reduced to primalityof $ x$ in $ K[x]$).\r\nBut now $ x^n|z$ for all $ n$ (by definition of $ R$).",
"Solution_2": "What is meant by $ K(x,z)$? The quotient field of $ K[x,z]$? And do you mean $ R: \\equal{} K[x,\\frac {z}{x},\\frac {z}{x^2},...,\\frac {z}{x^n},...]$ ?",
"Solution_3": "[quote=\"Natroc\"]What is meant by $ K(x,z)$? The quotient field of $ K[x,z]$?[/quote]\nYes.\n\n[quote=\"Natroc\"]And do you mean $ R: \\equal{} K[x,\\frac {z}{x},\\frac {z}{x^2},...,\\frac {z}{x^n},...]$ ?[/quote]\r\nUmm, what's the difference\u00bf",
"Solution_4": "Ok, thank you.",
"Solution_5": "I would read what you (ZetaX) first wrote as stopping at $ n$- which means that it doesn't make sense without $ n$ defined in advance. A terminal ... makes it unambiguous.",
"Solution_6": "Ah, sorry, yes :oops:"
}
{
"Tag": [
"geometry",
"3D geometry",
"tetrahedron"
],
"Problem": "All the point in 3-space are coloured with 5 different colours (there is at least one point of each colour) prove that there exists at least one plane such that the points that make up that plane are made up of at least four of those colours.",
"Solution_1": "Isn't that trivial? Just take any four points of different colors (they exist by the problem definition) and connect them. We've got our plane.",
"Solution_2": "[quote=\"nr1337\"]Isn't that trivial? Just take any four points of different colors (they exist by the problem definition) and connect them. We've got our plane.[/quote]\r\n\r\nA plane is defined by three points, not four.",
"Solution_3": "[quote=\"dts\"][quote=\"nr1337\"]Isn't that trivial? Just take any four points of different colors (they exist by the problem definition) and connect them. We've got our plane.[/quote]\n\nA plane is defined by three points, not four.[/quote]\r\n\r\nWhy can't we have a rectangular plane? Planes aren't necissarily triangular.",
"Solution_4": "[quote=\"nr1337\"][quote=\"dts\"][quote=\"nr1337\"]Isn't that trivial? Just take any four points of different colors (they exist by the problem definition) and connect them. We've got our plane.[/quote]\n\nA plane is defined by three points, not four.[/quote]\n\nWhy can't we have a rectangular plane? Planes aren't necissarily triangular.[/quote]\r\n\r\nIf you have four points, you cannot guarantee that they are all on the same plane. They could form a tetrahedron, in which case they do not define a plane.",
"Solution_5": "This is one of those things I can see is true by just thinking about it for a little bit, but for which I can't come up with a nice proof. \r\nMaybe this works. Let P be a plane that contains three points of different colors. Now let l be a line that contains points of the other two colors. Consider the set of all planes that contain l. If one of them contains 4 different color points we are done. Otherwise, none do. But that means that the colors of the plane P must be divided into lines. Simply take a plane that contains one of the other two colors, but which is tilted at a different angle. This plane will contain that color and the three original colors. As I warned, I can't really explain my logic."
}
{
"Tag": [
"group theory",
"abstract algebra",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "$G$ a finite group of order $n$. Suppose that if $G$ has a subgroup of order $k$, then there is only one subgroup of order $k$.\r\nIs $G$ necessary cyclic?",
"Solution_1": "In particular, each $p$-Sylow of $G$ is unique and whence normal in $G$. Also their intersection is pairwise disjoint, and $G$ can be written as the internal direct product of its $p$-Sylows. Now if each $p$-Sylow is cyclic, we are done. Otherwise, for each $p$-Sylow $S_{p}$ with cardinal say $p^{n}$ ($n >1$) we must have $\\sum_{k=0, \\dots,n-1} p^{k} \\geq p^{n}$ (since each unique subgroup of $S_{p}$ has a cardinal dividing $p^{n}$ strictly). This is obviously absurd.\r\n\r\nP.S : in the case $G$ is not abelian, the $p$-Sylow subgroups of $G$ are still determining an internal direct product but I didn't verify carefully that this sum must equal $G$ (and is not strictly included in $G$). It seems kind of clear by induction.\r\n\r\nI hope I did not commit any big mistake this time :D",
"Solution_2": "[quote=\"julien_santini\"] Also their intersection is pairwise disjoint, and $G$ can be written as the internal direct product of its $p$-Sylows. [/quote]\r\n\r\njust a question : why can $G$ be written as the internal direct product of its $p$-Sylows. The $p$-Sylows are just pairwise distinct and normal. Does it suffice ?",
"Solution_3": "Yes:\r\nLet $S: = S_{p_1} \\times \\S_{p_2} \\times ... \\times S_{p_n}$ be the direct product of the sylow subgroups.\r\nThen the claim is that $f: (x_1,x_2,...,x_n) \\mapsto x_1x_2...x_n$ is a group isomorphism.\r\n\r\nLemma:\r\nif $U,V \\subset G$ are normal subgroups with $U \\cap V = \\{ e \\}$, then the map $F: U \\times V \\mapsto G, (u,v) \\mapsto uv$ is injective and $UV=f(U \\times V)$ is a normal subgroup in $G$.\r\nProve: try yourself, it's easy.\r\n\r\nBut multiple use of this lemma implies the result.",
"Solution_4": "Yes, it does. What you want to prove is this: given two normal subgroups $H,K$ of a group $G$ satisfying $H\\cap K=\\{1\\}$, the subgroup $HK$ of $G$ is the internal direct product $H\\times K$. \r\n\r\nIt suffices to prove that every element $h$ of $H$ commutes with every element $k$ of $K$. Look at the commutator $[h,k]=hkh^{-1}k^{-1}$. On the one hand, it must belong to $H$, since both $h$ and $kh^{-1}k^{-1}$ do. On the other hand, it must belong to $K$, since both $hkh^{-1}$ and $k^{-1}$ do. Since $H$ and $K$ have trivial intersection, we must have $[h,k]=1$, meaning that $h$ and $k$ commute, as desired.\r\n\r\nEdit:\r\n\r\nI see ZetaX answered before I did.",
"Solution_5": "oops :blush: I thought a necessary condition was $S_i \\cap _{j \\neq i}=\\{1\\}$ and not just $S_i \\cap S_j=\\{1\\}$... sorry again ...",
"Solution_6": "For every proper divisor $k$ of $n$, there are, under the given assumption, at most $\\phi(k)$ elements of order $k$.\r\nSince\r\n\\[ \\sum_{k|n}\\phi(k)=n\\,, \\]\r\nthere still are at least $\\phi(n)$ elements of order not a proper divisor of $n$. These have necessarily order $n$, so they generate $G$, which is then cyclic.",
"Solution_7": "why at most $\\phi(k)$ elements of order $k$ ? (It makes me think to the proof that any finite subgroup of the multiplicative group of a field is cyclic. But here I can't understand your statement ?)",
"Solution_8": "Yes, it works very similar:\r\nIf $x$ has order $k$, then of the powers $x^i$, only the ones with $gcd(i,k)=1$ have order $k$ too.\r\nNow assume there is some $y \\not \\in $ with order $k$. But then $$ and $$ are different subgroups with $k$ elements."
}
{
"Tag": [
"superior algebra",
"superior algebra unsolved"
],
"Problem": "What notation could be used to denote the following field?:\r\n\r\n\\[ \\left\\{\\frac{a\\plus{}bx}{c\\plus{}dx}\\ : \\ a,b,c,d\\in\\mathbb{Z}_{2}\\right\\}.\\]\r\nCould we use $ \\mathbb{Z}_{2}\\left(x\\right)/x^{2}\\mathbb{Z}_{2}\\left(x\\right)$?",
"Solution_1": "In what sense is this a field? It does not look closed under multiplication to me. Also, I am not sure what you mean by $ x \\over 0$.\r\n\r\nBy the way, $ \\mathbb{Z}_{2}\\left(x\\right)/x^{2}\\mathbb{Z}_{2}\\left(x\\right)$ is not a field, as it has zero divisors: $ x^2\\equal{}0$.",
"Solution_2": "I'm looking for a field with $ 16$ elements but I guess that example doesn't make sense. Would the field $ \\mathbb{Z}_{2}\\left[x\\right]/\\left$ work?",
"Solution_3": "Yes, it's always the case that $ \\mathbb{F}_p [x] / \\left \\cong \\mathbb{F}_{p^k}$ when $ p$ is degree $ k$ and irreducible over $ \\mathbb{F}_p$."
}
{
"Tag": [
"ARML",
"algebra",
"polynomial",
"pigeonhole principle",
"probability",
"geometry",
"perimeter"
],
"Problem": "\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\r\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Final Week}\\\\ \\hline\r\n& Day 43 & Day 44 & Day 45 & Day 46 & Day 47 & Day 48 & Day 49 & Final Score\\\\ \\hline\r\nd343seven & 3 & - & 5 & - & - & 5 & 4 & 200\\\\ \\hline\r\npaladin8 & 3 & - & - & 6 & 6 & - & - & 163\\\\ \\hline\r\nalan & 3 & - & 5 & - & 6 & 5 & 4 & 123\\\\ \\hline\r\ndjshowdown2 & - & - & - & - & - & - & - & 89\\\\ \\hline\r\nxxreddevilzxx & 3 & - & - & - & - & 5 & 4 & 85\\\\ \\hline\r\nElemennop & 3 & - & - & - & - & - & - & 82\\\\ \\hline\r\nJesusFreak197 & 3 & - & 5 & 6 & 6 & 5 & 4 & 73\\\\ \\hline\r\nisaacchao & 3 & 7 & 5 & 6 & - & 5 & 4 & 71\\\\ \\hline\r\nRep123Max & - & - & - & - & - & - & - & 70\\\\ \\hline\r\nqweretyq & - & - & - & - & - & - & - & 58\\\\ \\hline\r\nThe Original Pie Guy & - & - & - & - & - & - & - & 47\\\\ \\hline\r\nBornForMath & - & - & - & - & - & - & - & 45\\\\ \\hline\r\nScrambled & - & - & - & - & - & - & - & 43\\\\ \\hline\r\nfurious & - & - & - & - & - & - & - & 38\\\\ \\hline\r\nwobster109 & 3 & - & 5 & - & - & 5 & - & 38\\\\ \\hline\r\nwhite-horse-king88 & - & - & - & - & - & - & - & 26\\\\ \\hline\r\nChigr & 3 & - & 5 & 6 & - & 5 & 4 & 26\\\\ \\hline\r\nsheepwarrior & - & - & - & - & - & - & - & 24\\\\ \\hline\r\n\r\n\\end{tabular}\\]\r\n\r\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\r\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Final Week(Continued)}\\\\ \\hline\r\n& Day 43 & Day 44 & Day 45 & Day 46 & Day 47 & Day 48 & Day 49 & Final Score\\\\ \\hline\r\nShogia & - & - & - & - & - & - & - & 23\\\\ \\hline\r\nmiraculouspostmaster & - & - & - & - & - & - & - & 21\\\\ \\hline\r\nSly Si & 3 & 7 & 5 & - & - & - & - & 18\\\\ \\hline\r\nh\\_s\\_potter2002 & - & - & - & - & - & - & - & 16\\\\ \\hline\r\nBictor717 & - & - & - & - & - & - & - & 15\\\\ \\hline\r\neryaman & - & - & - & - & - & - & - & 13\\\\ \\hline\r\nWindSlicer & - & - & - & - & - & - & - & 13\\\\ \\hline\r\nfrodo & - & - & - & - & - & - & - & 12\\\\ \\hline\r\nmathfanatic & - & - & - & - & - & - & - & 11\\\\ \\hline\r\ndongkooklim & - & - & - & - & - & - & - & 3\\\\ \\hline\r\nMaThWhIz2004 & - & - & - & - & - & - & - & 3\\\\ \\hline\r\nSlizzel &- & - & - & - & - & - & - & 3\\\\ \\hline\r\n\\end{tabular}\\]\r\n\r\n[hide=\"Week 6\"]\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 6}\\\\ \\hline\n& Day 36 & Day 37 & Day 38 & Day 39 & Day 40 & Day 41 & Day 42 & Total\\\\ \\hline\nd343seven & 3 & 3 & 3 & 5 & 4 & 3 & 3 & 183\\\\ \\hline\npaladin8 & 3 & 3 & 3 & 5 & 4 & 3 & 3 & 148\\\\ \\hline\nalan & 3 & 3 & 3 & 5 & 4 & 3 & 3 & 100\\\\ \\hline\ndjshowdown2 & - & - & - & - & - & - & - & 89\\\\ \\hline\nElemennop & 3 & - & - & 5 & 4 & - & - & 79\\\\ \\hline\nxxreddevilzxx & - & 3 & 3 & 5 & - & - & 3 & 73\\\\ \\hline\nRep123Max & - & - & - & - & - & - & - & 70\\\\ \\hline\nqweretyq & - & - & - & - & - & - & - & 58\\\\ \\hline\nThe Original Pie Guy & - & - & - & - & 4 & 3 & 3 & 47\\\\ \\hline\nBornForMath & 3 & 3 & 3 & 5 & 4 & 3 & 3 & 45\\\\ \\hline\nJesusFreak197 & 3 & 3 & 3 & 5 & - & 3 & 3 & 44\\\\ \\hline\nScrambled & - & - & - & - & - & - & - & 43\\\\ \\hline\nisaacchao & 3 & 3 & 3 & 5 & 4 & 3 & 3 & 41\\\\ \\hline\nfurious & - & - & - & - & - & - & - & 38\\\\ \\hline\nwhite-horse-king88 & - & - & - & - & - & - & - & 26\\\\ \\hline\nwobster109 & 3 & - & 3 & 5 & 4 & 3 & 3 & 25\\\\ \\hline\nsheepwarrior & 3 & - & - & - & - & - & - & 24\\\\ \\hline\nShogia & 3 & - & 3 & 5 & - & - & - & 23\\\\ \\hline\n\\end{tabular}\\]\n\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 6(Continued)}\\\\ \\hline\n& Day 36 & Day 37 & Day 38 & Day 39 & Day 40 & Day 41 & Day 42 & Total\\\\ \\hline\nmiraculouspostmaster & - & - & - & - & - & - & - & 21\\\\ \\hline\nh\\_s\\_potter2002 & - & - & - & - & - & - & - & 16\\\\ \\hline\nBictor717 & - & - & - & - & - & - & - & 15\\\\ \\hline\neryaman & - & - & - & - & - & - & - & 13\\\\ \\hline\nWindSlicer & - & - & - & - & - & - & - & 13\\\\ \\hline\nmathfanatic & - & - & - & - & - & - & - & 11\\\\ \\hline\nfrodo & - & - & 3 & 5 & 4 & - & - & 12\\\\ \\hline\nChigr &- & - & - & - & - & - & 3 & 3\\\\ \\hline\ndongkooklim & - & - & - & - & - & - & - & 3\\\\ \\hline\nMaThWhIz2004 & - & - & - & - & - & - & - & 3\\\\ \\hline\nSlizzel &- & - & - & - & - & 3 & - & 3\\\\ \\hline\nSly Si &- & - & - & - & - & - & 3 & 3\\\\ \\hline\n\\end{tabular}\\]\n[/hide]\n\n[hide=\"Week 5\"]\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{10}{|c|}{ARML Practice Series Scoreboard Week 5}\\\\ \\hline\n& Day 29 & Day 30 & Day 31 & Day 32 & PS1* & Day 33 & Day 34 & Day 35 & Total\\\\ \\hline\nd343seven & 2 & 3 & 2 & 4 & 30 & 3 & - & 4 & 159\\\\ \\hline\npaladin8 & 2 & 3 & 2 & 4 & 17 & 3 & - & - & 124\\\\ \\hline\ndjshowdown2 & 2 & - & - & - & - & - & - & - & 89\\\\ \\hline\nalan & 2 & 3 & 2 & - & - & 3 & - & 4 & 76\\\\ \\hline\nRep123Max & - & - & - & - & - & - & - & - & 70\\\\ \\hline\nElemennop & 2 & - & - & - & - & - & - & - & 67\\\\ \\hline\nxxreddevilzxx & 2 & 3 & 2 & 4 & - & 3 & 6 & 4 & 59\\\\ \\hline\nqweretyq & 2 & - & - & - & - & - & - & - & 58\\\\ \\hline\nScrambled & - & - & - & - & - & - & - & - & 43\\\\ \\hline\nfurious & 2 & - & - & - & - & - & - & - & 38\\\\ \\hline\nThe Original Pie Guy & 2 & - & - & - & - & - & - & - & 37\\\\ \\hline\nwhite-horse-king88 & - & - & - & - & - & - & - & - & 26\\\\ \\hline\nJesusFreak197 & - & - & - & 4 & 20 & - & - & - & 24\\\\ \\hline\nBornForMath & - & - & - & - & 17 & - & - & 4 & 21 \\\\ \\hline\nmiraculouspostmaster & - & - & - & - & - & - & - & - & 21\\\\ \\hline\nsheepwarrior & - & - & - & - & - & - & - & - & 21\\\\ \\hline\nisaacchao & - & - & - & 4 & - & 3 & 6 & 4 & 17\\\\ \\hline\nh\\_s\\_potter2002 & - & - & - & - & - & - & - & - & 16\\\\ \\hline\n\\end{tabular}\\]\n\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{10}{|c|}{ARML Practice Series Scoreboard Week 5(Continued)}\\\\ \\hline\n& Day 29 & Day 30 & Day 31 & Day 32 & PS1* & Day 33 & Day 34 & Day 35 & Total\\\\ \\hline\nBictor717 & - & - & - & - & - & - & - & - & 15\\\\ \\hline\neryaman & - & - & - & - & - & - & - & - & 13\\\\ \\hline\nWindSlicer & - & - & - & - & - & - & - & - & 13\\\\ \\hline\nShogia & 2 & - & - & - & 2 & - & 6 & - & 12\\\\ \\hline\nmathfanatic & - & - & - & - & - & - & - & - & 11\\\\ \\hline\nwobster109 &- & - & - & - & - & - & - & 4 & 4\\\\ \\hline\ndongkooklim &- & - & - & - & - & - & - & - & 3\\\\ \\hline\n\\end{tabular}\\]\n* PS1 is the problem set posted by joml88.\n[/hide]\n\n[hide=\"Week 4\"]\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 4}\\\\ \\hline\n& Day 22 & Day 23 & Day 24 & Day 25 & Day 26 & Day 27 & Day 28 & Total\\\\ \\hline\nd343seven & 4 & 5 & 7 & 4 & 5 & 1 & 1 & 111\\\\ \\hline\npaladin8 & - & 5 & - & 4 & 5 & 1 & 1 & 93\\\\ \\hline\ndjshowdown2 & 4 & 5 & 7 & 4 & - & 1 & 1 & 87\\\\ \\hline\nRep123Max & - & - & - & - & - & 1 & - & 70\\\\ \\hline\nElemennop & - & - & - & - & 5 & 1 & 1 & 65\\\\ \\hline\nalan & 4 & - & 4 & 4 & 5 & 1 & 1 & 62\\\\ \\hline\nqweretyq & - & - & - & - & 5 & 1 & 1 & 56\\\\ \\hline\nScrambled & - & - & - & - & 5 & 1 & 1 & 43\\\\ \\hline\nfurious & 4 & - & - & - & - & 1 & 1 & 36\\\\ \\hline\nThe Original Pie Guy & - & - & - & - & - & 1 & 1 & 35\\\\ \\hline\nxxreddevilzxx & - & - & - & 4 & 5 & 1 & 1 & 35\\\\ \\hline\nwhite-horse-king88 & - & - & - & - & - & - & - & 26\\\\ \\hline\nmiraculouspostmaster & - & - & - & - & - & - & - & 21\\\\ \\hline\nsheepwarrior & - & - & - & - & - & 1 & - & 21\\\\ \\hline\nh\\_s\\_potter2002 & - & - & - & - & - & - & - & 16\\\\ \\hline\nBictor717 & - & - & - & - & - & - & - & 15\\\\ \\hline\neryaman & - & - & - & - & 5 & - & - & 13\\\\ \\hline\nWindSlicer & - & - & - & - & - & - & - & 13\\\\ \\hline\nmathfanatic & - & - & - & - & - & - & - & 11\\\\ \\hline\ndongkooklim &- & - & 1 & - & - & 1 & 1 & 3\\\\ \\hline\n\\end{tabular}\\]\n\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 4(Continued)}\\\\ \\hline\n& Day 22 & Day 23 & Day 24 & Day 25 & Day 26 & Day 27 & Day 28 & Total\\\\ \\hline\nShogia & - & - & - & - & - & 1 & 1 & 2\\\\ \\hline\n\\end{tabular}\\]\n[/hide]\n\n[hide=\"Week 3\"]\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 3}\\\\ \\hline\n& Day 15 & Day 16 & Day 17 & Day 18 & Day 19 & Day 20 & Day 21 & Total\\\\ \\hline\nd343seven & 2 & 5 & 5 & 2 & 4 & 4 & 3 & 84\\\\ \\hline\npaladin8 & 2 & - & 5 & 2 & 4 & 4 & 3 & 77\\\\ \\hline\nRep123Max & 2 & - & - & - & 4 & 4 & 3 & 69\\\\ \\hline\ndjshowdown2 & - & - & 4 & 2 & 4 & 4 & 3 & 65\\\\ \\hline\nElemennop & 2 & - & - & 2 & 4 & 4 & 3 & 58\\\\ \\hline\nqweretyq & 2 & - & - & - & 4 & - & - & 49\\\\ \\hline\nalan & - & - & - & 2 & 4 & 4 & 3 & 43\\\\ \\hline\nScrambled & - & 5 & - & - & 4 & - & 3 & 36\\\\ \\hline\nThe Original Pie Guy & 2 & - & - & 2 & - & - & 3 & 33\\\\ \\hline\nfurious & 2 & 5 & 5 & 2 & 4 & 4 & 3 & 30\\\\ \\hline\nwhite-horse-king88 & 2 & 5 & - & - & 4 & - & 3 & 26\\\\ \\hline\nxxreddevilzxx & 2 & - & - & - & 4 & 4 & 3 & 24\\\\ \\hline\nmiraculouspostmaster & - & - & - & - & - & - & - & 21\\\\ \\hline\nsheepwarrior & 2 & - & - & 2 & 4 & 4 & 3 & 20\\\\ \\hline\nh\\_s\\_potter2002 & 2 & - & - & - & - & - & - & 16\\\\ \\hline\nBictor717 & 2 & - & - & - & 4 & - & - & 15\\\\ \\hline\nWindSlicer & 2 & - & - & - & 4 & 4 & - & 13\\\\ \\hline\nmathfanatic & - & - & - & - & - & - & - & 11\\\\ \\hline\neryaman & - & - & - & - & 4 & 4 & - & 8\\\\ \\hline\n\\end{tabular}\\][/hide]\n\n[hide=\"Week 2\"]\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 2}\\\\ \\hline\n& Day 8 & Day 9 & Day 10 & Day 11 & Day 12 & Day 13 & Day 14 & Total\\\\ \\hline\nd343seven & 5 & 6 & 5 & 10 & 8 & 4 & 5 & 59\\\\ \\hline\npaladin8 & 5 & 6 & 5 & 10 & - & 4 & 5 & 57\\\\ \\hline\nRep123Max & 5 & - & 5 & 10 & - & 4 & - & 56\\\\ \\hline\nfanaticsm & 5 & 6 & 5 & 10 & - & - & - & 48\\\\ \\hline\ndjshowdown2 & 5 & 6 & 5 & 4 & - & - & - & 48\\\\ \\hline\nqweretyq & - & 6 & 5 & 10 & - & 4 & - & 43\\\\ \\hline\nElemennop & - & 6 & 5 & 7 & 8 & 4 & - & 43\\\\ \\hline\nalan & 5 & 6 & 5 & - & - & - & - & 30\\\\ \\hline\nThe Original Pie Guy & - & - & - & - & - & 4 & 5 & 26\\\\ \\hline\nScrambled & - & - & - & 6 & - & 4 & - & 24\\\\ \\hline\nmiraculouspostmaster & - & - & - & - & - & - & - & 21\\\\ \\hline\nh\\_s\\_potter2002 & - & - & 5 & - & - & 4 & 5 & 14\\\\ \\hline\nwhite-horse-king88 & - & - & - & - & - & - & - & 12\\\\ \\hline\nmathfanatic & - & - & - & - & - & - & - & 11\\\\ \\hline\nxxreddevilzxx & - & 6 & - & 5 & - & - & - & 11\\\\ \\hline\nBictor717 & - & - & - & - & - & 4 & 5 & 9\\\\ \\hline\nfurious & - & - & 5 & - & - & - & - & 5\\\\ \\hline\nsheepwarrior & - & - & 5 & - & - & - & - & 5\\\\ \\hline\nWindSlicer & - & - & - & - & - & - & - & 3\\\\ \\hline\n\\end{tabular}\\]\n[/hide]\n\n[hide=\"Week 1\"]\n\\[\\begin{tabular}{|c|c|c|c|c|c|c|c|c|} \\hline\n\\multicolumn{9}{|c|}{ARML Practice Series Scoreboard Week 1}\\\\ \\hline\n& Day 1 & Day 2 & Day 3 & Day 4 & Day 5 & Day 6 & Day 7 & Total\\\\ \\hline\nRep123Max & 4 & 6 & 2 & 6 & 3 & 7 & 4 & 32\\\\ \\hline\ndjshowdown2 & - & 6 & 2 & 6 & 3 & 7 & 4 & 28\\\\ \\hline\nfanaticsm & - & - & 2 & 6 & 3 & 7 & 4 & 22\\\\ \\hline\npaladin8 & - & - & 2 & 6 & 3 & 7 & 4 & 22\\\\ \\hline\nmiraculouspostmaster & 4 & - & - & 6 & - & 7 & 4 & 21 \\\\ \\hline\nqweretyq & - & - & 2 & 6 & 3 & 7 & - & 18\\\\ \\hline\nScrambled & - & - & 2 & 6 & 3 & 7 & - & 18\\\\ \\hline\nThe Original Pie Guy & - & - & - & 6 & - & 7 & 4 & 17\\\\ \\hline\nd343seven & - & - & 2 & - & 3 & 7 & 4 & 16\\\\ \\hline\nalan & - & - & - & - & 3 & 7 & 4 & 14\\\\ \\hline\nElemennop & - & - & 2 & - & - & 7 & 4 & 13\\\\ \\hline\nwhite-horse-king88 & - & - & 2 & - & 3 & 7 & - & 12\\\\ \\hline\nmathfanatic & - & 6 & 2 & - & 3 & - & - & 11\\\\ \\hline\nWindSlicer & - & - & - & - & 3 & - & - & 3\\\\ \\hline\nfurious & - & - & - & - & - & - & - & -\\\\ \\hline\nh\\_s\\_potter2002 & - & - & - & - & - & - & - & -\\\\ \\hline\nsheepwarrior & - & - & - & - & - & - & - & -\\\\ \\hline\n\\end{tabular}\\]\n[/hide]",
"Solution_1": "Here are solutions to the first two days. Most of the times we will use solutions that are PMed to us by competitors.\r\n\r\n[size=150][b]Day 1[/b][/size]\r\nLet $Q(x)$ be the quotient when $37x^{73}-73x^{37}+36$ is divided by $x-1$. Compute the sum of the coefficients of $Q(x)$.\r\n\r\n[i]Solution-Complements of Rep123Max[/i]\r\n\r\n[quote=\"Rep123Max\"]\n$37x^{73}-73x^{37}+36=37x^{73}-73x^{37}+73-37=37(x^{73}-1)-73(x^{37}-1)$\n\n$(x-1)|(x^{73}-1)$ and $(x-1)|(x^{37}-1)$ leaving a quotient of\n\n$37(x^{72}+x^{71}+\\ldots+x+1)-73(x^{36}+x^{35}+\\ldots+x+1)$\n\nSubbing in $x=1$ since the question wants the sum of the coefficients of the quotient, it is $37(73)-73(37)=\\boxed{0.}$[/quote]\n\n[size=150][b]Day 2[/b][/size]\n\nLet $P(x)$ be a polynomial whose degree is $1996$. If $P(n)=\\frac 1n$ for $n=1,2,3,\\ldots,1997$, compute the value of $P(1998).$\n\n[i]Solution[/i]\n\n[quote]First, multiply both sides of $P(n)=\\frac 1n$ by $n$ and subtract both sides by one:\n\n\\[nP(n)-1=0.\\]\n\nNow, we know that this is satisfied for $1,2,\\ldots,1997$. In other words, $nP(n)-1$ has zeros at $1,2,\\ldots,1997$. So we can write it as:\n\n\\[nP(n)-1=k(n-1)(n-2)\\cdots(n-1997)\\]\n\nwhere $k$ is some constant. So pluggin in $n=0$ we can find $k$ as follows:\n\n\\[-1=k(-1)(-2)\\cdots(-1997)\\Rightarrow k=\\frac{1}{1997!}.\\]\n\nTherefore \\[1998P(1998)-1=\\frac{1}{1997!}(1997)(1996)\\cdots(1)\\Rightarrow P(1998)=\\boxed{\\frac{1}{999}.}\\][/quote]",
"Solution_2": "[size=150][color=black]Day 3[/color][/size]\r\n\r\nIf $P(x)$ is a polynomial in $x$, and $x^{23}+23x^{17}-18x^{16}-24x^{15}+108x^{14}=(x^4-3x^2-2x+9)\\cdot P(x)$ for all values of $x$, compute the sum of the coefficients of $P(x)$.\r\n\r\n\r\nThe most common solution given was straightforward polynomial division:\r\n[quote=\"Solution 1\"]Directly divide $x^{23}+23x^{17}-18x^{16}-24x^{15}+108x^{14}$ by $(x^4-3x^2-2x+9)$ to obtain $P(x)=x^{19}+ 3x^{17}+2x^{16}+12x^{14}$, so the sum of the coefficients is $1+3+2+12=\\fbox{18}$.[/quote]\n\nHowever, this problem originally appeared in an individual round, suggesting a faster way (although in most cases, people said the division process took them less than 5 minutes to do anyways):\n\n[i]Solution-Thanks to Elemennop[/i]\n\n[quote=\"Elemennop\"]Given, $\\displaystyle x^{23}+23x^{17}-18x^{16}-24x^{15}+108x^{14}=(x^4-3x^2-2x+9)\\cdot P(x)$,\n\nwe isolate $P(x)$ to one side.\n\n$P(x)=\\displaystyle\\frac{x^{23}+23x^{17}-18x^{16}-24x^{15}+108x^{14}}{x^4-3x^2-2x+9}$\n\nNow that we have $P(x)$, the sum of the coefficients of it is merely $P(1)$. Therefore,\n\n\n$P(1)=\\frac{1+23-18-24+108}{1-3-2+9}=\\frac{90}{5}=\\fbox{18}$.[/quote]",
"Solution_3": "Wow, that really is a much better solution.",
"Solution_4": "[size=150][color=black]Day 4[/color][/size]\r\nPoint value: 6\r\n\r\nNo calculators! :P\r\n\r\nThe number $85^9-21^9+6^9$ is divisible by an integer between $2,000$ and $3,000$. Compute that integer.\r\n\r\n\r\nThis probably was originally #8 on an ARML individual round. Most people used guess and check to quickly find that 5 and 7 divided this expression (thanks to Rep123Max):\r\n\r\n[quote=\"Rep123max\"]\nAnswer: $\\fbox{2240}$\n\nSolution: After realizing that the numbers are too big to go through all the factoring (in the normal, alloted time), I began to search for values that were divisible by it. \n\nFor some reason, I started with seeing if 7 divides the equation (probably because 21 was sticking out at me): $85^9-21^9+6^9\\equiv (1)^9-(0)^9+(-1)^9\\equiv 0\\pmod 7$. So we now have one divisor. Noticing that seeing if 49 divides it does not look promising, I moved downwards to 5.\n\n$85^9-21^9+6^9\\equiv (0)^9-(1)^9+(1)^9\\equiv 0\\pmod 5$. Again, noticing that 25 does not look to promising (although, I did try it for about a minute), we move down to 2.\n\n\n$85^9-21^9+6^9\\equiv (1)^9-(1)^9+(0)^9\\equiv 0\\pmod 2$. I continued with powers of 2 because it seemed easy to work with:\n\n$85^9-21^9+6^9\\equiv (1)^9-(1)^9+(0)^9\\equiv 0\\pmod 4$.\n$85^9-21^9+6^9\\equiv (-3)^9-(-3)^+(0)^9\\equiv 0\\pmod 8$.\n$85^9-21^9+6^9\\equiv (5)^9-(5)^9+(0)^9\\equiv 0\\pmod {16}$.\n$85^9-21^9+6^9\\equiv (21)^9-(21)^9+(0)^9\\equiv 0\\pmod {32}$.\n$85^9-21^9+6^9\\equiv (21)^9-(21)^9+(0)^9\\equiv 0\\pmod {64}$.\n\nWe can stop there, we have that $2^6\\cdot5\\cdot7=2240|(85^9-21^9+6^9)$, so we are done.[/quote]\n\nAlthough his solution took less than 5 minutes anyways, the \"most elegant\" solution was algebraic manipulation:\n\n[quote=\"Solution 2\"]\nConsider the general case $a^k-b^k+c^k$.\nSince $a^k-b^k$ is divisible by $a-b$, then $a^k-b^k+c^k$ is divisible by the greatest common factor of $a-b$ and $c^k$. Hence, our expression is divisible by $GCD(85-21,6^9)=GCD(64,6^9)=2^6$.\nBy the same token $GCD(21-6,85^9)=GCD(15,85^9)=5$ is a factor.\nAlso, for odd $k$, the expression $a^k+c^k$ is divisible by $a+c$, so $GCD(85+6,21^9)=7$ must also be a factor. Since all these divisors are relatively prime, $85^9-21^9+6^9$ is divisible by $2^6\\cdot 5\\cdot 7=\\fbox{2240}$.\n[/quote]\r\n\r\nProps to those who jumped on that immediately.",
"Solution_5": "[size=150][color=darkblue]Day 5[/color][/size]\r\n\r\n3 points this time, and still no calculator...but at least we get a break from too much more algebra. :)\r\n\r\nYou have 121 marbles, some of which are red, some white, and the rest blue. You also have 10 jars. If all the marbles are distributed into the jars, there [i]must[/i] be a jar with at least $n$ marbles of the same color. Compute the maximum possible value for $n$.\r\n\r\n\r\nThis problem was originally a Team question #1 on ARML. Apparently, what the question was asking for was confusing, since only 11 out of 17 people got it overall and only around 5 on their first try. It was basically two applications of Pigeonhole, but several people misread the number of marbles or jars. Anyways, here's a solution (Thanks to mathfanatic):\r\n\r\n[quote=\"mathfanatic\"]\nAnswer: $\\fbox{5}$\n\nSolution:\n\nSince 121/3 = 40.333, at least 41 marbles must have the same color.\n\nSince 41/10 = 4.1, at least 5 marbles of the same color must be in some jar.\n\nCheck that this is minimum: take {5R,4W,4B} and 9 sets of {4R,4W,4B}. \n\nAlternatively, we can think of the problem as having thirty jars containing 121 marbles of the same color. Again, the answer is 5.[/quote]\n\nOr we could pigeonhole jars first:\n\n[quote=\"Solution 2\"]\nSince there are 121 marbles to put into 10 jars, there must be at least one jar with at least 13 marbles. In that jar, since there are 3 total different colors for marbles, there must be at least one color with at least 5 marbles.\nThus, the answer is $\\fbox{5}$\n[/quote]\r\n\r\nGood job to everyone who got it!",
"Solution_6": "[size=150][color=green]Day 6[/color][/size]\r\n\r\nNo calculators, though not sure how they could help anyways.\r\n\r\nWorth 7 points.\r\n\r\nIn order to list the integers from 0 through 7 in binary notation [0, 1, 10, 11, 100, 101, 110, 111], we must write twelve 1's. Compute the number of 1's needed to list, in binary notation, the integers from 0 through 1023.\r\n\r\n\r\nSo, as expected, there were many different approaches to this problem:\r\n\r\n[quote=\"Elemennop\"]We can see that if we order all the numbers in base $2$ into a certain order, there is symmetry between one subset and the subset of the next group. Specifically, between the subset of all the numbers between $2^n$ and $2^{n+1}-1$, inclusive, has the same number of $1's$ written as between $0$ and $2^n-1$, plus an additional $n$ $1's$ at the beginning of the number (This is obvious for smaller cases of $n$, such as $n=1,2,3$).\n\nWe can define this recursively: Define $f(n)$ as the number of $1's$ written to list out all integers between $2^n$ and $2^{n+1}-1$. By our previous definition, $f(n)$ is defined as:\n\n$f(n)=f(1)+f(2)+...+f(n-1)+2^n$\n\nNow we can quickly list out all the ordered pairs, getting the next $f(n)$ by adding up all the terms in the f(n) part of the ordered pair, then adding $2^n$. We are looking for everything up to $f(9)$ (The $1's$ between $512$ and $1023$ in base $2$).\n\n$(0,1)$\n$(1,3)$\n$(2,8)$\n$(3,20)$\n$(4,48)$\n$(5,112)$\n$(6,256)$\n$(7,576)$\n$(8,1280)$\n$(9,2816)$\n\nNow to find our answer, we merely add up $f(1)+f(2)+...+f(9)=2f(9)-2^9$ (because $f(9)-2^9=f(1)+f(2)+...+f(8)$ by our formula) $=2(2816)-512=5632-512=\\fbox{5120}$.\n\nNote: I tried to find some general formula for $f(n)$ in terms of $f(0)$, but ran into some problems. If anyone else wants to follow up on this, feel free to.[/quote]\n\n\n[quote=\"alan\"]im fairly certain this method works. try smaller cases to notice a pattern in the amount of \"1\"s used. indeed, it begins to resemble binomial expansion very quickly.\n\n$0-0 \\rightarrow 0$\n$0-1 \\rightarrow 0 + 1$\n$0-3 \\rightarrow 0 + 2(1) + 2$\n$0-7 \\rightarrow 0 + 3(1) + 3(2) + 3$\n$\\vdots$\n$0-(2^n-1) \\rightarrow \\binom{n}{0}(0) + \\binom{n}{1} (1) + \\binom{n}{2} (2) \\dots + \\binom{n}{n-1}(n-1) + \\binom{n}{n}(n)$\n\n$1023=2^{10}-1$\n\n$=\\binom{10}{0}(0) + \\binom{10}{1} (1) + \\binom{10}{2} (2) \\dots + \\binom{10}{9}(9) + \\binom{10}{10}(10)$\n$=(0+10) + \\binom{10}{1}(1+9) + \\binom{10}{2}(2+8) + \\binom{10}{3}(3+7) + \\binom{10}{4}(4+6)+\\binom{10}{5}(5)$\n$=5*2^{10}=\\boxed{5120}$\n[/quote]\n\n[quote=\"djshowdown2\"]When dealing with binary, powers of two are always useful, so we see that $1023 = 2^{10} - 1$ and is therefore nine 1's in binary. So I listed out the first couple numbers in reverse order and recognized the pattern within the digits (first digit cycles through $2^0$ 1's then $2^0$ 0's, second digit is $2^1$ 1's and $2^1$ 0's, third is $2^2$ and so on). \n111111111\n111111110\n111111101\n111111100\n111111011\n111111010\n\nif this pattern is continued through 000000000, there are 1024 rows and 10 columns. Clearly half of the digits in the each column are 1's, so the total is $\\frac{1}{2}1024 \\times 10=\\fbox{5120}$[/quote]\n\n[quote=\"fanaticsm\"]Day 6:\n\nThe answer is $\\boxed{5120}$\n\nFor each binary number $a$ there is a complimentary number $b$, such that $a+b=1023$. The reason for calling them complimentary number is because their binary representations completes each to other to 10 units Thus, the answer to the question is $512\\cdot 10=5120$\n\n[b]Note:[/b] If you have no idea as to what I mean when I say 'complimentary', here is an example:\n\nConsider $a=618$ and $b=405$, and then compare the binary representation for $a$ and $b$: \n\n$1001101010=a=618$\n$+$\n$0110010101=b=405$\n$=$\n$1111111111=1023=a+b$[/quote]\n\n\n[quote=\"paladin8\"]ANSWER: $\\fbox{5120}$\n\nWe notice that if we write all the numbers in binary from $0$ to $2^n-1$, exactly half of the numbers have a $1$ as the last digit, half have $1$ as the second to last digit, and so on. $2^n-1$ has $n$ digits and there are $2^n$ total numbers. So $2^{n-1}$ of them have a $1$ in a certain spot. Thus there are $n(2^{n-1})$ total $1$'s. Since $1024=2^n$, $n=10$ and we have $10(2^9)=5120$ appearances of $1$.[/quote]\r\n\r\npaladin8 has the general formula, $n(2^{n-1})$.",
"Solution_7": "[size=150][color=indigo]Day 7[/color][/size]\r\n\r\nNo calculators, 4 points.\r\n\r\nCompute $x$ if $\\text{Arctan }x+\\text{Arctan }1=2\\left(\\text{Arctan }x-\\text{Arctan }\\frac{1}{3}\\right)$.\r\n\r\n\r\nThis problem came from a team round, so don't worry about it if you spent a lot more time getting your hands dirty with the algebra. Some people immediately took the tangent of both sides, but the undistributed 2 on the right hand side makes for an ugly cubic. Instead, if the terms were collected first, the work is a lot easier.\r\n\r\n[quote=\"paladin8\"]ANSWER: $\\boxed{7}$\n\nRearrange to get $\\arctan{x}=\\arctan{1}+2\\arctan{\\frac{1}{3}}$.\nSo $x=\\tan{(\\arctan{1}+2\\arctan{\\frac{1}{3}})}$.\nUsing the tangent sum formula,\n$x=\\frac{\\tan{(\\arctan{1})}+\\tan{(2\\arctan{\\frac{1}{3}})}}{1-\\tan{(\\arctan{1})}\\tan{(2\\arctan{\\frac{1}{3}})}}$.\nWe use the tangent sum formula again to evaluate $\\tan{(2\\arctan{\\frac{1}{3}})}$.\n$\\tan{(2\\arctan{\\frac{1}{3}})}=\\frac{2\\tan{(\\arctan{\\frac{1}{3}})}}{1-\\tan^2{(\\arctan{\\frac{1}{3}})}}=\\frac{\\frac{2}{3}}{1-\\frac{1}{9}}=\\frac{3}{4}$\nPlugging that back in we get $x=\\frac{1+\\frac{3}{4}}{1-\\frac{3}{4}}=7.$\n[/quote]\n\nAlso, here's another approach that considers an actual triangle:\n\n[quote=\"alan\"]let $y=\\arctan{x}, z=\\arctan{\\frac{1}{3}}$\n$y+45=2(y-z)$\n$45+2z=y$\ndraw triangle for $z$\n$\\sin{z}=\\frac{1}{\\sqrt{10}}, \\cos{z}=\\frac{3}{\\sqrt{10}}$\n$\\sin{2z}=2\\sin{z}\\cos{z}=\\frac{3}{5}$\nso it forms a 3-4-5\n$x=\\tan{y}=\\tan{(45+2z)}=\\frac{\\tan{45}+\\tan{2z}}{1-\\tan{45}\\tan{2z}}=\\frac{1+\\frac{3}{4}}{1-(1)(\\frac{3}{4})}=\\boxed{7}$[/quote]\r\n\r\nAgain, thanks to all those who provided solutions. :)",
"Solution_8": "[size=150][color=red]Day 8[/color][/size]\r\n\r\n5 points and no calculator\r\n\r\nCompute the length of the tangent segment from the origin to the circle that passes through the points (3, 4), (6, 8), and (5, 13).\r\n\r\n[quote=\"djshowdown2\"]If you draw a coordinate plane and plot the three points, you can see the general curve of circle, so put a center at point $(h,k)$. Draw 3 segments: a radius to the point of tangency, the tangent from the origin, and a segment connecting the origin and the center of the circle, forming a right triangle. The length of the hypotenuse is $h^2+k^2$ and the radius is $r$ so calling the length of the tangent segment $x$ you can write the equation $h^2 + k^2 = r^2 + x^2$ by Pythagorean Theorem.\n\nDescribe the circle based on two different points. $(3,4)$ gives you $k^2+h^2=r^2+8h+6k-25$ and $(6,8)$ gives $k^2+h^2=r^2+16h+12k-100$ so you have $x^2 = 8h+6k-25 = 16h +12k - 100$ so $8h+6k = 75$ and thus $x^2 = 75 - 25 = 50$ so $x=\\boxed{5\\sqrt{2}}$[/quote]\n\n[quote=\"paladin8\"]ANSWER: $\\boxed{5\\sqrt{2}}$\n\nConsider the line from the origin through $(3,4)$ and $(6,8)$ and the tangent line of length $x$. The distance from the origin to $(3,4)$ is $5$ and the distance to $(6,8)$ is $10$. By Power of a Point, $(5)(10)=x^2 \\Rightarrow x=5\\sqrt{2}$.[/quote]",
"Solution_9": "[size=150][color=violet]Day 9[/color][/size]\r\n\r\nNo calculators and 6 points.\r\n\r\nLet $x_n+iy_n=(1+i\\sqrt{3})^n$, where $x_n$ and $y_n$ are real and $n$ is a positive integer. If $x_{19}y_{91}+x_{91}y_{19}=2^k\\sqrt{3}$, compute $k$\r\n\r\n\r\nThis problem was from a team round, so no worries if it took a little longer to think of. The core of the problem seemed to be DeMoivre's Theorem for everybody, and it was pretty much straightforward from then on.\r\n\r\nHere's one solution in the complex plane (Thanks to djshowdown2):\r\n\r\n[quote=\"djshowdown2\"]Write $1+i\\sqrt{3}$ as the polar coordinate $(2,\\frac{\\pi}{3})$. Using that $(r,\\theta)^n = (r^n,n\\theta)$, \n\n$(2,\\frac{\\pi}{3})^{19} = (2^{19},\\frac{19\\pi}{3}) \\Rightarrow (2^{19},\\frac{\\pi}{3}) \\Rightarrow 2^{18} + 2^{18}\\sqrt{3}i$\n\n$(2,\\frac{\\pi}{3})^{91} = (2^{91},\\frac{91\\pi}{3}) \\Rightarrow (2^{91},\\frac{\\pi}{3}) \\Rightarrow 2^{90} + 2^{90}\\sqrt{3}i$\n\nso the value in question is \n\n${2^{18}2^{90}\\sqrt{3} + 2^{90}2^{18}}\\sqrt{3}$\n$(2^{108} + 2^{108})\\sqrt{3}$\n$2(2^{108})\\sqrt{3}$\n$2^{109}\\sqrt{3} = 2^k\\sqrt{3}$\n$\\framebox[.5 in]{k=109}$[/quote]\n\nAnd another approach involving cis (Thanks to Elemennop):\n\n[quote=\"Elemennop\"]$(1+i\\sqrt{3})$ can be rewritten as $2\\displaystyle\\left(\\frac{1}{2}+\\frac{\\sqrt{3}}{2}i\\right)$. Conveniently, the second expression is $\\cos60+i\\sin60$, or just $\\text{cis}60$. Thus, $x_n+iy_n=(2)^n(\\text{cis}60)^n$.\n\nNow that we've simplified the first expression, we can rewrite what we're looking for as:\n\n$\\text{Re}[(2^{19})(\\text{cis}60)^{19}]\\cdot\\text{Im}[(2^{91})(\\text{cis}60)^{91}] + \\text{Re}[(2^{91})(\\text{cis}60)^{91}]\\cdot\\text{Im}[(2^{19})(\\text{cis}60)^{19}]$\n\n$2^{110}\\text{Re}(\\text{cis}1140)\\cdot\\text{Im}(\\text{cis}5460)+2^{110}\\text{Re}(\\text{cis}5460)\\cdot\\text{Im}(\\text{cis}1140)$\n\nSince we can subtract any measure of $360k$ from the value of $\\text{cis}$, we do this to the two expressions to see that are both $60$! How nice.\n\n$2^{110}\\cdot\\frac{1}{2}\\cdot\\frac{\\sqrt{3}}{2} + 2^{110}\\cdot\\frac{1}{2}\\cdot\\frac{\\sqrt{3}}{2}$\n\n$2^{108}\\sqrt{3} + 2^{108}\\sqrt{3} = 2^{109}\\sqrt{3}$\n\nTherefore, the value of $k$ is $\\fbox{109}$.[/quote]",
"Solution_10": "[size=150][color=olive]Day 10[/color][/size]\r\n\r\nLet $x$ be a randomly selected integer from the set $\\{100,101,\\ldots,999\\}.$ What is the probability that $x^2$ and $(x+100)^2$ have the same number of digits?\r\n\r\nAll the solutions were basically variations of this one (thanks to fanaticism):\r\n\r\n[quote=\"fanaticism\"]\nThe probability is $\\boxed{\\frac{7}{9}}$\n\nThe squares of three-digit numbers have either 5 or 6 digits. If $10,000 \\leq x^2 < 100,000,$ then $100 \\leq x \\leq 316$ and $x$ and $x+100$ both lie in this range if $x \\in \\{100, 101, ..., 216\\}.$ If $100,000 \\leq x^2 < 1,000,000,$ then $317 \\leq x \\leq 999$ and $x$ and $x+100$ both lie in this range if $x \\in \\{317, 318, ..., 899\\}$. Out of 900 three-digit numbers, there are $(216-100+1) + (899-317+1) = 117 + 583 = 700$ values of $x$ satisfying the condition stated in the problem.\n\nTherefore, the required probability is $\\frac{700}{900} = \\frac{7}{9}$.[/quote]",
"Solution_11": "[size=150][color=orange]Day 11[/color][/size]\r\n\r\nPart 1 (6 pts):\r\n\r\nProve that if $n=\\sqrt{a}+\\sqrt{b}+\\sqrt{c}$ for integers n, a, b, and c, then a, b, and c must be perfect squares.\r\n\r\nPart 2 (4 pts):\r\n\r\nUsing Part 1 or otherwise, prove that if $T=T(a,b,c)$ is a right triangle with integer sides, then the perimeter of $\\sqrt{T}=(\\sqrt{a},\\sqrt{b},\\sqrt{c})$ cannot be an integer.\r\n\r\n[quote=\"qwertyq\"]\n\\[\\displaystyle n=\\sqrt{a}+\\sqrt{b}+\\sqrt{c}\\]\n\\[n-\\sqrt{c}=\\sqrt{a}+\\sqrt{b}\\]\nsquare both sides and rearrange to get\n\\[n^2+c-a-b=2\\sqrt{a}\\sqrt{b}+2n\\sqrt{c}\\]\n\\[n^2+c-a-b=2(\\sqrt{a}\\sqrt{b}+n\\sqrt{c})\\]\nThe left side is obviously an integer since $n$,$a$,$b$, and $c$ are integers\ntherefore the right side has to be an integer\nfor that to happen $\\sqrt{a}\\sqrt{b}$ has to be an integer and $n\\sqrt{c}$ has to be an integer\nIgnore the first expression for now\nFor the second expression to be an integer $c$ has to be a perfect square since $n$ is an integer that multiplys it. If $c$ were anything other than a perfect square then that expression would not be an integer.\nSimilarly for $a$ and $b$ we can rearrange the original to make\n\\[n^2+a-b-c=2\\sqrt{b}\\sqrt{c}+2n\\sqrt{a}\\] and\n\\[n^2+b-c-a=2\\sqrt{c}\\sqrt{a}+2n\\sqrt{b}\\]\nThus, $a$, $b$, and $c$ are all perfect squares. [/quote]\n\n[quote=\"Rep123Max\"]\nThis is easily proved by contradiction. Supposing that the perimeter of \\sqrt{T}=T(\\sqrt{a}, \\sqrt{b}, \\sqrt{c}) is an integer, then a=d^2, b=e^2 and c=f^2 for integers d, e, and f. Since the trianlge is right, a^2+b^2=c^2\\Longrightarrow d^4+e^4=f^4 which is not true for integral d, e, and f. Thus, the perimeter is not integral.[/quote]",
"Solution_12": "Solution to problem I posted:\r\n\r\n[quote=\"Elemennop\"]The obvious way to see how many solutions there are is to see how many intersections there between the graphs of the two equations (this allows us to not actually have to find the solutions $(x,y)$).\n\nSo, the first equation is $\\sin(x+y)=\\cos(x+y)\\Longrightarrow\\tan(x+y)=1$. A tan function has a value of 1 at values of $\\frac{\\pi}{4}+k\\pi$ for all integer values of k.\n\nLooking at the second equation, we see it's a circle with center at $(0,0)$ and a radius of $\\frac{1995\\pi}{4}$.\n\nWe essentially have the graph of a circle with a large radius, and an infinite amount of lines with equation of $x+y=\\frac{(4k+1)\\pi}{4}\\Rightarrow y=-x+\\frac{(4k+1)\\pi}{4}$ for all integers k.\n\nEach line (excluding tangents) will make two intersections with the circle. Thus, we only need to find how many lines will go inside the circle, then multiply that by two and account for any tangents that might occur.\n\nTo see how many lines appear, we must find what two lines are the tangent lines of the circle. These two lines will have equation y=-x+m for some real number m. All equations of that form make an $45^{\\circ}-45^{\\circ}-90^{\\circ}$ triangle with the coordinate axes.\n\nThus, imagine quadrant I of the graph. There will be a quarter circle, and some line l that is tangent to the circle and has the form mentioned above. Because it's an isosceles triangle, it's tangent from the center of the circle will be both the altitude of the triangle from the origin as well as the median.\n\nKnowing that the median of a right triangle will split the hypotenuse into two equal parts, each with length equal to the median, and seeing that the median is the radius of the circle, $m=\\frac{1995\\pi}{4}=\\frac{1}{2}h$, and the hypotenus has length $\\frac{1995\\pi}{2}$. Since it's base angles are $45^{\\circ}$, it's legs have length $\\frac{1995\\sqrt{2}\\pi}{4}$. Thus, any of our lines such that $|m|>\\frac{1995\\sqrt{2}\\pi}{4}$ will be outside the circle (and since the $\\sqrt{2}$ is there, it's clear we won't have any tangent lines formed from our first equation).\n\nNow we need to see what's the highest k we can get. $1995\\cdot\\sqrt{2}$ is approximately equal to 2821 and some fractional part. Thus, all $(4k+1)\\le2821$ or $(4k+1)\\ge2821$ will be within the circle and have two intersection points. Our answer is twice however many integer solutions there are to $|4k+1|\\le2821$. Solving this, we see $-705.5\\le k\\le705$.\n\nTherefore, the answer is $2(705+705+1)=2(1411)=\\fbox{2822}$.[/quote]\n\n[quote=\"fanaticism\"]\n\\begin{eqnarray*} \n\\sin{(x+y)} - \\cos{(x-y)} = 0 &\\rightarrow& \\left(\\sin{x} \\cdot \\cos{y} + \\cos{x} \\cdot \\sin{y}\\right) - \\left(\\cos{x} \\cdot \\cos{y} + \\sin{x} \\cdot \\sin{y}\\right) = 0\\\\ &\\rightarrow& (\\sin{x})(\\cos{y} - \\sin{y}) - (\\cos{x})(\\cos{y} - \\sin{y}) = 0\\\\ &\\rightarrow& (\\sin{x} - \\cos{x})(\\cos{y} - \\sin{y}) = 0\\\\\n\\end{eqnarray*}\n\nTherefore, $\\sin{x} = \\cos{x}$ or $ \\sin{y} = \\cos{y}$\n\nTherefore, $x = \\frac{\\pi}{4} + n\\pi$ or $y = \\frac{\\pi}{4} +n\\pi$, where n is an integer. The radius of the circle is $\\frac{1995\\pi}{4}$. For $ x \\geq 0$, there are vertical lines at $\\left\\{x = \\frac{\\pi}{4}, \\frac{5\\pi}{4}, ... , \\frac{1993\\pi}{4}\\right\\}$, for a total of 499 lines intersecting the circle at two points, making 998 points of intersections. Similarly, for $y \\geq 0$, there are 499 vertical lines intersecting the circle at 998 points. For $x < 0$, there are vertical lines at $\\left\\{x = -\\frac{3\\pi}{4}, -\\frac{7\\pi}{4}, ... , -\\frac{1995\\pi}{4}\\right\\}$. There are $\\frac{1995-3}{4} + 1$, or 499 vertical lines, 498 of which intersect the circle at two points, the other is a tangent.\n\nThus, there are 997 points of intersection. Similarly for y < 0, there are 997 points of intersections. The total number of points of intersections, or rather, solutions equals $2(998) + 2(997) = 3990$.[/quote]",
"Solution_13": "so you get the same answer both ways...?",
"Solution_14": "In the ARML book, it actually reads:\r\n$\\sin (x+y) = \\cos (x+y)$\r\n$x^2+y^2 = \\left(\\frac{1995\\pi}{4}\\right)^2$\r\n\r\nAnd in the provided solutions:\r\n$\\sin(x+y)-\\cos(x-y) = 0$\r\n$(\\sin{x}\\cdot\\cos{y}+\\cos{x}\\cdot\\sin{x})-(\\cos{x}\\cdot\\cos{y}+\\sin{x}\\cdot\\sin{y}) = 0$\r\n\r\nSo the intended problem was probably:\r\n$\\sin (x+y) = \\cos (x-y)$\r\n$x^2+y^2 = \\left(\\frac{1995\\pi}{4}\\right)^2$\r\n\r\nThe actual problem is a lot harder, unless I'm missing something really obvious or I messed up somewhere down the line:\r\n$\\sin (x+y) = \\cos (x+y) \\rightarrow \\tan(x+y)=1$\r\n$\\displaystyle x+y=\\frac{\\pi}{4}+n\\pi$\r\nThis equation describes lines that are separated by a perpendicular distance of $\\frac{\\pi}{\\sqrt{2}}$.\r\n$x^2+y^2 = \\left(\\frac{1995\\pi}{4}\\right)^2$ is a circle with diameter $\\frac{1995\\pi}{2}$, enough distance for at least $\\displaystyle\\frac{\\left(\\frac{1995\\pi}{2}\\right)}{\\left(\\frac{\\pi}{\\sqrt{2}}\\right)}\\approx1410.67\\rightarrow1410$ of these lines. If any of these lines are tangent to the circle, it must be either at $(\\frac{1995\\sqrt{2}}{8},\\frac{1995\\sqrt{2}}{8})$ or at $(-\\frac{1995\\sqrt{2}}{8},-\\frac{1995\\sqrt{2}}{8})$.\r\nHowever, neither of these points satisfies $\\displaystyle x+y=\\frac{\\pi}{4}+n\\pi$ for integral $n$, so there are no tangent lines to the circle.\r\nThus, there should be $1410*2=\\fbox{2820}$ solutions to the system since each intersecting line produces two solutions.\r\n\r\nI found this problem pretty hard to do without calculator, so you're probably right Max. The original problem should read:\r\n\r\n\r\n\r\nFor $x$ and $y$ in radians, compute the number of solutions in ordered pairs $(x,y)$ to the following system:\r\n\r\n\\begin{eqnarray*}\r\n\\sin (x+y) &=& \\cos (x-y) \\\\\r\nx^2+y^2 &=& \\left(\\frac{1995\\pi}{4}\\right)^2\r\n\\end{eqnarray*}",
"Solution_15": "[quote=\"towersfreak2006\"]Yes, you can always work on a problem until the solutions are posted. And the day 44 problem was really tough, as only three people got the right answer. I also stole it from joml88's book :) and the recent years seem to be more tricky.\n\n[size=150][color=indigo]Day 46[/color][/size]\n\nNo calculators, 6 points\n\nPoint $P$ is inside a rectangular box. The distances from point $P$ to four of the vertices of the box are $1, 2, 3,$ and $4$. If the distance from $P$ to another vertex is greater than $5$, find that distance.[/quote]\n\n[quote=\"ARML Solution\"]\nThis solution is based on the easily proved theorem that given rectangle $ABCD$ and point $P$, then $PA^2+PC^2=PB^2+PD^2$; this holds even when $P$ is not in the plane of the rectangle! Let the vertices of one face of the box be $A$, $B$, $C$, and $D$, and the diagonally opposite vertices (with respect to the center of the box) be $A'$, $B'$, $C'$, and $D'$ respectively. To simplify notation, let the distance $PA$ be represented simply by $A$, $PB$ by $B$, etc. Thus we know that $A^2+C^2=B^2+D^2$, and (from rectangle $ABA'B'$) $A^2+A'^2=B^2+B'^2$, etc.\n\n1. Let $A[=PA]=1$. Then either $2=B$ or $2=C$ $[2%Error. \"neqA\" is a bad command.\n'$ since $A^2+A'^2=1^2+2^2<3^2$+anything, so no distance from $P$ to another vertex can equal to $3]$. Suppose $2=C$. Then $3\\neq A$, $C$, $B$, $D$, $A'$, or $C'$. Therefore $3=B'$ or $D'$ (symmetrical situation). Let $3=B'$. Then $4\\neq A$, $C$, $B$, $D$, $A'$, $C'$, or $B'$, so $4=D'$. But we cannot accommodate a distance greater than $5$. Therefore $2=B$.\n\n2. Now $3\\neq A$, $B$, $A'$, or $B'$, so either $3=C$ $[$or $D']$ or $3=D$ $[$or $C']$. Suppose $3=C$. Then we have $D=\\sqrt{6}$. Now $4\\neq A'$, $B'$, $C'$, OR $D'$, since we could not accommodate a distance greater than $5$. Therefore $3=D$.\n\n3. Finally, we have $C=\\sqrt{12}$ and $4=C'$ $[$this is the only way to get a pair of \"opposite points\" whose squares total more than $25]$. Then $'D=\\sqrt{19}$, $B'=\\sqrt{24}$, and $A'=\\sqrt{27}$ or $3\\sqrt{3}$. Note: It can be shown that such a box [i]does[/i] exist.\n[/quote]\n\nAlso, the following is an interesting solution that assumes the box to be a cube. I don't think it works, but then again, what do I know about geometry? Perhaps someone else can comment on it:\n\n[quote=\"tarquin\"]To make it convenient, put the point inside a cube. On a cartesian plane, the coordinates of the cube are then\n\n$(0,0,0)$\n$(0,a,0)$\n$(0,0,a)$\n$(a,0,0)$\n$(a,a,0)$\n$(a,a,a)$\n$(0,a,a)$\n$(a,0,a)$\n\nIf the coordinates of $P$ are $(m,p,q)$ using the distance formula we have\n\n$m^2 + p^2 + u^2 = 1$\n\n$m^2 + (p-a)^2 + u^2 =4$\n\n$m^2 + p^2 + (u-a)^2 = 9$\n\n$(m-a)^2 + p^2 + u^2 = 16$\n\nAdding them all up and subtracting $3m^2 + 3p^2 + 3u^2 = 3$ yields $(m-a)^2 + (p-a)^2 + (u-a)^2 = 27$\n\n$\\sqrt {27} = 3 \\sqrt 3$[/quote]",
"Solution_16": "Well if I wanted to type a little bit more than I could have put it in any arbitrary rectangular prism I'd just add on some value $x$ and I would have ended up with $(m - a - x)^2 + (p - a- x)^2 + (q - a -x)^2 = 27$\r\n\r\nEr I think.",
"Solution_17": "[quote=\"towersfreak2006\"][size=150][color=violet]Day 47[/color][/size]\n\nNo calculators, 6 points\n\nThere are several values for a prime, $p$, with the property that any 5-digit multiple of $p$ remains a multiple of $p$ under \"cyclic permutation\". One such value is $41$ (for example, since $50635$ is a multiple of $41$, so are $55063$, $35506$, $63550$, and $[0]6355$); another such value of $p$ is $3$. Compute the value of $p$ that is greater than $41$.[/quote]\n\n[quote=\"paladin8\"]ANSWER: $271$\n\nSo we know $p$ divides some number $abcde = 10^4a+10^3b+10^2c+10d+e$ and its permutations $bcdea$, $cdeab$, $deabc$, and $eabcd$. Thus it must divide the sum of these five numbers, $11111(a+b+c+d+e)$. Since we have $11111 = (41)(271)$, $p = 271$ is the only possible number.[/quote]",
"Solution_18": "Why has the answer been posted already? I got home from school, saw the question, had to go straight out, got the answer while I was out (and it's the correct answer, by the way), and I come back and the answer is posted? I don't think so. :dry: You didn't give nearly enough time for this, especially considering that this was a 6-point question.",
"Solution_19": "Since these are from before 1995, no use of calculator was permitted for any of these (for some obvious reasons...)\r\n\r\nCompute $\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}$.\r\n\r\n[quote]\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\sqrt{(\\frac{999,999,999,999}{9})(1,000,000,000,005)+1}$\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\sqrt{\\left[\\frac{10^{12}-1}{9}\\right]\\left[10^{12}+5\\right]+1}$\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\sqrt{\\frac{10^{24}+4\\cdot10^{12}+4}{9}}$\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\sqrt{\\frac{(10^{12}+2)^2}{9}}$\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\frac{10^{12}-1+3}{3}$\n$\\sqrt{(111,111,111,111)(1,000,000,000,005)+1}=\\frac{999,999,999,999+3}{3}=\\boxed{333,333,333,334}$\n[/quote]\n\nLet $z$ and $z+1$ be complex $n$th roots of $1$. If $n$ is a positive integer multiple of $5$, compute the smallest possible value for $n+z^3$.\n\n[quote]\nDrawing a unit circle around the origin $O$ in the complex plane, we know that points representing the $n$th roots of unity lie on this circle and are the vertices of a regular $n$-gon. The points $P$ and $Q$, representing the complex numbers $z$ and $z+1$ respectively, lie on this circle; $PQ$ is parallel to the x-axis and $PQ=1=OP=OQ$, so angle $POQ=60^\\circ$. Thus, additional vertices must lie at $A(1,0), B(-1,0)$ and at points $P'$ and $Q'$ representing $-z$ and $-(z+1)$ respectively, producing the six vertices of a regular hexagon. Since $P, A, Q'$ represent the cube roots of $1$, then $z^3=1$. If there are additional vertices along arc $PQ$, there must be corresponding vertices along each of the other five arcs of the circle. Thus $n$ must be a multiple of $6$. Since it is given to be a multiple of $5$, the smallest possible value for $n$ is $30$, and $n+z^3=\\boxed{31}$.\n[/quote]\n\nCompute all positive integer values of $n$, $1AC)$ tangents to BC at E. AE meets (I) at E and D. Let $ F\\in AC$ suct that $ CF\\equal{}CE$. Let G be the intersection of CF and BD. Prove that $ CF\\equal{}FG$.",
"Solution_1": "I don't think that those problems are true :huh: \r\n\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=204652\r\nhttp://www.mathlinks.ro/Forum/viewtopic.php?t=204655",
"Solution_2": "I made a solution for solving the problem, i am sure that it is true and very nice.",
"Solution_3": "Should be: Let $ F \\in AE$ be such that $ CE \\equal{} CF.$\r\n_________________________________________\r\n\r\n$ U, V$ are tangency point of the incircle $ (I)$ with $ AB, AC.$ $ UV$, tangent to $ (I)$ at $ D,$ and tangent $ BC$ to $ (I)$ at $ E$ are polars with respect to $ (I)$ of the collinear points $ A, D, E,$ they are concurrent at $ K.$ The cross ratio $ \\frac {\\overline{KB}}{\\overline{KC}} \\cdot \\frac {\\overline{EC}}{\\overline{EB}} \\equal{} \\minus{} 1$ is harmonic (by Menelaus for the transversal $ UVK$), hence $ \\overline{EK} \\equal{} \\overline{EC} \\plus{} \\overline{CK} \\equal{} \\overline{EC} \\plus{} \\frac {\\overline{EC} \\cdot \\overline{BC}}{\\overline{BE} \\minus{} \\overline{EC}} \\equal{} \\frac {2\\ \\overline {BE} \\cdot \\overline{EC}}{\\overline{BE} \\minus{} \\overline{EC}}.$ The isosceles $ \\triangle KDE \\sim \\triangle CFE$ are centrally similar with center $ E$, $ CF \\parallel KD.$ Let a parallel to $ BD$ through $ F$ cut $ BC$ at $ M.$ The $ \\triangle MCF \\sim \\triangle BKD$ are centrally similar with center $ E$ and coefficient $ k \\equal{} \\frac {\\overline{EC}}{\\overline{EK}} \\equal{} \\frac {\\overline{BE} \\minus{} \\overline{EC}}{2\\ \\overline{BE}},$ hence $ \\overline{EM} \\equal{} k \\cdot \\overline{EB} \\equal{} \\minus{} \\frac {\\overline{BE} \\minus{} \\overline{EC}}{2}$ and $ \\overline{BM} \\equal{} \\overline{BE} \\plus{} \\overline{EM} \\equal{} \\frac {\\overline{BE} \\plus{} \\overline{EC}}{2} \\equal{} \\frac {\\overline{BC}}{2}.$ Thus $ M$ is the midpoint of $ \\overline{BC}.$ Since $ BG \\parallel MF,$ $ MF$ is a midline of the $ \\triangle BCG$ and $ F$ is the midpoint of $ \\overline{CG}.$",
"Solution_4": "[quote=\"mr.danh\"] [color=darkred]The incircle $ w \\equal{} C(I)$ of triangle $ ABC\\ \\ (c > b)$ touches it in the points $ D\\in BC$ , $ E\\in CA$ , $ F\\in AB$ .\n\nDenote $ \\{D,T\\} \\equal{} AD\\cap w$ and let $ K\\in AD$ be a point suct that $ CK \\equal{} CD$ . Denote $ L\\in CK\\cap BT$ . Prove that $ KC \\equal{} KL$.[/color] [/quote]\r\n\r\n[color=darkblue][b][u]Proof (similarly with Yetti's).[/u][/b] Denote the midpoint $ M$ of the side $ [BC]$ and $ N\\in EF\\cap BC$ . Observe easily that $ \\boxed {\\ MD \\equal{} \\frac {c \\minus{} b}{2}\\ }$ . \n\nFrom the well-known property $ IN\\perp \\overline {ATD}$ obtain $ N\\in TT$ . Thus, $ CK \\equal{} CD$ $ \\implies$ $ \\widehat {DKC}\\equiv\\widehat {CDK}\\equiv\\widehat {DTN}$ $ \\implies$ \n\n$ \\widehat {DKC}\\equiv\\widehat {DTN}$ $ \\implies$ $ KC\\parallel TN$ . Denote the point $ M_1\\in BC$ , $ KM_1\\parallel BT$ . Therefore, $ \\triangle M_1KC\\sim\\triangle BTN$ $ \\implies$ $ \\frac {DM_1}{DC} \\equal{} \\frac {DB}{DN}$ .\n\nSince $ \\frac {NB}{NC}\\equal{}\\frac {DB}{DC}\\equal{}\\frac {p\\minus{}b}{p\\minus{}c}$ ascertain easily that $ DN \\equal{} \\frac {2(p \\minus{} b)(p \\minus{} c)}{c \\minus{} b}$ . Thus, $ M_1D \\equal{} \\frac {DB\\cdot DC}{DN} \\equal{} (p \\minus{} b)(p \\minus{} c)\\cdot\\frac {c \\minus{} b}{2(p \\minus{} b)(p \\minus{} c)}$ , \n\ni.e. $ \\boxed {\\ M_1D \\equal{} \\frac {c \\minus{} b}{2}\\ }$ . In conclusion, $ M_1D \\equal{} MD$ , i.e. $ M_1\\equiv M$ $ \\implies$ $ KM\\parallel BT$ $ \\implies$ $ KC \\equal{} KL$ .[/color]",
"Solution_5": "We restate the notations of the problem, like to the [b][size=100]Virgil Nicula\u2019s[/size][/b] ones.\r\n\r\nWe denote as $ D,\\ E,\\ F,$ the tangency points of the incircle $ (I)$ of $ \\bigtriangleup ABC,$ with its side-segments $ BC,\\ CA,\\ AB$ respectively and let be the points $ T\\equiv (I)\\cap AD$ and $ K$ on $ AD,$ such that $ CD = CK.$\r\n\r\nThen, we must prove that $ CK = KL,$ where $ L\\equiv CK\\cap TB.$\r\n\r\n$ \\bullet$ Let $ S$ be, the intersection point of the sideline $ BC,$ from the tangent line of the incircle $ (I),$ at point $ T.$\r\n\r\nThe pencil $ T.SBDC,$ is in harmonic conjugation ( easy to prove ) as well and so, it is enough to prove that $ CK\\parallel TS.$\r\n\r\nBut it is true, because of $ \\angle CKT = \\angle CDK = \\angle SDT = \\angle STK.$\r\n\r\nSo, we conclude that $ CK = KL$ and the proof is completed.\r\n\r\nKostas Vittas."
}
{
"Tag": [
"geometry",
"number theory",
"greatest common divisor",
"invariant"
],
"Problem": "The vertices of $\\triangle ABC$ are lattice points on the xy-plane.\r\nIf both $AB$ and $AC$ contain odd numbers of lattice points (excluding the vertices themselves), prove that the number of lattice points on $BC$ (excluding the vertices) is also an odd number.",
"Solution_1": "this might have to do with the formula for area that gives the area of a triangle with lattice point verticies",
"Solution_2": "I would guess that Pick's theorem may be useful.",
"Solution_3": "Position the axes so that A is the origin. Suppose $B = (bp,bq)$ and $C = (cr,cs)$, where p,q and r,s are coprime.\r\n\r\nWe have the equation of the line AB to be $y = bqx/bp$. We are concerned with integer solns to $bpy = bqx$. Obviously (0,0) and (bp,bq) work. So clearly the equation reduces to $py = qx$ and we must have $x$ a multiple of $p$ etc. (since p,q coprime) so that we have solutions $kp, kq$ for k = 0...b. So an equivalent condition for AB to have odd # of lattice points is that the gcd of the ordinate and abscissa is even. (*)\r\n\r\n\r\nNow position the axes so B is the origin by a translation. Then the gcd of the ordinate and abscissa of C is even (since we have translated by an even number to both the ordinate and abscissa of C, keeping invariant their parity). But similar to (*) we deduce that BC has an odd # of lattice points."
}
{
"Tag": [
"geometry",
"rectangle",
"geometric transformation",
"reflection",
"homothety",
"trigonometry",
"IMO Shortlist"
],
"Problem": "Let $ ABC$ be a triangle and take three similar rectangles inscribed $ ABC$ (a rectangle inscribed $ ABC$ has two vertices on side a and two remaining vertices on two other sides of $ ABC$). Let $ A',B',C'$ be center of these rectangles (cyclic with $ A,B,C$), respectively.\r\n\r\na) Prove that $ AA',BB',CC'$ are concurrent.\r\n\r\nb) Let $ A'',B'',C''$ are reflections of $ A',B',C'$ through $ BC,CA,AB$, resp. Prove that $ AA'',BB'',CC''$ are concurrent.\r\n\r\nNote that, part a) is generalization of the problem on the post [url=http://www.mathlinks.ro/viewtopic.php?t=17463]IMO ShortList 2001, geometry problem 1[/url].",
"Solution_1": "$ \\bullet$ Let $ PQRS$ be the rectangle relative to $ A$ such that $ P \\in AB$ and $ Q \\in AC.$ Construct outwardly on $ BC$ the rectangle $ BCC_aB_a$ similar to $ PQRS.$ Homothety with center $ A$ and coefficient $ \\frac {_{AQ}}{^{AC}}$ takes $ PQRS$ into $ BCC_aB_a$ $\\Longrightarrow$ their centers $ A'$ and $ A_1$ are collinear with the homothetic center $ A$ $ \\Longrightarrow$ $ A_1 \\in AA'.$ Likewise, $ B_1,C_1$ lie on $ BB',CC'.$ \n\n$ \\triangle BCA_1,\\triangle CAB_1, \\triangle ABC_1$ are 3 similar isosceles triangles erected outwardly on the sides of $ \\triangle ABC.$ By Kiepert theorem, $ AA',BB',CC'$ concur at the outer Kiepert perspector relative to $ \\angle CBA_1 \\equal{} \\angle ACB_1 \\equal{} \\angle BAC_1 \\equal{} \\phi.$\n\n$ \\bullet$ Homothety with center $ A,$ taking $ PQRS$ into $ BCC_aB_a,$ takes the reflection $ A''$ of $ A'$ about $ BC$ into the reflection $ A_2$ of $ A_1$ about $ B_aC_a.$ Let $ M_a$ denote the midpoint of $ BC.$\n\n$ \\tan \\widehat{M_aCA_2} \\equal{} \\frac {M_aA_2}{M_aC} \\equal{} \\frac {3A_1M_a}{M_aC} \\equal{} 3\\tan\\phi$ \n\nTherefore, $ \\triangle A_2BC,\\triangle B_2CA$ and $ \\triangle C_2AB$ are three similar isosceles triangles erected outwardly on the sides of $ \\triangle ABC$ $ \\Longrightarrow$ $ AA'',BB'',CC''$ concur at the outer Kiepert perspector relative to $ \\tan^{ \\minus{} 1}(3 \\tan \\phi).$"
}
{
"Tag": [
"calculus",
"integration",
"logarithms",
"function",
"calculus computations"
],
"Problem": "a. Evaluate $ I_{1} \\equal{} \\int_{0}^{\\infty} \\frac{te^{\\minus{}t}}{1 \\minus{} e^{\\minus{}2t}} dt$\r\n\r\nb. Evaluate $ I_{2} \\equal{} \\int_{0}^{1} \\frac{t^{2} \\log t }{t^2 \\minus{} 1} dt$\r\n\r\nc. Find $ I_{3} \\equal{} \\int_{y\\equal{}0}^{\\infty} \\int_{x \\equal{}y}^{\\infty} \\frac{(x \\minus{} y)^{2}\\log \\left(\\frac{x \\plus{}y}{x \\minus{} y}\\right)}{xy \\sinh (x \\plus{}y)} dx dy$",
"Solution_1": "Here's $ I_1:$\r\n\r\nSince $ \\sum_{k\\equal{}0}^{\\infty}e^{\\minus{}2kt}\\equal{}\\frac1{1\\minus{}e^{\\minus{}2t}},$ we have \r\n\r\n$ I_1\\equal{}\\int_0^{\\infty}\\sum_{k\\equal{}0}^{\\infty}te^{\\minus{}(2k\\plus{}1)t}\\,dt\\equal{}\\sum_{k\\equal{}0}^{\\infty}\\int_0^{\\infty}te^{\\minus{}(2k\\plus{}1)t}\\,dt$\r\n\r\n$ \\equal{}\\sum_{k\\equal{}0}^{\\infty}\\frac1{(2k\\plus{}1)^2}\\int_0^{\\infty}ue^{\\minus{}u}\\,du\\equal{}\\sum_{k\\equal{}0}^{\\infty}\\frac1{(2k\\plus{}1)^2}$\r\n\r\nThis is a well-known sum closely related to $ \\zeta(2).$ Its value is $ \\frac34\\zeta(2)\\equal{}\\frac{\\pi^2}{8}.$",
"Solution_2": "$ I_{2}\\equal{}\\minus{}\\sum_{k\\equal{}0}^{\\infty }\\int_{0}^{1}t^{2k\\plus{}2}\\ln tdt\\equal{}\\sum_{k\\equal{}0}^{\\infty }\\frac{1}{\\left( 2k\\plus{}3\\right) ^{2}}\\equal{}\\allowbreak \\frac{\\pi ^{2}}{8}\\minus{}1$",
"Solution_3": "Thank you all for your solutions.\r\nBy setting $ x \\equal{} \\frac {u \\plus{} v}{2}$ and $ y \\equal{} \\frac{u \\minus{} v}{2}$ and then again setting $ v \\equal{} uw$ in the resultant integral reduce $ I_{3}$ to solving for $ I_{1}$ and $ I_{2}$.",
"Solution_4": "here i have a doubt \r\n\r\nIn the second problem it can also be reduced to $ I_{2}\\equal{}\\sum_{k\\equal{}0}^{\\infty }\\int_{0}^{1}t^{\\minus{}2k}\\ln tdt\\equal{}\\sum_{k\\equal{}0}^{\\infty }\\frac{\\minus{}1}{\\left( \\minus{}2k\\plus{}1\\right)^{2}}\\equal{}\\allowbreak\\frac{\\minus{} \\pi^{2}}{8}\\minus{}1$\r\n\r\nHow does this happen or am i wrong ?? please help me . :(",
"Solution_5": "The key in both of the solutions supplied by Kent and Indochina is to use the geometric series. So if your answer differs from the correct answers already supplied, then just show your work and someone will point out an incorrect step.",
"Solution_6": "vijaymenon: note that negative powers of $ t$ are greater than $ 1$ on the interval $ (0,1).$ Hence $ \\sum_{k\\equal{}0}^{\\infty}t^{\\minus{}2k}\\ln t$ is a divergent series on the interval in question. Furthermore, for $ k\\ge 1,$ $ \\int_0^1t^{\\minus{}2k}\\ln t\\,dt$ is a divergent integral.\r\n\r\nOne more note: on the interval in question, $ \\ln t<0$ and $ t^2\\minus{}1<0.$ Hence, $ \\frac{t^2\\ln t}{t^2\\minus{}1}>0$ and the result of the integral should be positive. (I would also add that the integrand tends to finite limits at both ends of the interval, and can be considered as if it were continuous on $ [0,1]$ - hence we can be sure of convergence of the integral.)\r\n\r\nIndoChina's computation is correct, and I don't see any plausible alternatives to his methods.",
"Solution_7": "[quote=\"Kent Merryfield\"]Here's $ I_1:$\n\nSince $ \\sum_{k \\equal{} 0}^{\\infty}e^{ \\minus{} 2kt} \\equal{} \\frac1{1 \\minus{} e^{ \\minus{} 2t}},$ we have \n\n$ I_1 \\equal{} \\int_0^{\\infty}\\sum_{k \\equal{} 0}^{\\infty}te^{ \\minus{} (2k \\plus{} 1)t}\\,dt \\equal{} \\sum_{k \\equal{} 0}^{\\infty}\\int_0^{\\infty}te^{ \\minus{} (2k \\plus{} 1)t}\\,dt$\n\n$ \\equal{} \\sum_{k \\equal{} 0}^{\\infty}\\frac1{(2k \\plus{} 1)^2}\\int_0^{\\infty}ue^{ \\minus{} u}\\,du \\equal{} \\sum_{k \\equal{} 0}^{\\infty}\\frac1{(2k \\plus{} 1)^2}$\n\nThis is a well-known sum closely related to $ \\zeta(2).$ Its value is $ \\frac34\\zeta(2) \\equal{} \\frac {\\pi^2}{8}.$[/quote]\r\nSorry, but I dont understand this:\r\n \r\n$ \\sum\\limits_{k\\equal{}0}^{\\infty }{\\int\\limits_{0}^{\\plus{}\\infty }{t.{{e}^{\\minus{}(2k\\plus{}1)t}}\\,dt}}\\equal{}\\sum\\limits_{k\\equal{}0}^{\\infty }{\\frac{1}{{{(2k\\plus{}1)}^{2}}}}.\\int\\limits_{0}^{\\plus{}\\infty }{u.{{e}^{\\minus{}u}}\\,du}$\r\nAnd I have not learned this follow formula yet. Please tell me about this function:\r\n \r\n$ \\sum\\limits_{k\\equal{}0}^{\\infty }{\\frac{1}{{{(2k\\plus{}1)}^{2}}}}\\equal{}\\frac{3}{4}\\zeta (2)\\equal{}\\frac{{{\\pi }^{2}}}{8}$",
"Solution_8": "The first line you quote is simply the ordinary substitution $ u\\equal{}(2k\\plus{}1)t$ or $ t\\equal{}\\frac{u}{2k\\plus{}1}.$\r\n\r\nAs for the computation of $ \\zeta(2)\\equal{}\\sum_{n\\equal{}1}^{\\infty}\\frac1{n^2}\\equal{}\\frac{\\pi^2}6:$ there are many different computations, and there are links to places where these computations have been collected together. It's quite a famous problem, and much about it has appeared on this site over the years. (Someone else can supply the links.)",
"Solution_9": "chuong01,\r\nIf you are familiar with double integrals, then doing the following exercises may help in the evaluation of the series in question. \r\n\r\nEvaluate: $ J_{1} \\equal{} \\int_{\\minus{}1}^{1}\\int_{\\minus{}1}^{1} \\frac{1}{1 \\plus{} 2xy \\plus{} y^2}dxdy$\r\n\r\nEvaluate: $ J_{2} \\equal{} \\int_{\\minus{}1}^{1}\\int_{\\minus{}1}^{1} \\frac{1}{1 \\plus{} 2xy \\plus{} y^2}dydx$"
}
{
"Tag": [
"MATHCOUNTS"
],
"Problem": "Is it just me, or is no one going to the mathcounts forum anymore. If so, I think everyone is flocking to this site (like me), cause it's so good. This site is cool 8-) !",
"Solution_1": "Mathcounts.org forum? What's the address?",
"Solution_2": "just go to http://www.mathcounts.org and then meeting place. but this site and forum is much better to me. after you go to meeting place there are hmmmmmmm i think 4 different forums you can go to. but one of them you have to be a coordinator of somat like that. if you post something it takes a few days because of screening.",
"Solution_3": "Yah, it like took 4 days once. Well, this site is WAY better because of instant posting, and smileys and all sorta stuff.",
"Solution_4": "one acronym. \"lol\"",
"Solution_5": "Only to Syntax : Your LoL is usually \"lool\". You always do that, but maybe your forgot it this time and made a Synatx Error. (Don't kill me Neal)",
"Solution_6": "Just forget about the last post. And it's so amusing there aren't any posts on that MathCounts.org forum, and there's like 50 new posts here. :lol:",
"Solution_7": "humph chinaboi. one more acronym \"lau\" i made it up. it stands for laughing at u. lool!",
"Solution_8": "What is \"lool\"?\r\n\r\nWhat is \"lolol\"?\r\n\r\nI'll stick to the unabbreviated words. :lol:",
"Solution_9": "It's supposed to stand for laughing out loud, but with these guys, you never know...\r\n\r\nMaybe \"laughing out of leprosy\"? sort of like, laughing out of happiness?",
"Solution_10": "Hey, it's Syntax. You don't really know what he can think of. Right Neal?",
"Solution_11": "yup yup yup. dipdidpdip, hoot hoot hoot. lool. lau. make sure to spread lau to whoever you see. lool. maybe i should do something productive. ha. lool. namaste jall",
"Solution_12": "\"Namaste jall\"? Does that mean\r\n\r\n[color=#CC3399]Not As Much As Some Thought[/color] [Syntax] [color=#CC3399]Error[/color]\r\n\r\n[color=#3399CC]Just A Little Left[/color]\r\n\r\n:?: :mrgreen: :lol:",
"Solution_13": "whoa. you have a lot of freetime on ure hands. how long did it take for you to think that up. well namaste in hindi means like hello goodbye greeting type of thing. and jall is just like yall which is just like you. so i basically said bye everyone. nuff said",
"Solution_14": "Now you see what I mean. Anyways, I'll make sure not to spread the \"lau\", Neal. See Ya later.",
"Solution_15": "Chinaboy, what's the whole point in this lol thing? Didn't I tell you lol or LOL, not LoL? Are you trying to get yourself in the crowd? I'll try lOl! :shock: I think this site is way better than MATHCOUNTS site! By the way, is Syntax Error, Neal?",
"Solution_16": "Yep, Syntax is indeed Neal, and Neal, splash1u27 is Amy.",
"Solution_17": "representin from the hood jall. anyways. im the normal one. its martin, you dont have to worry about him. hes just oddly. anyways. spread the acronym \"lau\" it rocks the hizhouse. yep yep. martin or chinaboi is the one who has a bedtime of like 8:30 pm. shame. some people just have no life. and hes a teenager. tisk tisk. well if you havent figured out by now im sorta mad at him because of his stupidity. anyways. nuff said",
"Solution_18": ":-x Hey! I went to bed late enough um... this morning... hee. I don't wanna go to bed that early. It's parents. Oh yah, I was at Neal's house at 7:30, and all the computer's were used up, so couldn't get on. Was Josh also on at 7:30??? Anyways, gtg. Hey Neal, look at the time I posted this. Way past \"my bedtime\". anyways, bye.",
"Solution_19": "wooooooooot. i only went to sleep about 6 hours later. so shup",
"Solution_20": "[quote=\"Ex274513\"]\"Namaste jall\"? Does that mean\n\n[color=#CC3399]Not As Much As Some Thought[/color] [Syntax] [color=#CC3399]Error[/color]\n\n[color=#3399CC]Just A Little Left[/color]\n\n:?: :mrgreen: :lol:[/quote]\r\n\r\nThats no abbreviation! That means \"Hello, a deception!\"....doesn't really make much sense. :?",
"Solution_21": "of course it doesnt make much sense. just like i end every post with nuff said. to stop my ranting. anywho. nuff said",
"Solution_22": "lau is outdated lolt kicks ***(*)\r\n\r\n5 posts to 1000 NEAL GO FOR IT WHURM!!",
"Solution_23": "[quote=\"sugarcookie34\"]Chinaboy, what's the whole point in this lol thing? Didn't I tell you lol or LOL, not LoL? Are you trying to get yourself in the crowd? I'll try lOl! :shock: I think this site is way better than MATHCOUNTS site! By the way, is Syntax Error, Neal?[/quote]\r\nI'm going with AOL :lol: \r\n\r\nUm, what's the plural of a*s anywho?"
}
{
"Tag": [
"vector",
"linear algebra"
],
"Problem": "For a curve $x=y=z^{2}$, the differential length vector $d\\bold{l}$ at a point $(x,y,z)$ is $dx\\bold{i_{x}}+dy\\bold{i_{y}}+2zdz\\bold{i_{z}}$.\r\n\r\nWhat is the physical meaning of $d\\bold{l}$?",
"Solution_1": "$2 z dz\\bold{i_{x}}+2 z dz\\bold{i_{y}}+dz\\bold{i_{z}}$.\r\n\r\nAre you sure you don't mean that? If not, I don't think I can help?",
"Solution_2": "oh yes... you are absolutely right! Sorry my bad. :oops:",
"Solution_3": "Well, do you understand it, or do you still need an explanation.\r\nI'm sorry but I rather refrain from doing work for nothing, it is so so SO hot here right now :roll: :wink:",
"Solution_4": "Please don't double post -- you posted this problem [url=http://www.artofproblemsolving.com/Forum/viewtopic.php?p=578216]here[/url] as well.",
"Solution_5": "[quote=\"JBL\"]Please don't double post -- you posted this problem [url=http://www.artofproblemsolving.com/Forum/viewtopic.php?p=578216]here[/url] as well.[/quote]\r\nI am sorry, but I posted this problem 4-5 days ago in the Calculus/Real Analysis forum. After not getting any response for so many days, I started thinking that maybe topic would be more appropriate for Matrix/Linear Algebra forum. It wasn't until today that anybody took the time to answer my question in the Calculus/Real Analysis forum. :(\r\n\r\nSorry for the trouble. :oops:"
}
{
"Tag": [
"geometry",
"circumcircle",
"perpendicular bisector",
"USAMO",
"power of a point"
],
"Problem": "Let ${\\cal C}_1$ and ${\\cal C}_2$ be concentric circles, with ${\\cal C}_2$ in the interior of ${\\cal C}_1$. From a point $A$ on ${\\cal C}_1$ one draws the tangent $AB$ to ${\\cal C}_2$ ($B\\in {\\cal C}_2$). Let $C$ be the second point of intersection of $AB$ and ${\\cal C}_1$, and let $D$ be the midpoint of $AB$. A line passing through $A$ intersects ${\\cal C}_2$ at $E$ and $F$ in such a way that the perpendicular bisectors of $DE$ and $CF$ intersect at a point $M$ on $AB$. Find, with proof, the ratio $AM/MC$.",
"Solution_1": "Let $O$ be the common center of the concentric circles $c_1, c_2$. The tangency point B is the midpoint of the chord AC, because AC is perpendicular to the radius OB of the circle $c_2$, and O is also the center of the circle $c_1$. The power of the point A to the circle $c_2$ is $AE \\cdot AF = AB^2$. But since B is the midpoint of AC and D the midpont of AB, $AD \\cdot AC = \\frac{AB}{2} \\cdot 2 AB = AB^2$ as well. Hence, the quadrilateral CDEF is cyclic. The intersection M of the perpendicular bisectors of its diagonals CE, DF is its circumcenter. If this circumcenter is to be on one of its side CD, it must be the midpoint of this side, $DM = MC = \\frac{DC}{2}$. Since $DC = \\frac 3 2 AB$, $DM = MC = \\frac 3 4 AB$ and $AM = AD + DM = \\frac{AB}{2} + \\frac 3 4 AB = \\frac 5 4 AB$ and $\\frac{AM}{MC} = \\frac 5 3$.",
"Solution_2": "[quote]But since M is the midpoint of AC[/quote] But since B is the midpoint of AC. :)",
"Solution_3": "[b]Problem:[/b] Two circles are concentric. A chord $AC$ of the outer circle touches the inner circle at $Q$. $P$ is the midpoint of $AQ$. A line through $A$ intersects the inner circle at $R$ and $S$. The perpendicular bisectors of $PR$ and $CS$ meet at $T$ on the line $AC$. What is the ratio $AT/TC$?\n\nBy Power of a Point we have $(AR)(AS)=AQ^2 = \\frac{1}{2}2(AQ)^2 = (AP)(AC)$, so $\\triangle ARP\\sim\\triangle ACS$. Hence we have $\\angle ACS=\\angle ARP\\implies PRSC$ is cyclic. Hence $T$ must be the center of its circumcircle and must also lie on the perpendicular bisector of $CP$. Hence, $T$ is the midpoint of $CP$. This yields $CT=\\frac{3}{8}(CA)\\implies\\frac{AT}{TC}=\\frac{5}{3}$.",
"Solution_4": "You guys are wrong its actually 3/5...",
"Solution_5": "Why, maplestory? Is there something wrong with the other solutions?",
"Solution_6": "yes, they assumed that M is not on AB but rather reflected across line OB (O is the center of the concentric circles) so when they found AM/MC they actually found MC/AM, now they have to take the reciprocal to get 3/5.",
"Solution_7": "I don't follow :maybe:",
"Solution_8": "[hide=\" Solution\"]\n[asy]\n /* Geogebra to Asymptote conversion, documentation at artofproblemsolving.com/Wiki, go to User:Azjps/geogebra */\nimport graph; size(10cm); \nreal labelscalefactor = 0.5; /* changes label-to-point distance */\npen dps = linewidth(0.7) + fontsize(10); defaultpen(dps); /* default pen style */ \npen dotstyle = black; /* point style */ \nreal xmin = -4.300000000000004, xmax = 15.96000000000002, ymin = -5.740000000000006, ymax = 6.300000000000007; /* image dimensions */\n\n /* draw figures */\ndraw(circle((2.120000000000002,0.5600000000000007), 2.308246087400564)); \ndraw(circle((2.120000000000002,0.5600000000000007), 3.548182633405449)); \ndraw((-1.290876635526356,1.537507328474018)--(3.747148166480568,2.197189312302931)); \ndraw((-1.290876635526356,1.537507328474018)--(3.609775926061166,3.780274474346656)); \ndraw((0.1261857316220342,1.723058323220483)--(-0.06571349512947573,2.098199114942177)); \ndraw((3.609775926061166,3.780274474346656)--(3.747148166480568,2.197189312302931)); \n /* dots and labels */\ndot((2.120000000000002,0.5600000000000007),dotstyle); \nlabel(\"$O$\", (2.200000000000004,0.6800000000000008), NE * labelscalefactor); \ndot((-1.290876635526356,1.537507328474018),dotstyle); \nlabel(\"$A$\", (-1.220000000000001,1.660000000000002), NE * labelscalefactor); \ndot((0.1261857316220342,1.723058323220483),dotstyle); \nlabel(\"$E$\", (0.2000000000000008,1.840000000000002), NE * labelscalefactor); \ndot((3.747148166480568,2.197189312302931),dotstyle); \nlabel(\"$F$\", (3.820000000000006,2.320000000000002), NE * labelscalefactor); \ndot((1.159449645267405,2.658890901410337),dotstyle); \nlabel(\"$B$\", (1.240000000000002,2.780000000000003), NE * labelscalefactor); \ndot((-1.290876635526356,1.537507328474018),dotstyle); \ndot((3.609775926061166,3.780274474346656),dotstyle); \nlabel(\"$C$\", (3.680000000000006,3.900000000000004), NE * labelscalefactor); \ndot((-0.06571349512947573,2.098199114942177),dotstyle); \nlabel(\"$D$\", (0.02000000000000059,2.220000000000003), NE * labelscalefactor); \ndot((1.772031215465845,2.939236794644417),dotstyle); \nlabel(\"$M$\", (1.860000000000003,3.060000000000004), NE * labelscalefactor); \nclip((xmin,ymin)--(xmin,ymax)--(xmax,ymax)--(xmax,ymin)--cycle); \n /* end of picture */ [/asy]\nLet $O$ be the common center of the circles. Note that $\\triangle OAB \\cong \\triangle OCB,$ so $B$ is the midpoint of $AC$. Let $AC = d.$ We then have $AB = BC = \\dfrac{d}{2}, AD = \\dfrac{d}{4},$ and hence $AD \\times AC= \\dfrac{d^2}{4} = AB^2$. By PoP, we have that $AB^2= AE \\times AF,$ so $AD \\times AC = AE \\times AF,$ so by converse of PoP, points $D, E, F, C$ are concyclic. Hence, the point of intersection of the perpendicular bisectors of $DE$ and $CF$, $M$, must be equidistant from $D, E, F$ and $C,$ implying $M$ must be the midpoint of $CD.$ Now, we have that $CD = \\dfrac{d}{2} + \\dfrac{d}{4}= \\dfrac{3}{4}d,$ so $CM = \\dfrac{CD}{2} = \\dfrac{3}{8}d$ and $AM = d - \\dfrac{3}{8}d = \\dfrac{5}{8}d.$ Hence, our desired answer is $\\dfrac{AM}{MC} = \\boxed{\\dfrac{5}{3}}.$ $\\blacksquare$ [/hide]",
"Solution_9": "can any one please give a bash type solution as well ;) I tried both complex and bary bash but the calculations are getting huge(alas!)",
"Solution_10": "Between I agree with [quote=maplestory]You guys are wrong its actually 3/5...[/quote] The answer is coming 3/5\nSee p8 in the attachment",
"Solution_11": "Yufei probably just made a typo. Note that from $\\tfrac{MC}{AC}=\\tfrac38$ we actually get \\[\\dfrac{AM}{MC}=\\dfrac{AC-MC}{MC}=\\dfrac{AC}{MC}-1=\\dfrac83-1=\\dfrac53,\\] not $\\tfrac35$ as he suggests.",
"Solution_12": "Wait it says that M is on AB, so I took it to be M on segment AB. However, according to the solution, M is on segment BC. Did they mess up the problem or did they mean line AB?",
"Solution_13": "[quote=Phie11]Wait it says that M is on AB, so I took it to be M on segment AB. However, according to the solution, M is on segment BC. Did they mess up the problem or did they mean line AB?[/quote]\n\nI believe they are referring to line $AB$, not segment $AB$ because that is how they defined point $C$ as well.",
"Solution_14": "Some wishful thinking tells us that M should be the midpoint of $\\overline{DC}$. Since $M$ is the intersection of the perpendicular bisectors of $\\overline{DE}$ and $\\overline{CF}$, it suffices to show that quadrilateral $DEFC$ is cyclic. However notice that $\\overline{AC}$ is the radical axis of $C_1$ and $C_2$, hence by PoP, we have that $\\overline{AE}\\cdot\\overline{AF}=\\overline{AD}\\cdot\\overline{AC}$, implying that $DEFC$ is cyclic. Through direct computation, we see that $\\frac{\\overline{AM}}{\\overline{MC}}=\\frac{5}{3}$. $\\square$",
"Solution_15": "By power of a point, $$AE \\cdot AF = AB^2 = AD \\cdot AC,$$ so $CDEF$ is cyclic. But by definition, $M$ must be the cyclic center of $(CDEF)$, and thus $MC = MD$. Then $$\\frac{MA}{MC} = \\frac{\\frac 52 AD}{\\frac 32 AD} = \\boxed{\\frac 53}.$$",
"Solution_16": "Note that $$AD\\cdot AC=(1/2 AB)(2AB)=AB^2=AE\\cdot EF,$$ so $CDEF$ is cyclic. Then, $M$ is the circumcenter of $CDEF$, and since $M$ lies on $DC$, $M$ is the midpoint of $DC$. Hence, $M$ is 5/8 of the way from A to C, so the answer is $5/3.$",
"Solution_17": "We have AF*AE=1/2AB*2AB=AD*AC, so CDEF is cyclic. Then M is the circumcenter and because it lies on CD consequently the midpoint. Calculating simply (find BM=1/8AC) we get an answer of 5/3.",
"Solution_18": "By Power of a Point, $AE \\cdot AF = AB^2$ and by the midpoint condition we see that $AD \\cdot AC = \\tfrac12 AB \\cdot 2AB = AB^2$, so $AE \\cdot AF = AD \\cdot AC$. By Converse of Power of a Point we find that $CDEF$ is cyclic. \n\n[b]Claim:[/b] $M$ is the midpoint of $CD$.\n[i]Proof.[/i] Since $M$ is the intersection of the perpendicular bisectors of $DE$ and $CF$, we see that it is the circumcenter of $(CDEF)$. From here, it immediately follows that $M$ is the midpoint of $CD$, as desired. $\\square$\n\nFinally, note that $MC=\\tfrac38 AC$, so $AM=\\tfrac58 AC$, which gives $\\tfrac{AM}{MC}=\\tfrac53$. $\\blacksquare$",
"Solution_19": "Headsolving without coherency [verbatim]:D[/verbatim]\n\nWrite $AC\\cdot AD=AB^2=AE\\cdot AF$ so that $CDEF$ is cyclic. It follows that $M$ is the center with $MC=MD$, whence it's clear that $\\tfrac{AM}{MC}=\\tfrac{5}{3}$. $\\square$",
"Solution_20": "[asy]\n size(250,250); \n draw(circle((0,0),5)); draw((-4,3)--(4,3)); draw(circle((0,0),3)); \n\n label(\"$A$\", (-4,3), NW); filldraw(circle((-4,3),0.05),black);\n label(\"$B$\", (0,3), N); filldraw(circle((0,3),0.05),black);\n label(\"$C$\", (4,3), NE); filldraw(circle((4,3),0.05),black);\n label(\"$D$\", (-2,3), NW); filldraw(circle((-2,3),0.05),black);\n label(\"$M$\", (1,3), N); filldraw(circle((1,3),0.05),black);\n\n label(\"$E$\", (-1.918677,2.306225775), W); filldraw(circle((-1.918677,2.306225775),0.05),black);\n label(\"$F$\", (2.918677,0.6937742252), SE); filldraw(circle((2.918677,0.6937742252),0.05),black);\n\n draw(circle((1,3),3)); draw((-4,3)--(2.918677,0.6937742252));\n[/asy]\n\nI claim that the ratio $\\frac{AM}{MC}$ is always $\\frac{5}{3}$. \n\n[color=#00f][b]1. $B$ is the midpoint of $AC$.[/b][/color]\nNote that since $AC$ is tangent to $\\mathcal C_2$, we have that $\\angle OBA=90$. However, since $AC$ is also a chord of $\\mathcal C_1$, which has center $O$, this also means that $B$ is the midpoint of $AC$. \n\n[color=#00f][b]2. (Power of a Point Converse) $CDEF$ is cyclic.[/b][/color]\nNote that the power of $A$ with respect to $\\mathcal C_2$ is \n\\[AB^2=BC^2=AD(AC),\\]\nsince by \\textbf{(1)}, $B$ is the midpoint of $AC$, and by problem conditions, $D$ is the midpoint of $AB$. However, note by Power of a Point on $A$ with respect to $\\mathcal C_2$, we have \n\\[AB^2=AE(AF) \\iff AE(AF)=AD(AC).\\]\nFinally, using the converse of Power of a Point, we have that $CDEF$ must be cyclic. \n\n[color=#00f][b]2. (Similar Triangles) $CDEF$ is cyclic.[/b][/color]\nSimilarly, as in the first proof of \\textbf{(2)}, we find that \n\\[AE(AF)=AC(AD),\\]\nwhich actually gives us that\n\\[\\frac{AD}{AF}=\\frac{AE}{AC},\\]\nmeaning that by SAS similarity, $\\triangle ADE \\sim \\triangle AFC$. Through angle chasing, we then have that\n\\[\\angle AFC + \\angle EDC = \\angle AFC + (180-\\angle ADE) = \\angle AFC + (180-\\angle AFC) = 180,\\]\nmeaning that $CDEF$ is cyclic. \n\n[color=#00f][b]3. (Finishing) $M$ is the midpoint of $CD$.[/b][/color]\nNote that since $CDEF$ is cyclic, and the perpendicular bisectors of $DE$ and $CF$ intersect at $M$, $M$ must be the circumcenter of $(CDEF)$. However, since $M$ is on $CD$, this means that $M$ must be the midpoint of $CD$. Therefore, since $D$ is the midpoint of $AB$, and $B$ is the midpoint of $AC$, we find that $\\frac{AM}{MC}=\\frac{5}{3}$, finishing the problem.",
"Solution_21": "Let the common centre be $O$ and the radius of $C_1$ be $R_1$ and of $C_2$ be $R_2$.\nNow from PoP we have: $AB^2 = AO^2-R_2^2 = R_1^2-R_2^2 = CO^2-R_2^2 = CB^2$.\nSo: $AB=BC$.\nAgain from a PoP: $AB.AC=2AB.\\frac{AC}{2}=AB^2=AE.AF$. $ \\Leftrightarrow DEFC$ is cyclic.\nNow we have that $M$ is the midpoint of $DC=3AD$, so $AM=\\frac{5}{2}AD$ and because of $AC=4AD$, $MC=\\frac{3}{2}AD$.\nSo: $\\frac{AM}{MC}=\\frac{5}{3}$",
"Solution_22": "We first claim that $CFED$ is cyclic. This is because\n\n$AB^2 = AE \\cdot AF \\Rightarrow AD \\cdot AC = AE \\cdot AF \\Rightarrow \\triangle ACF \\sim \\triangle AED$\nThis proves our claim. From the given perpendicularity constraints we can see that $M$ is the circumcenter of $CFED$. This gives our answer of $\\boxed{\\frac{5}{3}}$ . $\\blacksquare$",
"Solution_23": "Notice that\n\n\\[AE \\cdot AF = AB^2=AB \\cdot BC = AD \\cdot AC\\]\n\nimplying that $CDEF$ is cyclic. Hence, it is clear that $M$ must be the midpoint of $CD$ at which point we easily find the ratio to be $\\boxed{\\frac{5}{3}}$.\n\n[Refer to #48 for a diagram, I was too lazy to make one]",
"Solution_24": "From Power of a point, $AD \\cdot AC = AB^2 = AE \\cdot AF$, so quadrilateral $CDEF$ is cyclic. Point $M$ must be the circumcenter of $CDEF$, so $M$ is the midpoint of $CD$, and the ratio $AM/MC = \\frac{5}{3}$.",
"Solution_25": "[b][color=#00f]Claim:[/color][/b] $DEFC$ is cyclic\n\n[i]Proof:[/i]\nBy PoP $$AE \\cdot AF=AB^2=2AB \\cdot \\tfrac{AB}{2}=AD \\cdot AC$$\nhence $DEFC$ is cyclic. $\\square$\nSince $M$ is the intersection of the perpendicular bisectors of $DE$, and $CF$, $M$ is the center of $(DEFC)$. \nSo:\n$$MC=\\frac{AC-AD}{2}=\\frac{AC-\\tfrac{1}{2}AB}{2}=\\frac{AC-\\tfrac{3}{4}AC}{2}=\\frac{3}{8}AC.$$\n$$\\frac{AM}{MC}=\\frac{AC-MC}{MC}=\\frac{AC-\\tfrac{3}{8}AC}{\\tfrac{3}{8}AC}=\\frac{\\tfrac{5}{8}AC}{\\tfrac{3}{8}AC}=\\frac{5}{3}.$$\n",
"Solution_26": "Note $B$ is the midpoint of $\\overline{AC}$ so \\[AF\\cdot AE=AB^2=(\\tfrac{1}{2}AB)(2AB)=AD\\cdot AC\\] and $CDFE$ is cyclic. Hence, $M$ is the center of this circle and thus the midpoint of $\\overline{CD}$. WLOG if $AD=2$, then $DB=2$ so $BC=4$ so $CM=3$ so $AM/CM=5/3$. $\\square$",
"Solution_27": "By power of a point $AE*AF=AB^2=AD*AC$, so $DEFC$ is cyclic.\nBecause $M$ is the intersection of the perpendicular bisectors of $DE$ and $FC$, $M$ is the center of the circumcircle, so $MD=MC$\nLet $AD=a, BM=x$\n$MC=MD=a+x$, so $MC+MB=a+2x=2a$, so $x=\\frac{a}{2}$. Then $\\frac{AM}{MC}=\\frac{2.5a}{1.5a}=\\boxed{\\frac{5}{3}}$",
"Solution_28": "By Power of a Point with respect to $\\mathcal{C}_2,$ we get $$AB^2 = AE \\cdot AF.$$\n\nThen, since $D$ is the midpoint of $\\overline{AB},$ it follows that $$AD \\cdot AC = AB^2 \\implies AD \\cdot AC = AE \\cdot AF.$$\n\nHence, by the converse of Power of Point, quadrilateral $CDEF$ is cyclic.\n\nSince the perpendicular bisectors of $\\overline{DE}$ and $\\overline{CF}$ intersect at $M,$ this tells us that the circumcenter of $CDEF$ is $M,$ so $M$ is on the midpoint of $\\overline{CD}.$\n\nTherefore, if we let $AD=x,$ we then get\n\\begin{align*}\nCD=3x &\\implies DM=MC=1.5x \\\\\n&\\implies AM=2.5x \\\\\n&\\implies \\frac{AM}{MC} = \\frac{2.5x}{1.5x} = \\boxed{\\frac{5}{3}}.\n\\end{align*}\n$\\square$",
"Solution_29": "[asy]\ndraw(circle((2,1),1));\ndraw(circle((2,1), sqrt(5)));\npair O=(2,1), A=(0,0), B=(2,0), C=(4,0), D=(1,0), M=(5/2,0);\ndraw(A--C);\ndot(D);\ndot(B);\ndot(M);\ndot(O);\ndot(A);\ndot(C);\nlabel(\"$A$\", A, SW);\nlabel(\"$B$\", B, S);\nlabel(\"$C$\", C, SE);\nlabel(\"$D$\", D, S);\nlabel(\"$M$\", M, S);\nlabel(\"$O$\", O, S);\npair E=(2-2sqrt(5)/5, 1-sqrt(5)/5), F=(2+2sqrt(5)/5, 1+sqrt(5)/5);\ndot(E);\ndot(F);\nlabel(\"$E$\", E, NW);\nlabel(\"$F$\", F, NE);\ndraw(A--F);\npair G=D/2+E/2, H=F/2+C/2;\ndraw(G--M--H);\ndraw(D--E);\ndraw(F--C);\n\n[/asy]\nWe claim the answer is $\\boxed{\\dfrac{5}{3}}.$\n\nLet $O$ be the center of both $C_2$ and $C_1.$ Since $AB$ is tangent to $C_2,$ $\\angle OBA=90^{\\circ},$ so $AC$ is bisected by $OB$ and $AB=BC.$ By power of a point, $AE\\cdot AF=AB^2=\\dfrac{AB}{2}\\cdot 2AB=AD\\cdot AC,$ so $DEFC$ is cyclic. Note that the perpendicular bisectors of chords $DE$ and $FC$ in the circumcircle of $DEC$ intersect at the circumcenter of $DEC,$ so $M$ is the circumcenter of $DEC$ and $DM=CM.$ As such, $\\dfrac{AM}{MC}=\\dfrac{AD+DM}{MC}=\\dfrac{\\left(AD+\\dfrac{DC}{2}\\right)}{\\left(\\dfrac{DC}{2}\\right)}=\\dfrac{\\left(AD+\\dfrac{3AD}{2}\\right)}{\\left(\\dfrac{3AD}{2}\\right)}=\\dfrac{5}{3},$ as desired. $\\Box$"
}
{
"Tag": [
"calculus",
"calculus computations"
],
"Problem": "Let $ a_{1},a_{2}$\r\n$ a_{n\\plus{}1}\\equal{}\\frac{1}{2n}a_{n\\minus{}1}\\plus{}\\frac{2n\\minus{}1}{2n}$\r\nFind $ lim a_{n}$",
"Solution_1": "let $ b_{n}\\equal{}a_{n}\\minus{}1$, so $ b_{n\\plus{}1}\\equal{}\\frac{b_{n\\minus{}1}}{2n}$ so $ b_{n}\\rightarrow 0$ so $ a_{n}\\rightarrow 1$."
}
{
"Tag": [],
"Problem": "what is the different betwin molal and molar?",
"Solution_1": "did u mean what is the diff between molal and molar?\r\n\r\nwell molal is mols of solute over kg of solvent, while molar is mols of solute over liters of solution...",
"Solution_2": "[quote=\"g1sk8terz\"]did u mean what is the diff between molal and molar?\n\nwell molal is mols of solute over kg of solvent, while molar is mols of solute over liters of solution...[/quote]\r\n\r\nthanks,,,\r\n\r\nthe mistake edited. ;)"
}
{
"Tag": [
"algebra",
"polynomial",
"algebra solved"
],
"Problem": "P(x) and Q(x) are polynomials of positive degree such that for all x P(P(x))=Q(Q(x)) and P(P(P(x)))=Q(Q(Q(x))). Does this necessarily mean that P(x)=Q(x)?",
"Solution_1": "We have from the given that:\r\nP(Q(Q(x)))=Q(Q(Q(x))) , so polYnomials P and Q take for a sufficiet big nr of values(as many as Q(Q(x))) take the same value, so they must be identical!\r\nI hope I've explained it well enough :) ,at least the idee should be clear!",
"Solution_2": "I think a problem from a recent Romanian TST stated that $P=Q$ if $P\\circ P=Q\\circ Q$, but I'm not sure right now. I'd have to check.",
"Solution_3": "What about $P(x)=x$ and $Q(x)=-x$?",
"Solution_4": "Yes, sorry about that, $P$ and $Q$ were monic complex polynomials in that problem. It's from a TST in $2000$.",
"Solution_5": "Myth, your polynomials do not satisfy the second condition.",
"Solution_6": "[quote=\"grobber\"]$P$ and $Q$ were monic complex polynomials in that problem.[/quote]\r\nIt is obvious statement for real polynomials, but complex... I need additional reasoning.",
"Solution_7": "[quote=\"Xixas\"]Myth, your polynomials do not satisfy the second condition.[/quote]\r\nIt was counterexample for grobber's post.",
"Solution_8": "[quote=\"Myth\"][quote=\"grobber\"]$P$ and $Q$ were monic complex polynomials in that problem.[/quote]\nIt is obvious statement for real polynomials, but complex... I need additional reasoning.[/quote]\r\n\r\nYes, for real polynomials you don't need any messy coefficient identification: you can do it by some analysis, studying what happens when $x\\to\\infty$. I think that's what I did when I first saw the problem, because I didn't read the text properly and missed \"complex polynomials\". On the other hand, I think we have to resort to identifying the coefficients if $P,Q$ are complex, and I don't like such problems, so I won't try to solve it :).",
"Solution_9": "Actually, at this time I am writing solution, which uses behavour when $x\\to\\infty$.",
"Solution_10": "[quote=\"Xixas\"]P(x) and Q(x) are polynomials of positive degree such that for all x P(P(x))=Q(Q(x)) and P(P(P(x)))=Q(Q(Q(x))). Does this necessarily mean that P(x)=Q(x)?[/quote]\r\nFrom $P(P(P(x)))=Q(Q(Q(x)))$ we immediately obtain $P$ and $Q$ have the same leading coefficient.\r\nSuppose $P(x)\\neq Q(x)$, then $P(x)=a_nx^n+...+a_kx^k+p_{k-1}x^{k-1}+...+p_0$ and $P(x)=a_nx^n+...+a_kx^k+q_{k-1}x^{k-1}+...+q_0$, where $p_{k-1}>q_{k-1}$ (for example).\r\n0) $n=1$. Any questions?\r\n1) $a>0$. Then there is $M$ s.t. $P(x)$ and $Q(x)$ increase on $[M,+\\infty]$, $P(M)>M$, $Q(M)>M$, $P(x)>Q(x)$ on $[M,+\\infty]$. We see that $P(P(x))>Q(P(x))>Q(Q(x))$ on $[M,+\\infty]$.\r\n2) $a<0$. Almost the same argumentation, but a bit more boring. :?",
"Solution_11": "[quote=\"Xixas\"]P(x) and Q(x) are polynomials of positive degree such that for all x P(P(x))=Q(Q(x)) and P(P(P(x)))=Q(Q(Q(x))). Does this necessarily mean that P(x)=Q(x)?[/quote]\r\n\r\nLet $S$ be the image of $P(P(x))$. Since $\\deg P > 0$, $S$ has infinitely many elements.\r\n\r\nSince $P(P(P(x)))=Q(Q(Q(x)))$, we see that if $y \\in S$, then $P(y) = Q(y)$. Two polynomials agree on infinitely many values imply that they're identical. Hence $P = Q$.",
"Solution_12": "It is strange. What if we remove condition $P(P(x))=Q(Q(x))$.",
"Solution_13": "[quote=\"grobber\"]Yes, sorry about that, $P$ and $Q$ were monic complex polynomials in that problem. It's from a TST in $2000$.[/quote]Yes I remember that problem. Nice one, solved it quickly, with equalizing coefficients ;)"
}
{
"Tag": [
"inequalities",
"induction"
],
"Problem": "How do we solve this with no use of calculators or logs \r\n4^n < 2^100 + 3^100 < 4^(n+1)",
"Solution_1": "The trick is clearly the 3^100, right? So,[hide] you're going to want to try and estimate something like (3/4)^n and see if you can use that. Because what's going to happen is that the 2^100/4^n is probably going to be very, very, very small, so you may not even have to think about it.[/hide]",
"Solution_2": "Maybe you can use the fact that [hide]44 is just a little larger than 35[/hide].",
"Solution_3": "[hide]An idea perhaps :\n\n\n\n2^100 + 3^300 = 4^50 * (1 + 4^25 * (1 + 1/8)^50))\n\n\n\nSo we can can now say that n >= 75. \n\n\n\n(1+x)^50 > f(x)=1 + 50x + 1225x (binomial)\n\n\n\nf(1/8) is still tractable without a computer (But of course I used one since I'm at home and noone is looking at me ). f(1/8) = 1689/64 ~26 > 4^2.\n\n\n\nSo n >= 77.\n\n\n\nSince C(50,n) (1/8)^n quickly becomes tiny, 77 is not to far from the actual value.[/hide]",
"Solution_4": "Not sure but here goes nothing (warning, this is just a guess!!):\n\n[hide]\n\nx^n+y^n:approx:((x+y)^n)/(2^(n-1)) note: if x and y are close. So we have 1<(5^100)/((2^(99))*2^(2n))<2^2.\n\n\n\n5^100<2^(2n+101), (5/2)^100<2^(2n+1), (5/2)^50<2^n:rt2: \n\n\n\n5/2 :approx: 2 :rt2: , so we have 2^75<2^(n+1/2). n+1/2>75, n>74\n\n[/hide]\n\nBut this is just a guess.",
"Solution_5": "[hide] We must have 4^n-2^100<3^100 and n is as big as possible. factor 2^100 out, we have (4^(n-50))-1<(3/2)^100. Note that (3/2)^4 is very close to 5. sub that in, (4^(n-50))-1<5^25+S, S=(3/2)^100-5^25. Obviously, n>75. If n=80, then 4^30>5^25 because 4^6>5^5. 80 is very close to what the answer will be because S is not going to be a small number. I couldn't get the exact answer although looks like n=79 or 78. [/hide] \n\n I edited my solution, I just figured out how to get the exact answer. [hide]4^6=5^5+971. (3125+997)^5-1>3125+S, because 3125+S=(5+1/16)^25, or (3125+625(5)/16+125(10)/256+25(10)/4096+..)^5 <(but very close) to (3125+201)^5. Let's say n=79, then ((3125+997)^5)/4 should be<(3125+201)^5, or (4096/3326)^5<4.(4096/326)^5<(1+1/4)^5 and (1+1/4)^5<4, which is true. so 79 works[/hide]",
"Solution_6": "Is it possible to get an exact answer without using a calculator? No one seemed have gotten an exact answer without using calculator or logs.I think the answer is [hide]79 [/hide] by using logs and calculator.",
"Solution_7": "I agree with you on that one.",
"Solution_8": "I gave you folks a hint, but you didn't listen. So I'll finish it off.\n\n\n\nI will use the inequality (1 + x)n :ge: 1 + nx for x :ge: -1 and n a nonnegative integer. The proof of that inequality is by induction on n.\n\n\n\nThe rest is in spoiler: [hide]\n\nWe have 480 / 3100 = (256/243)20 = (1 + 13/243)20 :ge: 1 + (13)(20)/243 = 1 + 260/243 > 2. Thus 480 > 2 3100 > 2100 + 3100. That's one direction.\n\n\n\nWe have 325 / 420 = (243/256)5 = (1 - 13/256)5 :ge: 1 - (13)(5)/256 = 1 - 65/256 = 191/256. Thus 350 / 440 :ge: (191/256)2 = 36841/65536 > 1/2. Thus 3100 / 480 > (1/2)2 = 1/4. Hence 479 < 3100 < 2100 + 3100. That's the second direction.\n\n\n\nThus, assuming that n must be an integer, we have n = 79.\n\n[/hide]"
}
{
"Tag": [],
"Problem": "Both $ \\bigtriangledown$ and $ \\oslash$ are operations in the set $ \\{ \\plus{} , \\minus{} , \\times , \\div \\}$, and $ \\frac{12 \\bigtriangledown 2}{9 \\oslash 3} \\equal{} 8$. What is the value of $ \\frac{4 \\bigtriangledown 3}{2 \\oslash 5}$?",
"Solution_1": "We find out that the triangle means multiplication and the circle means division. Therefore, answer:\r\n4*3/(2/5)=4*3*5/2=[b]30[/b]"
}
{
"Tag": [
"infinity"
],
"Problem": "Does it bug anyone else how there are the same number of natural numbers as there are whole numbers (and rational numbers and integers and blah blah). I mean, I know the reasons and everything about one to one mappings but it just doesn't seem [i]right[/i] to me.. All these numbers, add on one more number and you get the same. Argh.",
"Solution_1": "It used to bother me, but I have now accepted that arithmetic for infinities doesn't work quite the same way that it does for finite quantities. What really bothered me is that between any two rationals is an irrational, and between any two irrationals is a rational, but even though any transfinite quantity +/-1 is itself, there are more irrationals than rationals. But now I accept these things, since I do know their proofs, and you can't argue with a proper proof.",
"Solution_2": "Question: Are there just as many real numbers as there are irrational numbers?",
"Solution_3": "I would suppose so. Seeing as the number of real numbers is a \"bigger\" infinity than the number of rational numbers, subtracting a \"smaller\" infinity from the \"bigger\" infinity should leave the \"bigger\" infinity. Just like subtracting one from infinity leaves infinity.",
"Solution_4": "Speaking of how [color=#0099FF]infinity - 1 = infinity[/color], then\r\n\r\nShouldn't [color=#0099FF]1/infinity[/color] be 0?\r\n\r\nShouldn't [color=#0099FF]infinity/infinity[/color] be 1?\r\n\r\nAnd what do you do with [color=#0099FF]infinity/0[/color]?\r\n\r\nSomehow infinite numbers seem a little more real than they used to. :mrgreen:",
"Solution_5": "Shouldn't infinity/infinity be 1?\r\n\r\nNo, consider 2x/x. For any x except zero, that will equal two, since it's a horizontal line. But, if you plug in x = infinity, you get inf./inf. = 2. Inf./inf. can equal anything, depending on how you got the infinities.\r\n\r\ninfinity/0\r\nEither +inf. or -inf., depending on how you got it.",
"Solution_6": "1/(+ or - infinity) does equal 0.",
"Solution_7": "There's some junk with limits that will help with your questions regarding (infinity/infinity) or 0/0. As for the \"equal number\", the accepted method that helped me \"comprehend\" the situation is by establishing a bijection. For example, with whole numbers and natural numbers, if you list them \"all\" out, you have:\r\n\r\nWholes: 0, 1, 2, 3, 4, ...\r\nNaturals: 1, 2, 3, 4, 5, ...\r\n\r\nNotice how 0 and 1 pair up, 1 and 2, 2, and 3, and so on. This is a bijection (one to one correspondence), and this can continue forever, so there are an equal number of wholes and naturals. The same \"logic\" can be applied to natural numbers and multiples of some integer as well.\r\n\r\n1, 2, 3, 4, 5, ...\r\n2, 4, 6, 8, 10, ...",
"Solution_8": "How do you figure out what infinity/0 is? I thought you couldn't divide by zero.",
"Solution_9": "[quote=\"Bookworm\"]How do you figure out what infinity/0 is? I thought you couldn't divide by zero.[/quote]\r\n\r\nLimits...take the case y = 1/x. As you get closer to 0 for x, y approaches +/- infinity. So when you divide by 0, you get +/- infinity in any case. I think. Any ideas?",
"Solution_10": "Basically (read: very informally) infinity/0 = infinity * (1/0) = infinity * infinity = infinity. The +/- thing is a bit weird, but we can reasonably say that infinity=-infinity. In complex analysis, we usually let all infinities be the same point (for example on a Riemann sphere, it's the north pole).",
"Solution_11": "Yeah. Well basically if we have a fraction x/y and increase x the fraction increases, and decrease y it increase. Either way as x->oo or y->0 it will go to infinity.\r\n\r\nWhen we get to things like 0/0 and infinity/infinity, you can use L'Hopitals rule, the limit equals the limit of the thing where you differentiate top and bottom separately. Eg lim x->0 of (sin x)/x = lim x->0 (cos x)/1 = 1.\r\n(Just learnt that this year so felt like telling you all :))",
"Solution_12": "L'hopital's rule also works for 0/0, and other forms which are equivalent to one of those two.\r\n\r\nBTW, here's an example of a really weird problem. Let's say you choose one integer out of the set of all integers. What is the probability that it's odd? Common sense tells you that it is 1/2. But remember, probability can be defined as /, or the number of desired possibilities over the total number of possibilities. In this case, that would be <# of odd integers>/, and since those two sets are the same size, the answer should be 1! (read as 1 in a very surprised way, not as 1 factorial)",
"Solution_13": "Thanks for answering my question. \r\nSorry if I ask a lot of questions, but what do you mean by \"the limit equals the limit of the thing where you differentiate top and bottom separately\"?",
"Solution_14": "[quote=\"MysticTerminator\"]BTW, here's an example of a really weird problem. Let's say you choose one integer out of the set of all integers. What is the probability that it's odd? Common sense tells you that it is 1/2. But remember, probability can be defined as /, or the number of desired possibilities over the total number of possibilities. In this case, that would be <# of odd integers>/, and since those two sets are the same size, the answer should be 1! (read as 1 in a very surprised way, not as 1 factorial)[/quote]\r\n\r\nI don't think that's right. Just because there's a bijection between the two doesn't mean that the quotient is the same. It's still lim_{x\\rightarrow \\infty} {x/(2x)}=1/2.",
"Solution_15": "[quote=\"Bookworm\"]Thanks for answering my question. \nSorry if I ask a lot of questions, but what do you mean by \"the limit equals the limit of the thing where you differentiate top and bottom separately\"?[/quote]\r\n\r\nSorry I was trying to say it as informally as possible, probably got too informal. Basically lim (x->a) f(x)/g(x) = lim (x->a) f'(x)/g'(x).\r\n\r\nAs for the odd number probability thing, I think it could be equal to 1 I'm not sure. I know there are examples where even though its obviously not 1, it does equal 1. For example, whats the probability a random positive integer has the digit 3 in it? That turns out to be 1, I think.",
"Solution_16": "Whoa... don't quite follow you there about every number having a three in it. Does that use the same sort of logic that I used in my \"ever number is an odd number\" thing?\r\n\r\nBTW, about L'Hopital's rule, does anyone have a good way to visualize it, or \"intuitively\" understand it. I mean, it doesn't make any sense that if the two functions are both 0 or infinity at the point where you are taking the limit, their ratio will be equal to their slopes at that point. Should I just accept it and move on, or does anyone have an intuitive explanation of it?",
"Solution_17": "Oh yeah, \r\nAs an example of L'Hopital's rule, in case it isn't clear, let's say we take the limit as x goes to 3 of (x^2-9)/(x^2-2x-3). If you just plug in x=3, that will be 0/0. Now, we can do that the nice little way of canceling out a factor of (x-3) on both the numerator and denominator, leaving (x+3)/(x+1). Pluggin in x=3 now gives 1.5.\r\n-OR-\r\nWe could differentiate the top and bottom to give 2x/(2x-2)=x/(x-1), and the plug in x=3 to get the same answer. Of course, you don't usually use L'hopital's rule on things you can do other ways, I just wanted to show it worked.",
"Solution_18": "[quote=\"TripleM\"]For example, whats the probability a random positive integer has the digit 3 in it? That turns out to be 1, I think.[/quote]\r\n\r\nYes, that's definitely 1, since if you pick a random positive integer, chances are it's going to be really large and have many digits. Basically, the probability that an n-digit number has a 3 in it is 1-(9/10)^n, and the average number of digits in a random positive integer is infinitity, so when n goes to infinity we get 1. That's really handwavy, but that's the general idea behind problems like that.",
"Solution_19": "really? so it's not [u]exactly[/u] one? b/c if u let handwavy things start to get by you, you could say 0 is about 1, and then you could get all sorts of results, like your school is actually a giant carrot.",
"Solution_20": "In this case, ``handwavy'' means ``correct, but probably not good enough to go in a proof for, say, the USAMO.''",
"Solution_21": "It's .999bar. Call that n.\r\n10n = 9.9999bar\r\n- n = .9999bar\r\n 9n = 9\r\n n = 1\r\n\r\nSo it's 1, not just \"about 1.\"",
"Solution_22": "Yes, I agree. But in the case of the problem with the 3s, it is slightly more difficult since you have to do work with the expected value of the number of digits. This is infinity with uniform distribution, but I think you have to use umbral calculus to prove the result rigorously. I don't think it's really that easy, but it's not too hard to give an intuitive ``proof'' of the desired result.",
"Solution_23": "Yeah I think it comes down to how you can pick a random integer. Which is basically impossible..",
"Solution_24": "It's impossible with uniform distribution. It's not impossible with, say, the Poisson Distribution.",
"Solution_25": "There are two general themes here. One, is not to confuse ordinal numbers with cardinal numbers. For example, the number 4 is different from a *set* with four members. In one instance 4 is in itself an algebraic object, and in the other it is just describing the cardinality of the set. This was one of the pitfalls of early set theory when Whitehead tried to define natural numbers by the equivalence class of all sets having that cardinality. There are some serious problems with doing that. In the 70s however, John Conway came up with a system of incorporating transfinite numbers into the system of natural numbers with what he called surreal numbers. There's a very good exposition of this in his book [u]On Numbers and Games[/u]. I highly recommend it.\r\n\r\nThe other theme I see here is to not confuse a sequence of numbers with it's limit. This is best illustrated by Zeno's Paradox, but in the opposite sort of way than some of you are doing it. For example with the 3s digit problem, you are applying properties of the finite sample spaces (i.e. numbers having n digits), to the limit space (numbers having a countably infinite number of digits). But there's no justification for being able to do that. There are many examples in mathematics where something is true for each finite case in a sequence, but not true for the limit case. The best you can say about that 3s digit problem is that the probability can be made as arbitrarily close to 1 as you want by increasing the number of digits. But once you talk about an infinite number of digits you have to redefine the probability distribution on your sample space.[/u]",
"Solution_26": "Precisely. See, so even though if you let n=infinity and the probability is one, that won't work for like a two digit number (which is possible) (even though choosing a two digit number technically has probability 90/infinity=0). I suppose some guy who (again) had two much spare time showed that there were an equal number of the numbers that had threes in them as there were numbers, or something like that.",
"Solution_27": "Another cute example of the pitfals of infinity is something I found on the math.sci newsgroup. Suppose you have the following infinite sequence 3 31 314 3141 31415 314159... where the nth term in the suquence is the truncation of the first n digits of pi. Then is pi contained in this sequence?",
"Solution_28": "Wow...[color=#3399CC]infinity[/color] is a hard idea to comprehend.\r\n\r\nIt is an idea, right? :mrgreen:",
"Solution_29": "[quote=\"Ex274513\"]Wow...[color=#3399CC]infinity[/color] is a hard idea to comprehend.\n\nIt is an idea, right? :mrgreen:[/quote]\r\n\r\nThe reason why people have so much trouble with the idea of infinity is because we don't deal with it on an everyday basis. In our everyday lives we only deal with the finite; so our intuitions are bad when it comes to infinity. This is why Cantor, who first developed much of our modern understanding of infinity, was derided by even well established mathematicians of his day.\r\n\r\nBut this is where the strength of mathematical thinking comes through. It forces you to put aside your everyday intuitions, which can be misleading, and to use the mathematical tools which you know to look at things in a different light. Then once you have done that enoough you build a new intuition, much the same was as we have done with complex numbers.",
"Solution_30": "In response to gauss's example, I say... no. The nth term is just a very good approximation to Pi* 10^(n-1). The [u]limit[/u] of the sequence will have Pi in it (actually it won't b/c it goes to infinity, but if you created a new sequence 3, 3.1, 3.14, 3.141, 3.1415, 3.14159, etc., the limit of that would have Pi in it.) The terms themselves will just be approximations. \r\n\r\nSort of thinking in calculus terms here: the [u]limit[/u] of (x^2-9)/(x-3) at 3 will be 6, meaning that as you get closer to 3, you can make the value get as close to 6 as you want, but the value itself will never be 6."
}
{
"Tag": [
"conics",
"parabola",
"hyperbola"
],
"Problem": "Let $a$ and $c$ be fixed positive numbers. For each real number $t$ let $(x_t, y_t)$ be the vertex of the parabola $y = ax^2+bx+c$. If the set of vertices $(x_t, y_t)$ for all real values of $t$ is graphed in the plane, the graph is\r\n\r\nA. a straight line\r\nB. a parabola\r\nC. part, but not all, of a parabola\r\nD. one branch of a hyperbola\r\nE. None of these",
"Solution_1": "[hide]we know that the vertex of this parabola is at the point $(x_t,y_t)=\\left(\\frac{-b}{2a},y_t\\right)$.\nwe then plug this into the equation and find that the set of verticies is the equation \n$\\displaystyle y_t-c=\\left(\\frac{b^2}{4a}-\\frac{b^2}{2a}\\right)= \\frac{-b^2}{4a}$\nsince the values of $a$ and $c$ are fixed we can disregard them. this basically leaves the equation $y_t=\\left(\\frac{-1}{4}\\right)(b^2)$ whcih is a parabola.\nsince the equation of the verticies is $\\displaystyle\\left(\\frac{-b}{2},\\frac{-b^2}{4}\\right)$\n\ndid i do this correctly of have i made a mistake somewhere? :oops:[/hide]",
"Solution_2": "You did it right, I also got a [hide=\"answer\"]parabola[/hide]."
}
{
"Tag": [
"algebra",
"polynomial",
"Vieta",
"inequalities",
"algorithm",
"number theory proposed",
"number theory"
],
"Problem": "find all positive integers such that $ \\frac{a^2\\plus{}b^2\\plus{}a\\plus{}b\\plus{}1}{ab}$ is an integer",
"Solution_1": "[hide]there is trivial solution (1,1) \nlet $ k \\equal{} \\frac {a^2 \\plus{} b^2 \\plus{} a \\plus{} b \\plus{} 1}{ab}$ fix k\n\\[ a^2 \\minus{} a(kb \\minus{} 1) \\plus{} b^2 \\plus{} b \\plus{} 1 \\equal{} 0\\]\nif we have a $ (x,y)$ couple satisfying that condition we can suppose that $ x > y$ (because x cannot be equal to y unless they are not equal to 1).\nso if $ (x,y)$ is a solution $ (y,ky \\minus{} 1 \\minus{} x)$ satisfies too. We have $ ky \\minus{} 1 \\minus{} x > 0$ because $ (ky \\minus{} 1 \\minus{} x).x \\equal{} y^2 \\plus{} y \\plus{} 1$\nso we have another couple $ (y,ky \\minus{} x \\minus{} 1)$ we have $ x > y$ there our aim is to prove $ y > ky \\minus{} x \\minus{} 1$\n\\[ ky \\minus{} x \\minus{} 1 < y < \\equal{} > kxy \\minus{} x < yx \\plus{} x^2\\]\n$ kxy \\equal{} x^2 \\plus{} y^2 \\plus{} x \\plus{} y \\plus{} 1$ so this inequality is equal with $ yx > y^2 \\plus{} y \\plus{} 1$\nwe have two case here if $ x \\geq y \\plus{} 2$ in every step of algorithm we can always reach a smaller solution and it should stop at (a,1) \nwhere a=3 or a=1.\n\n\ncase 2 : if x=y+1 somewhere, it is clear by parity there will be no solution.\nall of the cases we observe that k=5 and with this algorithm we can reach infinite number of solutions[/hide]\r\nnot a solution",
"Solution_2": "I think there are infinitely many solutions. I've got $ (1,3),(3,13),(13,61),\\ldots$. From one solution, we can generate another.. I don't know how to write the general form, but we certainly cannot use Vieta jumping.",
"Solution_3": "edit: we use vieta jumping for proving if this number is an integer it is 5. it might help to find recursive formula",
"Solution_4": "with $ x_1 \\equal{} 1$ and $ x_2 \\equal{} 1$ all solutions are $ (x_n,x_{n\\plus{}1})$ \r\n$ x_{n \\plus{} 2} \\equal{} 5.x_{n \\plus{} 1} \\minus{} x_n \\minus{} 1$",
"Solution_5": "sorry,Where can i find some information about Vieta jumping? :!:",
"Solution_6": "[quote=\"stvs_f\"]sorry,Where can i find some information about Vieta jumping? :!:[/quote]\r\n\r\nMaybe http://www.google.fr/search?q=vieta+jumping \r\n\r\n :P",
"Solution_7": "[quote=\"stvs_f\"]sorry,Where can i find some information about Vieta jumping? :!:[/quote]\r\n\r\nYimin Ge from austria wrote a great paper about it. [url]http://www.georgmohr.dk/tr/tr09taltvieta.pdf[/url]"
}
{
"Tag": [
"logarithms"
],
"Problem": "How do i obtain the logarithm of an expression like\r\n\r\nx = (a^m - b^m) / (a^n - b^n) also m > n (don't think this is significant)\r\n\r\n[hide]\n\nI tried \n\nmlog(a/b) - nlog(a/b) .. know this is wrong cant just figure out why :(\n\n[/hide]",
"Solution_1": "Let me re-type this:\r\n\r\n$ x\\equal{}\\frac{a^{m}\\minus{}b^{m}}{a^{n}\\minus{}b^{n}}$\r\n\r\nWhen you say \"obtain the logarithm\", what do you mean? Now to take the log of the expression, or express $ x$ using logs?",
"Solution_2": "[quote=\"JRav\"]Let me re-type this:\n\n$ x \\equal{} \\frac {a^{m} \\minus{} b^{m}}{a^{n} \\minus{} b^{n}}$\n\nWhen you say \"obtain the logarithm\", what do you mean? Now to take the log of the expression, or express $ x$ using logs?[/quote]\r\n\r\nOkay to be more clear. I need to obtain log x",
"Solution_3": "hello, it is clear that\r\n$ \\log(x)\\equal{}\\log\\left(\\frac{a^m\\minus{}b^m}{a^n\\minus{}b^n}\\right)\\equal{}\\log(a^m\\minus{}b^m)\\minus{}\\log(a^n\\minus{}b^n)$ with $ a^m\\minus{}b^m>0$ and $ a^n\\minus{}b^n>0$.\r\nSonnhard.",
"Solution_4": "[quote=\"Dr Sonnhard Graubner\"]hello, it is clear that\n$ \\log(x) \\equal{} \\log\\left(\\frac {a^m \\minus{} b^m}{a^n \\minus{} b^n}\\right) \\equal{} \\log(a^m \\minus{} b^m) \\minus{} \\log(a^n \\minus{} b^n)$ with $ a^m \\minus{} b^m > 0$ and $ a^n \\minus{} b^n > 0$.\nSonnhard.[/quote]\r\n\r\nI found out my mistake!!\r\n\r\nBtw, Is there any way to solve/Simplify further $ \\log\\left(a^m \\minus{} b^m\\right)$ in general?",
"Solution_5": "[quote=\"solveitall\"]\nBtw, Is there any way to solve/Simplify further $ \\log\\left(a^m \\minus{} b^m\\right)$ in general?[/quote]\r\n\r\n$ \\log\\left(a^m \\minus{} b^m\\right)$\r\n\r\n=$ \\frac{\\log a^m}{\\log b^m}$\r\n\r\n=$ \\frac{m\\log a}{m\\log b}$\r\n\r\nAny mistakes?",
"Solution_6": "Yes, there is a fundamental mistake. You're making your own rule for logarithms: difference of logarithms is a quotient of arguments, but the difference of arguments ISN'T quotient of logarithms.\r\n\r\n$ \\log_2 (32 \\minus{} 16) \\equal{} \\frac {\\log_2 32}{\\log_2 16} \\equal{} \\frac {5}{4}$\r\n\r\nbut, $ \\log_2 (32 \\minus{} 16) \\equal{} 4$",
"Solution_7": "[quote=\"Euclid_great\"][quote=\"solveitall\"]\nBtw, Is there any way to solve/Simplify further $ \\log\\left(a^m \\minus{} b^m\\right)$ in general?[/quote]\n\n$ \\log\\left(a^m \\minus{} b^m\\right)$\n\n=$ \\frac {\\log a^m}{\\log b^m}$\n\n=$ \\frac {m\\log a}{m\\log b}$\n\nAny mistakes?[/quote]\r\n\r\nYes there is. This is the same trap which i too fell into before i realised. Difference inside cant be expressed in the form of $ a/b$\r\n\r\nRight now the only way out to simplify this i can think of this expressing\r\n\r\n$ \\left(a^m \\minus{} b^m\\right) as \\left(a \\minus{} b\\right) \\left(a^{n\\minus{}1}...\\right)$ :(",
"Solution_8": "Ah yes, a very fundamental mistake I might want to highlight in a problem I will make in the TPP (see link below).",
"Solution_9": "It's not worth it to try to simplify some binomial expansion. Just leave it as it is."
}
{
"Tag": [],
"Problem": "H,A,T,B,O, and X are digits from 0 to 9. Repeats are allowed.HATBOX and BOXHAT are six digit numbers.\r\n$ 9\\times HATBOX \\equal{} 4\\times BOXHAT$\r\nFind a HATBOX and BOXHAT for which this is true.",
"Solution_1": "I think I might be on to something here, but I can't take it to the end...\r\n\r\nAssign to $ x$ the 3-digit number $ HAT$, and $ BOX$ to $ y$.\r\n\r\nSo...\r\n\r\n$ 9(1000x\\plus{}y) \\equal{} 4(1000y\\plus{}x)$\r\n\r\n$ \\frac{9}{4}\\equal{}\\frac{1000y\\plus{}x}{1000x\\plus{}y}$...so...taking the fraction into two equations...with k being some number\r\n\r\n$ 9k \\equal{} 1000y\\plus{}x$ and $ 4k \\equal{} 1000x\\plus{}y$, and adding them together...\r\n\r\n$ 13k \\equal{} 1001y\\plus{}1001x$, so $ x\\plus{}y \\equal{}\\frac{13k}{1001}$\r\n\r\nThis seems promising to me, except for the fact that I don't have an answer...and I don't know how to proceed from here. And I guess all the digits could just be 0, but I think there are more solutions.\r\n\r\nEDIT: Well...what if $ k$ was just assigned some multiple of 1001 such as...100100...then the numerator would be 1300, which could be the sum of many different 3 digit numbers... I extremely doubt it's right though, as it doesn't make too much sense on inspection...",
"Solution_2": "[quote=\"Icy\"]I think I might be on to something here, but I can't take it to the end...\n\nAssign to $ x$ the 3-digit number $ HAT$, and $ BOX$ to $ y$.\n\nSo...\n\n$ 9(1000x\\plus{}y) \\equal{} 4(1000y\\plus{}x)$\n\n$ \\frac{9}{4}\\equal{}\\frac{1000y\\plus{}x}{1000x\\plus{}y}$...so...taking the fraction into two equations...with k being some number\n\n$ 9k \\equal{} 1000y\\plus{}x$ and $ 4k \\equal{} 1000x\\plus{}y$, and adding them together...\n\n$ 13k \\equal{} 1001y\\plus{}1001x$, so $ x\\plus{}y \\equal{}\\frac{13k}{1001}$\n\nThis seems promising to me, except for the fact that I don't have an answer...and I don't know how to proceed from here. And I guess all the digits could just be 0, but I think there are more solutions.\n\nEDIT: Well...what if $ k$ was just assigned some multiple of 1001 such as...100100...then the numerator would be 1300, which could be the sum of many different 3 digit numbers... I extremely doubt it's right though, as it doesn't make too much sense on inspection...[/quote]\r\n\r\n 13 is a factor of 1001. :wink:",
"Solution_3": "So...$ \\frac{1001}{13}\\equal{}77$...but I don't know where to go from there....\r\n$ k\\equal{}77(x\\plus{}y)$, but I think I'm still missing something obvious that'll give an answer.",
"Solution_4": "[quote=\"Icy\"]I think I might be on to something here, but I can't take it to the end...\n\nAssign to $ x$ the 3-digit number $ HAT$, and $ BOX$ to $ y$.\n\nSo...\n\n$ 9(1000x\\plus{}y) \\equal{} 4(1000y\\plus{}x)$\n\n$ \\frac{9}{4}\\equal{}\\frac{1000y\\plus{}x}{1000x\\plus{}y}$...so...taking the fraction into two equations...with k being some number\n\n$ 9k \\equal{} 1000y\\plus{}x$ and $ 4k \\equal{} 1000x\\plus{}y$, and adding them together...\n\n$ 13k \\equal{} 1001y\\plus{}1001x$, so $ x\\plus{}y \\equal{}\\frac{13k}{1001}$\n\nThis seems promising to me, except for the fact that I don't have an answer...and I don't know how to proceed from here. And I guess all the digits could just be 0, but I think there are more solutions.\n\nEDIT: Well...what if $ k$ was just assigned some multiple of 1001 such as...100100...then the numerator would be 1300, which could be the sum of many different 3 digit numbers... I extremely doubt it's right though, as it doesn't make too much sense on inspection...[/quote]\r\n\r\n$ \\frac{9}{4}\\equal{}\\frac{1000y\\plus{}x}{1000x\\plus{}y}$\r\n$ 9(1000x\\plus{}y)\\equal{}4(1000y\\plus{}x)$\r\n$ 9000x\\plus{}9y\\equal{}4000y\\plus{}4x$\r\n$ 8996x\\equal{}3991y$\r\n$ 692x\\equal{}307y$\r\nso then if $ x\\equal{}307$ and $ y\\equal{}692$\r\n$ 2769228\\equal{}2769228$",
"Solution_5": "Are there any more solutions?",
"Solution_6": "If 692 and 307 are relatively prime (haven't checked, but I think so) then there are no more solutions.",
"Solution_7": "Well 307 is prime and 692 is not a multiple of 307 so I am pretty sure 307 and 692 are relatively prime :P ."
}
{
"Tag": [
"real analysis",
"real analysis solved"
],
"Problem": "compute the limit:\r\n\r\nlim(x-->0) x^4 (1+2+..+[1/x^2]) \r\nwhere [a]=a-{a}\r\n\r\nthank you very much! :)",
"Solution_1": "Let S(x) = 1+2+...+[1/x 2 ].\r\n\r\nAssume that x2 < 1.\r\n\r\nLet k be the integer such that 1/(k+1) < 1/x 2 \\leq 1/k.\r\nThen k/(2(k+1)) \\leq x4 S(x) \\leq (k+1)/(2k).\r\n\r\nSince k tends to +oo as x tends to 0, we deduce that lim x 4 S(x) = 1/2.\r\n\r\nPierre.",
"Solution_2": "1+2+..+[1/x^2]\r\nwhat are the terms after 2, the general term in the previous sum?",
"Solution_3": "but\r\n[quote]Let k be the integer such that 1/(k+1) < 1/x 1/k. [/quote]\r\nsince x tends to 0=>1/x tends to oo results that it can't be smaller than 1/k with k integer\r\ni think u meant k<=1/x0$, then $ \\phi(n\\plus{}1)\\equal{}2\\phi(n)$.\r\nProof: $ \\forall{y}\\in\\mathbb{R},\\phi(n)\\equal{}\\int_{\\minus{}\\infty}^{\\infty}\\binom{n}{x\\plus{}y}dx$; moreover $ \\forall{x,y}\\in\\mathbb{R},\\binom{n}{x}\\plus{}\\binom{n}{x\\plus{}1}\\equal{}\\binom{n\\plus{}1}{x\\plus{}1}$. Thus $ \\phi(n)\\plus{}\\phi(n)\\equal{}\\phi(n\\plus{}1)$. QED",
"Solution_2": "It is actually much better than \"a partial result\". Note that the identity is true for complex $ n$ with positive real part as well, so, when we multiply the integral by $ 2^{ \\minus{} z}$, we get an analytic function that is $ 1$-periodic in the right half-plane and, therefore, can be extended to the entire plane. Now all we need to do is to show that this function grows not too fast when we go up the imaginary axis (any estimate short of $ e^{\\pi|z|}$ will suffice) to conclude that it is actually constant. This reduces the problem to the [b]estimation[/b] of the integral instead of its [b]exact computation[/b]. I have no time to do it now but I hope it is not too difficult :)."
}
{
"Tag": [
"puzzles"
],
"Problem": "Which word from the bottom group belongs in the top group?\r\n\r\nblast,paper,box,bank\r\n\r\njuice,bag,cradle,carpet\r\n\r\nwhy?",
"Solution_1": "[hide]bag, because the first letter is either a b or an upside-down b[/hide]",
"Solution_2": "[quote=\"mustafa\"][hide]bag, because the first letter is either a b or an upside-down b[/hide][/quote]\r\n\r\nur answer is right but the reason isn't wat I was thinking of",
"Solution_3": "[quote=\"Ihatepie\"][quote=\"mustafa\"][hide]bag, because the first letter is either a b or an upside-down b[/hide][/quote]\n\nur answer is right but the reason isn't wat I was thinking of[/quote]What's your reason?",
"Solution_4": "try to figure it out \r\n\r\n[hide=\"the reason\"]u can put the word sand before each word and get another word.[/hide]",
"Solution_5": "[quote=\"Ihatepie\"]try to figure it out \n\n[hide=\"the reason\"]u can put the word sand before each word and get another word.[/hide][/quote]\r\n\r\nthat is creative..did u just think of that...or was that liek the actual answer? :D",
"Solution_6": "i've heard problems like that b4 so I made up a new one."
}
{
"Tag": [
"topology",
"real analysis",
"real analysis unsolved"
],
"Problem": "Hi guys,\r\n\r\nI have been working on this problem and made some progress but now I am stuck.\r\n\r\nSuppose A is an arbitrary indexing set, and C={B_a : a in A} is a collection of closed balls in R^n. Let E be the union of the elements of C and let D be the boundary of E. Prove D has measure zero.\r\n\r\nOk well the problem suggested to first assume that all the balls have radius at least one. I did that and managed to solve the problem, but now I am stuck because I am not sure how to use my case to solve the more general case. Any advice? P.S. This is a homework assignment so plz don't finish it for me, just give me a hint on how I can use my case to solve the more general one.\r\n\r\nThanks.",
"Solution_1": "You've got a wrong problem here: you can easily construct a closed set of positive measure with empty interior. Its complement can be represented as a union of closed balls of positive radii. I'm afraid that your instructor merely forgot that the boundary of the union may be larger than the union of the boundaries :)"
}
{
"Tag": [
"ARML"
],
"Problem": "Can any one post problem ARML 2000-2008?",
"Solution_1": "See [url]http://www.artofproblemsolving.com/Forum/viewtopic.php?t=201069[/url] for problems after 2003."
}
{
"Tag": [
"inequalities proposed",
"inequalities"
],
"Problem": "let a,b,c >0\r\na+b-c>0\r\nb+c-a>0\r\nc+a-b>0\r\nFind min $ \\sum \\frac {a}{\\sqrt{a^2 \\plus{} 3bc}}$",
"Solution_1": "[quote=\"fesha3t\"]let a,b,c >0\na+b-c>0\nb+c-a>0\nc+a-b>0\nFind min $ \\sum \\frac {a}{\\sqrt {a^2 \\plus{} 3bc}}$[/quote]\r\nSee here:\r\nhttp://www.artofproblemsolving.com/Forum/viewtopic.php?t=163238"
}
{
"Tag": [
"geometry"
],
"Problem": ":?: ABCD is a convex quadrilateral. Diagonals AC and BD are joined. Given tht area of pentagon ABOCD = Ar(BOC). Now two points are located on BC ( P and Q) such tht OP||AB and OQ||DC.\r\nProve that : 2 Ar(POQ) = Ar(AOB) + Ar(COD).",
"Solution_1": "Cmon guys pls answer this one."
}
{
"Tag": [],
"Problem": "A standard deck of cards has 52 cards divided into 4 suits, each of which has 13 cards. Two of the suits ($ \\heartsuit$ and $ \\diamondsuit$, called `hearts' and `diamonds') are red, the other two ($ \\spadesuit$ and $ \\clubsuit$, called `spades' and `clubs') are black. The cards in the deck are placed in random order (usually by a process called `shuffling'). In how many different ways can we pick two red cards? (Note: Order matters in the sense that drawing an ace of hearts followed by jack of diamonds is different than drawing a jack of diamonds followed by ace of hearts.)",
"Solution_1": "26P2\r\n$ 26*25\\equal{}\\boxed{650}$",
"Solution_2": "Can anyone explain why we don't divide by 2 in this case",
"Solution_3": "@above You only divide by 2 if order doesn't matter. However, the problem says that order matters, so you don't divide by 2.",
"Solution_4": "[hide=sol]The first position has $26$ options, the second has $25$, so the answer is $26\\cdot25=\\boxed{650}$[/hide]\n[hide=Why my answer is good]It is nice and quick. [/hide]"
}
{
"Tag": [
"geometry",
"rectangle",
"search",
"combinatorics proposed",
"combinatorics"
],
"Problem": "It is problem numbers six from one of the TST's of IRAN.\r\nWe have a polygon which can be coverd by $100$ rectangles but can not be covered by $99$ rectangles .\r\nProve that it can not be covered by $100$ triangles.\r\n\r\nBy covering I mean that all area of the shape must be covered and non of the rectangles shall cover each other.",
"Solution_1": "I really felt the life with solving this:\r\nHere are some things i have done ,but it is the most usuall approch :\r\n\r\n1.Each side of this polygon is prependicular to the two other sides near it.\r\n2.our dear polygon has even number of sides.\r\n3.any polygon with $2n$ sides can be covered with n-1 triangle(here with $\\frac{2*100}{2}-1$)",
"Solution_2": "I remember seeing this problem before around here. I think Myth solved it, and it was posted by a Romanian member of the forum, but I forgot which one. Anyway, you have enough key words to search for it (Myth, polygons, triangles, etc.) :).",
"Solution_3": "The last move for the bane of the polygon:\r\n\r\nThis polygon can not be covered by $99$ rectangles so $n-1 > 99$ .\r\nSo our dear polygon has at least $202$ sides.\r\nSo it has ,at least, $101\\pi$ degree in its internal angles.\r\nBut i assure you $99$ trinagle can not provide that much angles to cover the polygon.",
"Solution_4": "Well i didnt manage to find the one that grobber mentioned but i found a nice\r\ntopic somewhere from Myth to his very few friends:\r\n\r\n$\\text{Myth has a heart too}$\r\n\r\n[url]http://www.artofproblemsolving.com/Forum/viewtopic.php?highlight=covering&t=1896[/url]"
}
{
"Tag": [],
"Problem": "If $x$ is a positive integer find te solutions of \r\n\r\n$2^x\\equiv 2x (mod 3)$",
"Solution_1": "$2^6 = 1\\ mod \\ 3$ so $2^x - 2 \\cdot x$ is periodic with period at most $6$. Now it just the matter of 6 easy computations.\r\n\r\n$x \\in 1+6 \\cdot \\mathbb{N}$ or $x \\in 2+6 \\cdot \\mathbb{N}$",
"Solution_2": "It is a godd solution ,but not elementary at all. does ayone has an elementary solution?",
"Solution_3": "Well, $2^x \\equiv 2x$ is modulo $3$ equivalent with $(-1)^x \\equiv -x$. Now consider two cases:\r\n\r\n$x \\equiv 0 \\mod 2$, then $1\\equiv -x \\mod 3$, i.e. $x \\equiv 2 \\mod 3$. This has exactly one solution mod $6$ (Chinese remainder theorem); $x \\equiv 2 \\mod 6$.\r\n\r\n$x \\equiv 1 \\mod 2$, then $-1 \\equiv -x \\mod 3$, i.e. $x \\equiv 1 \\mod 3$. And this would result in $x \\equiv 1 \\mod 6$.\r\n\r\nSo, the solution set is $S=\\{x\\ |\\ x \\equiv 1 \\mod 6 \\ \\ \\wedge \\ \\ x \\equiv 2 \\mod 6\\}$"
}
{
"Tag": [
"induction",
"inequalities",
"algebra unsolved",
"algebra"
],
"Problem": "Let $ (w_{n})$ $ n\\in{\\mathbb{N}}$\r\n$ w_{n}\\equal{}\\prod_{k\\equal{}1}^{n}(1\\plus{}\\frac{1}{k^{3}})$\r\nprove that:\r\n$ (\\forall{n}\\in{\\mathbb{N}\\minus{}{0}})$ $ w_{n}<3$",
"Solution_1": "[quote=\"reason\"]Let $ (w_{n})$ $ n\\in{\\mathbb{N}}$\n$ w_{n} \\equal{} \\prod_{k \\equal{} 1}^{n}(1 \\plus{} \\frac {1}{k^{3}})$\nprove that:\n$ (\\forall{n}\\in{\\mathbb{N} \\minus{} {0}})$ $ w_{n} < 3$[/quote]\r\nProve by induction that the inequality $ \\prod_{k \\equal{} 1}^{n}(1 \\plus{} \\frac {1}{k^{3}}) < 3 \\minus{} \\frac{1}{n}$ holds for every $ n > 1$.\r\n\r\nNow try to prove the much stronger inequality :\r\n\r\n $ \\prod_{k \\equal{} 1}^{n}(1 \\plus{} \\frac {1}{k^{3}}) < \\frac{49}{20}$\r\n\r\nBTW,take a look here for some more interesting facts :\r\n\r\nhttp://www.mathlinks.ro/viewtopic.php?t=276003"
}
{
"Tag": [
"search",
"abstract algebra",
"superior algebra",
"superior algebra theorems"
],
"Problem": "Does anybody know why Lagrange's Theorem is attributed to him?? After all Guass never used the concept of a group in his famous Disquisitiones Arithmeticae(1801), so Lagrange could have hardly formulated such a result himself. Please have sources to back up your information. Also, does anyone know how Cauchy himself( if he did ), proved Cauchy's Theorem: that for any prime, p, dividing a finite group, the group has an element of order p. Did Cauchy really know about the Class Equation of a group??? I will be researching these questions myself, so I will let others know. If I should not be proposing such questions here, let me know also.\r\n\r\nMozart",
"Solution_1": "Lagrange essentially proved his theorem for Galois groups. He did not call the set of permutations a group though. Here is a reference (in Russian :oops: )\r\nS. S. Gindikin, \"Rasskazy o fizikah i matematikah\", 3rd edition, Moscow, MCCME, IMU, 2001, page 281.\r\nhttp://www.mccme.ru/free-books/gindikin/lagrange.zip",
"Solution_2": "I asked a colleague of mine who knows more about history than I do, and this is his reply:\r\n\r\n[quote]Here is what I know:\n\nThe word group is due to Galois who also proved Lagrange's periodicty of continued fractions theorem.\n\nLagrange did discuss the set of permutations that fixed polynomials of many variables--he probably observed easily that the set was closed under multiplication and also that the size had to be a divisor of n!\n\nGauss used cosets without naming them so--and he certainly used the fact that they partiition the group (again without using the word)\n\nI would consult Kline (the big book)--I do remember reading a long article on Lagrange in Historia Mathematica (I wonder if that is on line). Another book would be Van Der Waerden History of Algebra (but a quick scan did not provide any obvious references). It did say that absatract groups do not occur until 1870 by Jordan and Kronecker.\n[/quote]",
"Solution_3": "Do you remember the volume # in Historia Mathematica. It appears to be online,but there is no way to search the document. Thanks."
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "Prove:There exists no positive integer $ n$ such that for every rational number $ r$,\r\n$ \\exists b,a_1,a_2,...,a_n \\in Z$ so that $ r\\equal{}b\\plus{}\\frac{1}{a_1}\\plus{}...\\plus{}\\frac{1}{a_n}$",
"Solution_1": "[url=http://www.mathlinks.ro/Forum/viewtopic.php?t=188110&search_id=529936752]See here[/url]",
"Solution_2": "Next time when someone links to an old thread I suggest continuing there :wink:"
}
{
"Tag": [
"function",
"induction",
"algebra unsolved",
"algebra"
],
"Problem": "Find all functions $ f: \\mathbb Z \\rightarrow \\mathbb Z$ satisfying:\n(i) $ f(0)\\equal{}1$,\n(ii) $ f(f(n))\\equal{}n$, and\n(iii) $ f(f(n\\plus{}2)\\plus{}2)\\equal{}n$ for all integers $\\mathbb Z$.",
"Solution_1": "[quote=\"amparvardi\"]find all functions $ f: Z \\rightarrow Z$ satisfying:\n$ ( \\imath)$ $ f(0) \\equal{} 1$\n$ ( \\imath \\imath)$ $ f(f(n)) \\equal{} n$\n$ (\\imath \\imath \\imath)$ $ f(f(n \\plus{} 2) \\plus{} 2) \\equal{} n$ . $ \\forall n \\in Z$[/quote]\r\n\r\n$ f(f(n))\\equal{}n$ shows that $ f(n)$ is injective.\r\n\r\nThen $ f(f(n\\plus{}2)\\plus{}2)\\equal{}n\\equal{}f(f(n))$ $ \\implies$ $ f(n\\plus{}2)\\equal{}f(n)\\minus{}2$\r\n\r\nThen $ f(0)\\equal{}1$ and $ f(n\\plus{}2)\\equal{}f(n)\\minus{}2$ $ \\implies$ $ f(2n)\\equal{}1\\minus{}2n$\r\n\r\nThen $ f(0)\\equal{}1$ $ \\implies$ $ f(1)\\equal{}f(f(0))\\equal{}0$ and $ f(n\\plus{}2)\\equal{}f(n)\\minus{}2$ $ \\implies$ $ f(2n\\plus{}1)\\equal{}\\minus{}2n\\equal{}1\\minus{}(2n\\plus{}1)$\r\n\r\nAnd it is easy to check back that this indeed is a solution.\r\n\r\nHence the solution: $ \\boxed{f(x)\\equal{}1\\minus{}x}$",
"Solution_2": "[quote=\"pco\"]\n\nThen $ f(0) \\equal{} 1$ and $ f(n \\plus{} 2) \\equal{} f(n) \\minus{} 2$ $ \\implies$ $ f(2n) \\equal{} 1 \\minus{} 2n$\n\n[/quote]\r\n\r\nHow do you can say since ''$ f(0) \\equal{} 1$ and $ f(n \\plus{} 2) \\equal{} f(n) \\minus{} 2$ $ \\implies$ $ f(2n) \\equal{} 1 \\minus{} 2n$''?",
"Solution_3": "[quote=\"amparvardi\"][quote=\"pco\"]\n\nThen $ f(0) \\equal{} 1$ and $ f(n \\plus{} 2) \\equal{} f(n) \\minus{} 2$ $ \\implies$ $ f(2n) \\equal{} 1 \\minus{} 2n$\n\n[/quote]\n\nHow do you can say since ''$ f(0) \\equal{} 1$ and $ f(n \\plus{} 2) \\equal{} f(n) \\minus{} 2$ $ \\implies$ $ f(2n) \\equal{} 1 \\minus{} 2n$''?[/quote]\r\n\r\nIn order to see it :\r\n$ f(0)\\equal{}1$ and then $ f(2)\\equal{}f(0\\plus{}2)\\equal{}f(0)\\minus{}2\\equal{}\\minus{}1$ and then $ f(4)\\equal{}f(2\\plus{}2)\\equal{}f(2)\\minus{}2\\equal{}\\minus{}3...$\r\n$ f(0)\\equal{}1$ and then $ f(\\minus{}2)\\equal{}f(0\\minus{}2)\\equal{}f(0)\\plus{}2\\equal{}3$ and then $ f(\\minus{}4)\\equal{}f(\\minus{}2\\minus{}2)\\equal{}f(\\minus{}2)\\plus{}2\\equal{}5...$\r\n\r\nIn order to prove it : just use induction"
}
{
"Tag": [
"logarithms",
"trigonometry",
"limit",
"algebra",
"polynomial",
"function",
"real analysis"
],
"Problem": "For x > 0, lim ((sin x)^1/x + (1/ x)^ sin x ) is\r\n x----0",
"Solution_1": "Well, IMHO this is a Real analysis problem...\r\n\r\nnote that $ \\ln (\\sin x)^{\\frac {1}{x}} \\equal{} \\frac {\\ln \\sin x}{x}$.\r\n\r\n$ \\lim_{x\\to 0}\\frac {\\ln \\sin x}{x} \\equal{} \\minus{} \\infty$ (prove it using L'H\u00f4pital rule or knowing the growth rate of log and polynomial function) so $ \\lim_{x\\to 0} (\\sin x)^{\\frac {1}{x}} \\equal{} e^{ \\minus{} \\infty} \\equal{} 0$.\r\n\r\nOn the other hand, $ \\ln \\left(\\frac {1}{x}\\right)^{\\sin x} \\equal{} \\minus{} \\sin x \\ln x$\r\n\r\n$ \\lim_{x\\to 0} \\minus{} \\sin x \\ln x \\equal{} 0$ which you can also prove knowing that sine is asymptotically equivalent to $ x$ in the limit process towards zero, or by L'H\u00f4pital's rule, taking the sine down in the denominator.\r\n\r\nHence, $ \\lim_{x\\to 0}\\left(\\frac {1}{x}\\right)^{\\sin x} \\equal{} e^{0} \\equal{} 1$\r\n\r\nand as both limits exist, your desired limit is their sum: $ 1$.",
"Solution_2": "thanks sir"
}
{
"Tag": [
"linear algebra",
"matrix"
],
"Problem": "Solve for the system of linear equations\r\n\r\n\r\n[img]http://webwork2.math.utah.edu/webwork_system_html/tmp/equations/f0/693379284ca3188721313e621d10211.png[/img]\r\n\r\n\r\n\r\n\r\n\r\nThanks",
"Solution_1": "[hide]\nInput the 4x5 matrix on your trusty TI-84 and use rref\n\n$(w,x,y,z)=(2,4,-1,3)$.[/hide]",
"Solution_2": "ok, thanks for that answer. i already knew how to do that. the problem is when i go to solve it manually i dont get that answer. Anyone what to take the time and post the work on this. I know theres going to be alot",
"Solution_3": "well, im sure playing around with substitution and subtracting/adding the equations will also work... eventually. Right?",
"Solution_4": "hello, you can solve this system by the so called Gaussian elimination, see here\r\nhttp://mathworld.wolfram.com/GaussianElimination.html\r\nSonnhard.",
"Solution_5": "Add 4 times the 3rd to the 1st to obtain $ 13w\\minus{}2x\\minus{}5y\\equal{}23$. Add 2 times the 3rd to the 2nd to obtain $ 8w\\minus{}3x\\minus{}3y\\equal{}7$. Subtract 3 times the 3rd from the 4th to obtain $ \\minus{}5w\\minus{}x\\plus{}4y\\equal{}\\minus{}18$. Now we wish to solve the systems: \r\n$ 13w\\minus{}2x\\minus{}5y\\equal{}23$ $ (1)$\r\n$ 8w\\minus{}3x\\minus{}3y\\equal{}7$ $ (2)$\r\n$ \\minus{}5w\\minus{}x\\plus{}4y\\equal{}\\minus{}18$ $ (3)$\r\n\r\nSubtract 2 times the 3rd from the 1st to get $ 23w\\minus{}13y\\equal{}59$. Subtract 3 times the 3rd from the 2nd to get $ 23w\\minus{}15y\\equal{}61$. Now simply subtract these two equations to get $ 2y\\equal{}\\minus{}2$ so $ y\\equal{}\\minus{}1$ and from this $ w\\equal{}2$. Then everything else follows: $ x\\equal{}4$ and finally $ z\\equal{}3$."
}
{
"Tag": [
"calculus",
"integration",
"trigonometry",
"function",
"calculus computations"
],
"Problem": "Can we use $\\int_{0}^{1}x^{x}\\ dx=\\sum_{n=1}^{\\infty}\\frac{(-1)^{n+1}}{n^{n}}$ to solve this $\\int_{0}^{\\pi}(\\sin{x})^{\\sin{x}}\\ dx$ or $\\int_{0}^{1}(\\sin{x})^{\\sin{x}}\\ dx$ ?",
"Solution_1": "[quote=\"kg2100\"]Can we use $\\int_{0}^{1}x^{x}\\ dx=\\sum_{n=1}^{\\infty}\\frac{(-1)^{n+1}}{n^{n}}$ to solve this $\\int_{0}^{\\pi}(\\sin{x})^{\\sin{x}}\\ dx$ or $\\int_{0}^{1}(\\sin{x})^{\\sin{x}}\\ dx$ ?[/quote]\r\n\r\nLet me try, instead,\r\n$\\int_{0}^{\\pi/2}\\sin x^{\\sin x}dx$\r\nWrite this as,\r\n$\\int_{0}^{\\pi /2}\\sin x^{\\sin x}\\cdot \\frac{\\cos x}{\\cos x}dx = \\int_{0}^{\\pi /2}\\sin x^{\\sin x}\\cdot \\frac{\\cos x}{\\sqrt{1-\\sin^{2}x}}dx$\r\n\r\nLet $t=\\sin x$ then,\r\n\r\n$\\int_{0}^{1}t^{t}\\cdot \\frac{1}{\\sqrt{1-t^{2}}}dt$\r\n\r\nBut I dont think this can be simplified, to $\\int_{0}^{1}t^{t}dt$. :huh:",
"Solution_2": "no. we cant. because x is a variable, not a function, so we cannot make a substituin of x for a function",
"Solution_3": "[quote=\"yagaron\"]no. we cant. because x is a variable, not a function, so we cannot make a substituin of x for a function[/quote]\r\nWhat do you mean?",
"Solution_4": "Suppose we have ${x^{x}dx}$\r\n\r\nLet's make a substituition ${x=\\sin (y)}$\r\n\r\n${x^{x}dx=\\cos (y) \\sin^{\\sin (y)}(y) dy}$\r\n\r\nwe will get the same solution as your formula if and only if ${\\cos (y)=1}$ for all $y$, but this is impossible.\r\n\r\nMore general case\r\n\r\n${f(x) dx}$\r\n${x=G(y)}$\r\n${f(x) dx= f(G(y)) G'(y) dy}$\r\n\r\nWill have the same formula if and only if ${G'(y)=1\\land G(y)=y}$\r\n\r\nso we cannot make this substituition"
}
{
"Tag": [
"geometry"
],
"Problem": "Three circles, $ C_{1}, C_{2}$ and $ C_{3}$ intersect. $ C_{1}$ intersects $ C_{2}$ at $ X$ and $ W$. $ C_{1}$ intersects $ C_{3}$ at $ W$ and $ Z$, $ C_{2}$ intersects $ C_{3}$ at $ W$ and $ Y$.\r\n\r\nFrom the point $ A$ on $ C_{1}$ a line is drawn through $ X$ to meet $ C_{2}$ at $ B$. From $ B$ a line is drawn through $ Y$ to meet $ C_{3}$ at $ C$. From $ C$ a line is drawn through $ Z$ to meet $ C_{1}$ at $ A$. So we have a triangle $ ABC$. Is it necessarily the case that we get a triangle?\r\n\r\nBeginning with another point $ A'$ on $ C_{1}$, the process is repeated to get another triangle $ A'B'C'$ Are the triangles similar? Are they congruent?\r\n\r\nSorry I can't post a diagram, but I hope this is sufficient to draw a picture.",
"Solution_1": "[hide=\" One approach\"]\n\nJoin $ X$, $ Y$ and $ Z$ to $ O$, the point where all three circles meet.\n\nThen angle chasing shows that $ OYCZ$ is a cylic quadrilateral since $ AXOZ$ and $ XBYO$ are cyclic quadrilaterals.\n \nFor the second part, it seems that the trianlges are similar but not necessaily congruent.\n\n[/hide]"
}
{
"Tag": [
"LaTeX",
"quadratics",
"algebra",
"quadratic formula"
],
"Problem": "5x(1+1/x^2+y^2)=12\r\n5y(1-1/x^2+y^2)=4",
"Solution_1": "hello, do you mean \r\n$ 5x\\left(1\\plus{}\\frac{1}{x^2\\plus{}y^2}\\right)\\equal{}12$\r\n$ 5y\\left(1\\minus{}\\frac{1}{x^2\\plus{}y^2}\\right)\\equal{}4$\r\nSonnhard.",
"Solution_2": "ya, i didn't know how to type in latex",
"Solution_3": "hello, writing your both equations in the form\r\n$ 1\\plus{}\\frac{1}{x^2\\plus{}y^2}\\equal{}\\frac{12}{5x}$\r\n$ 1\\minus{}\\frac{1}{x^2\\plus{}y^2}\\equal{}\\frac{4}{5y}$\r\nand adding them we get\r\n$ \\frac{6}{5x}\\plus{}\\frac{2}{5y}\\equal{}1$ and solving for $ x$ we get\r\n$ x\\equal{}\\frac{6y}{5y\\minus{}2}$\r\nsetting this in your equation (1) and simplifying and factoring we get\r\n$ (y\\minus{}1)(5y\\plus{}1)(5y^2\\minus{}4y\\plus{}4)\\equal{}0$, from here you will get all solutions of your system.\r\nSonnhard.",
"Solution_4": "Take (1)+ i (2) to get\r\n\r\n$ x\\plus{}yi\\plus{}\\frac{x\\minus{}yi}{x^2\\plus{}y^2}\\equal{}\\frac{12}{5}\\plus{}i\\cdot \\frac{4}{5}$\r\n\r\nLetting $ z\\equal{}x\\plus{}yi$, we get\r\n\r\n$ z\\plus{}\\frac{1}{z}\\equal{}\\frac{12}{5}\\plus{}i\\cdot \\frac{4}{5}$\r\n\r\nwe can use the quadratic formula to finish from here..."
}
{
"Tag": [],
"Problem": "Hi, I'm having a hard time figuring out how to do this;\r\n\r\nIf $ a,b,c,d$ are in continued proportion, meaning $ \\frac{a}{b}\\equal{}\\frac{b}{c}\\equal{}\\frac{c}{d}$, prove that\r\n\r\n$ \\frac{a}{b\\plus{}d} \\equal{} \\frac{c^3}{c^2d \\plus{}d^3}$\r\n\r\nIs there a way to do this by means of invertendo, alternando, componendo and dividendo?",
"Solution_1": "[hide=\"Hint\"]Let $ r\\equal{}\\frac{a}{b}$. Now express everything in terms of $ d$ and $ r$.[/hide]"
}
{
"Tag": [
"function",
"probability and stats"
],
"Problem": "The lower and upper deciles of any distribution are the points that mark off the lowest 10% and the highest 10%. What are the lower and upper deciles of the standard normal distribution?",
"Solution_1": "I doubt that there is a closed form formula for that one: take your computer and compute a good approximation ( the erf function should be helpful)"
}
{
"Tag": [],
"Problem": "A number $a$ has $27$ digits and all of them are equal to $1$.Which is the remainder in the division $a: 27$ ?\r\n\r\n[u] Babis[/u]",
"Solution_1": "[hide]$0$[/hide]",
"Solution_2": "So is the original number 111,111,111,111,111,111,111,111,111?\r\nAssuming it is that, [hide] \\[111,111,111,111,111,111,111,111,111/27=4115226337448559670781893\\]\nSince there are no decimals or fractions, the remainder is 0[/hide]",
"Solution_3": "If \r\n$a=a_{1}a_{2}....a_{27}$\r\nthen\r\n$a=a_{1}+10a_{2}+19a_{3}+a_{4}+...+19a_{27}$ in modulo 27\r\n$a=1+10+19+1+...+19=30\\times9=0$ in modulo 27",
"Solution_4": "wow i was kind of lost on that expl. could you explain more...?(it looks sort of like base conversion) :lol:",
"Solution_5": "[hide=\"You might want to think about it like this instead...\"] We can write this number as\n\n$111111111 \\cdot 1000000001000000001 = 111 \\cdot 1001001 \\cdot 1000000001000000001$\n\nNow,\n\n$111 = 99+9+3$\n$1001001 = 999999+999+3$\n$1000000001000000001 = 999999999999999999+999999999+3$\n\nWhich are all divisible by $3$, so the whole thing is divisible by $27$. \n\nYou could do this in a somewhat more advanced way if you remember that a number is divisible by $3, 9$ if and only if the sum of its digits is divisible by $3, 9$. This is basically because $10^{k}-1 = 999...$ which is divisible by $3, 9$. [/hide]",
"Solution_6": "wow thanks",
"Solution_7": "[hide]just take \\frac{111111111111111111111111111}{27} = 4115226337448559670781893 since this is a whole number then there is no remainder so the remainder = 0[/hide]",
"Solution_8": "Care to repeat that operation (without mechanical help) for something larger? Perhaps if I asked you to divide $\\underbrace{111...111}_{1997 times}$ by $1997$? :o"
}
{
"Tag": [],
"Problem": "[size=200]Teh MAZE[/size]\r\n\r\nI will allow up to 30 people to play. The object of the game is to go from the Start to the Finish. \r\nYou have no map, but it gradually grows as you travel with trial and error.\r\n\r\n[u]Rules[/u]\r\n\r\n1. I was evil in creating this maze. Acknowledge that.\r\n2. There are only 4 directions you can go: Up, Down, Right, Left (Do not post North, South, East, West. Just don't)\r\n3. To move, put it in the form: (your username) goes (up, down, right or left). For example, I might do: andrewjjiang97 goes right. (This is not implying anything) If you do not use this format, I will ignore your move.\r\n4. Do not post if you are not in this game.\r\n5. You do not have to put /join or something like that. Just post the direction you want to go in. (see format above)\r\n6. I will not draw any maps except for the one below.\r\n7. Have fun!!\r\n\r\n\r\n\r\nS\r\n\r\n\r\n\r\nS=start",
"Solution_1": "[b]Bluecarneal moves DOWN[/b]",
"Solution_2": "[i]bluecarneal goes down. S/he finds a note. There seems to be a word etched into paper.It has the word \"in\" on it.\n[/i]\r\n\r\nTo see what you have put /view inventory.",
"Solution_3": "me is he.\r\n\r\n[b]Bluecarneal moves left[/b]",
"Solution_4": "[i]bluecarneal hits a brick wall. Ouch!! If you hit 2 more walls, you die.[/i]",
"Solution_5": "[b]tennis123 moves down[/b]\r\nIf we die, are we simply out of the game? Do we only have 1 life?",
"Solution_6": "[b]math_explorer moves right[/b]\r\nSo can we make our own maps? (And post them here?)\r\nAre our mazes all the same? (Or we start at different places in the same maze? Or something else?)\r\nCan we interact with each other?\r\nCan we make random guesses on the forum about the design or anything?\r\nSorry for all these questions, but, well, I don't want to be modkilled. BTW, I'm a he as well :)",
"Solution_7": "[i]tennis123 goes down. S/he finds a note. There seems to be a word etched into paper. It has the word \"in\" on it.[/i]\r\n\r\n@tennis123 you have unlimited lives\r\n\r\n[i]math_explorer moves right. Nothing happens... [/i]\r\n\r\n@math_explorer Yes you can make your own maps and post then here. However you want to be the first to the finish. All the mazes are the same, but as you just saw, there are unlimited \"notes\". You can interact with each other.",
"Solution_8": "[b]math_explorer moves up[/b]\r\nWhen we die, do we go back to the origin?\r\nDo we really need the notes to, say, unlock something, or is it merely information? Or should you tell us that?",
"Solution_9": "[i]math_explorer moves up. When he reaches there, there seems to be a black metallic door. When you test the handle, it clicks open. Now, choose a direction to venture forth.[/i]\r\n\r\nYes, when you die you go back to the origin. I will not answer your last question.",
"Solution_10": "I don't see the map... :huh:",
"Solution_11": "I know. You're suppose to create it by trial and error",
"Solution_12": "[b]Bluecarneal moves up[/b]",
"Solution_13": "[i]bluecarneal goes up. Hey, isn't this were I started? he wonders...[/i]\r\n\r\nJust FYI, you can post multiple moves at a time.",
"Solution_14": "btw I'm to the left of where I started.\r\n\r\nbluecarneal moves up, left, up, right, as far as walls allow.",
"Solution_15": "pacman moves right, up, up, right, right, down.",
"Solution_16": "JOIN LEGEND OF THE BLACK DRAGON",
"Solution_17": "Previous moves: down, left\r\nHits wall once\r\n[b]FlyAgaric moves right[/b]",
"Solution_18": "In-thread mod prod. On vacation?",
"Solution_19": "Me moves like this: DLRULDRUURDLRURLDURDLRUDRULLRD\r\nI bet I won't die.",
"Solution_20": "he's not on vacation, he's playing my game.\r\n\r\nBTW, my game is Legend of the Black Dragon.\r\nThere's still one spot! Join!",
"Solution_21": "Please stop advertising your RPG.\r\n\r\n[b]randomguy64 goes right and then down.[/b]\r\n\r\nBefore this post, I successfully did RUURL. I omitted wall-crashes.",
"Solution_22": "[i]pirgraph goes left, then up... Nothing happens...\n\nmath_explorer goes down, ,finds note with \"in\", then goes down into a wall, then he goes right down left right and as he goes right again, a mystic fog covers his body and kills him.\n\npacman2812 goes right, up, up through the door, right, and right into a wall, then down.\n\nFlyAgaric moves right...\n\naback goes down, finds note with \"in\" then left into a wall, right, up into a wall, left, and finally down into another wall dying...\n\nrandomguy64 goes right, then down...[/i]\r\n\r\npiegraph, please STOP advertising otherwise I will have you banned...\r\n\r\nlol",
"Solution_23": "[quote=\"andrewjjiang97\"]piegraph, please advertising otherwise I will have you banned...[/quote]\r\nWhaaaaaaaaaaat? :| \r\n\r\n[b](Respawned again, :( ) math_explorer moves down, right, right, right, left, down, left, down, left, down, right.[/b]",
"Solution_24": "pacman goes up, left, up, up, right, right, down.\r\n*fails*",
"Solution_25": "sorry for being a jerk",
"Solution_26": "left \r\nup\r\nright",
"Solution_27": "down up down right left right down left up right left down left",
"Solution_28": "[b]randomguy goes up twice[/b]\r\n\r\n(I did RUURLRD successfully)",
"Solution_29": "RLURU successful\r\n\r\n[b]RLLRUD[/b]"
}
{
"Tag": [],
"Problem": "The topic basically says it all, are you scene?",
"Solution_1": "scene is a noun?",
"Solution_2": "Scene is an adjective.",
"Solution_3": "uh. ok. definition = ?",
"Solution_4": "defenition??????",
"Solution_5": "[url=http://www.urbandictionary.com/define.php?term=scene]www.urbandictionary.com/define.php?term=scene[/url]",
"Solution_6": "um....i guess i'm not scence...",
"Solution_7": "UHH....No! I've never heard of that definition of scene be4.",
"Solution_8": "no...I learned a new phrase though.",
"Solution_9": "*bursts into laughter* i think AoPS and the scenesters are like, polar opposites by definition.",
"Solution_10": "[quote=\"h_s_potter2002\"]The topic basically says it all, are you scene?[/quote]\r\n\r\nShravas...whats the point of this? Were you trying to find out how many people knew what scene meant or something?",
"Solution_11": "I think I speak for many people when say: What the heck!?",
"Solution_12": "Yeah scene kids are pretty gay by definition, lots of them at my HS.",
"Solution_13": "[quote=\"bleumoose\"]*bursts into laughter* i think AoPS and the scenesters are like, polar opposites by definition.[/quote]\r\n\r\nThat's why I started it.",
"Solution_14": "for the most part scene/emo people annoy me.\r\n\r\nhaha, the moment I saw the topic I knew everyone would be like \"WTH??\"",
"Solution_15": "[quote=\"WarpedKlown1335\"]Yeah scene kids are pretty gay by definition, lots of them at my HS.[/quote]\r\n\r\nI'm going to \"call you out\" on this. It may not seem like a big deal to you, or maybe not to anyone here, but I take very personal offense when someone uses \"gay\" as a derogatory term. I have some extremely intelligent, beautiful, and talented homosexual friends, and I love them dearly. So please, think before you say something like that.",
"Solution_16": "yeah...well about i would say 70% of ppl at my high school are like that..\r\n\r\ni don't like stoners, and there is a whole bunch at my school",
"Solution_17": "[quote=\"bleumoose\"]I have... homosexual friends.... [b]I love them dearly[/b].[/quote]\r\nWas this intentional?It seems like a funny blunder, but yes, gay people are fine.\r\nExecpt for the annoying ones who think they should get special treatment because they're gay...but that is another story.",
"Solution_18": "Musta missed something. I know 'about' these kids.. never met one. No self identity.",
"Solution_19": "no. :huh: \r\nnever met any of these kinda people anyway..."
}
{
"Tag": [],
"Problem": "Hey MA people,\r\nMAML I is in about a week, how are YOU prepping for it?",
"Solution_1": "Wow its already in about a week? I need to sign up asap then..\r\n\r\nWhen is the exact date?",
"Solution_2": "I believe it's this Thursday."
}
{
"Tag": [
"analytic geometry"
],
"Problem": "A set $ S$ of points in the $ xy$-plane is symmetric about the origin, both coordinate axes, and the line $ y \\equal{} x$. If $ (2, 3)$ is in $ S$, what is the smallest number of points in $ S$?\n\n$ \\textbf{(A)}\\ 1 \\qquad\n\\textbf{(B)}\\ 2 \\qquad\n\\textbf{(C)}\\ 4 \\qquad\n\\textbf{(D)}\\ 8 \\qquad\n\\textbf{(E)}\\ 16$",
"Solution_1": "[hide=\"Solution\"]\nBecause the point (2,3) is not on the origin, an axis, or the line y=x, we know that there will be no weird situations. If (2,3) is reflected across the origin, we have 2 points. When those points are reflected across the axes we have 4 points. When [i]those[/i] points are reflected across the line y=x, we have o total of [u][b]8[/b][/u] points in S. [b]D[/b][/hide]"
}
{
"Tag": [
"geometry",
"AMC",
"USA(J)MO",
"USAMO"
],
"Problem": "Saw this on a part of the site (About Us\\Contact Us):\r\n\r\n[quote]If you would like to report an incident of copyright infringement by a member of our Community on our Forum, please contact Richard Rusczyk via one of the methods above.[/quote]\r\n\r\nJust wondering, but what does this mean? And what does this mean for us?\r\n\r\nDoes \"copyright infringement\" mean posting questions from copyrighted sources, like the AMC 10/12 exams? Or does it have to do with posting large quantities of questions? Because in that case, the \"Resources\\Websites\" section might be dangerous, both for AoPS Incorporated and for me, mathfanatic.\r\n\r\nAny input from anyone would be appreciated.",
"Solution_1": "Why would the resources/websites list be dangerous for you? If anything on those sites violates copyright, then that's the owner of that site's problem, not yours (or ours at AoPS).\r\n\r\nWhat it means is a legal grey area in some regards, but I think pretty clear-cut in others. Throwing out problems from various sources for conversational purposes on the board is fine. If we (meaning AoPS) gathered up, say, a bunch of Olympiad problems and started selling them on the site, that wouldn't be fine. I'm not comfortable reproducing them en masse, either (i.e. I'm happy linking to the Scholes site, but I wouldn't host such a site without a clear ok from either a lawyer or the folks whose tests I was reproducing). \r\n\r\nFurthermore, for educational purposes, the copyright laws are decidedly less strict than they are for commercial purposes. I can use USAMO problems in the weekend classes we hold here in San Diego (and do quite often) to my heart's content. I can't box 'em up and sell 'em to the students as USAMO collections. \r\n\r\nAlso, I wouldn't worry too much about the AMC coming after you for citing a problem here or there (particularly if you note where it came from). I would want you to discuss my problems if I produced them. I wouldn't want you releasing every single one to everyone in an easy to get at format if I were selling them, but I'd be very happy to have you throwing out problems occasionally (moreover, unless you were throwing them out and claiming you wrote them, I think I'd be powerless to stop you unless you did it on a scale that was damaging to me).\r\n\r\nAs for why that information about copyright infringement is on the site at all, it's a result of the Digital Millenium Act, which requires that information be present. There's an ongoing war over intellectual property (see the recent Disney laws for one of the more absurd recent government decisions). The internet intensifies it, as it's easier than ever to reproduce and distribute information.\r\n\r\nHope that answers your questions.",
"Solution_2": "Okay, thanks for clearing that up. I just didn't like the sound of the sentence:\r\n[quote]...an incident of copyright infringement by a member of our Community on our Forum...[/quote]\r\n\r\nWhich now makes perfect sense legally: it identifies that if there is a copyright infringement made [i]by a member of the community[/i] (for instance - although I don't think so - me), it is made by a member of the community, and not AoPS Inc. That is, if I started posting advertisements that I was selling collections of problems, then it ought to be me that gets in trouble, not you.\r\n\r\nNot that I think anyone would try to do that because 1) there are enough problems out there for free that 2) it would be too much trouble for 3) not enough profit. And 4) because we are \"mathletes\" with morals and selling someone else's things on someone else's site would be completely wrong.\r\n\r\nSo in summary, no problems. That disclaimer is, as I understand, just there for legal purposes?",
"Solution_3": "You got it.",
"Solution_4": "'k.\r\n_____________________\r\n\r\n\"Boy howdy, all this legal stuff is making me hungry...for math!\" \r\n\r\n=> mathFANATIC"
}
{
"Tag": [],
"Problem": "It's christmas, so as a gift, I will attempt to write my first problem. :) \r\n\r\n$4xy({(x+y)}^2-2xy)=a$\r\n\r\n$6\\left({\\frac{a}{4(x^2+y^2)}}\\right)^2=b$\r\n\r\n$(x^2+y^2)(x^2+y^2)-\\frac{2b}{6}=c$\r\n\r\nFind $a+b+c$ given that $x=\\frac{3}{17}$ and $y=\\frac{31}{17}$\r\n\r\nWhat would $a+b+c$ be in terms of x and y?\r\n\r\nDon't try to brute force this, the solution is simpler than it looks.",
"Solution_1": "This is a very simple question.\r\n\r\nAll you have to do is plugging in the values.\r\n\r\nThis is quite nice for Christmas! :D",
"Solution_2": "Actually, I didn't intend for you to plug in x and y from the beginning, I changed their values to make it harder to do that. The answer remains the same though.",
"Solution_3": "[hide]\n$a = 4xy(x^2+y^2)\\;\\;\\Rightarrow\\;\\;a^2 = 16x^2y^2(x^2+y^2)^2$\n\nThen: $b \\;= \\;\\frac{3a^2}{(8(x^2+y^2)^2} \\;= \\;\\frac{3\\cdot16x^2y^2(x^2+y^2)^2}{8(x^2+y^2)^2} \\;= \\;6x^2y^2 $\n\nAnd: $c \\;= \\;(x^2+y^2)^2 - \\frac{b}{3} \\;= \\;(x^2+y^2)^2 - \\frac{6x^2y^2}{3} \\;= \\;(x^2+y^2)^2 - 2x^2y^2 \\;= \\;x^4+y^4$\n\n\nHence: $a + b + c \\;= \\;4xy(x^2+y^2) + 6x^2y^2 + x^4+y^4$\n\n$.\\qquad\\;\\;a + b + c \\;= \\;x^4 + 4x^3y + 6x^2y^2 + 4xy^3 + y^4$ \n\nTherefore: $a + b + c \\;= \\;(x + y)^4$\n\n\nFor $x = \\frac{3}{17},\\;y = \\frac{31}{17}:\\;\\;a + b + c \\;= \\;\\left(\\frac{3}{17}+\\frac{31}{17}\\right)^4 \\;= \\;2^4 \\;= \\;16$[/hide]",
"Solution_4": "we can imply the result might be (x+y) or something related, because the sum of the given numbers is 3/17+31/17=2"
}
{
"Tag": [
"MATHCOUNTS"
],
"Problem": "This is the problem:\r\n\r\n\r\nTwo trains are headed directly towards each other at rates of 8o miles per hour and 120 miles per hours. How many miles apart are they 1 minute before impact? Express your answer as a mixed number.",
"Solution_1": "[hide]Ok, so they are effectively rushing towards each other at 200 miles per hour. Thus, 1 minute before impact, they will be $ \\frac{200}{60}=3 \\frac{1/3}$ miles apart. [/hide]",
"Solution_2": "[quote=\"k00lperson\"]This is the problem:\n\n\nTwo trains are headed directly towards each other at rates of 8o miles per hour and 120 miles per hours. How many miles apart are they 1 minute before impact? Express your answer as a mixed number.[/quote]\r\n\r\nOk....so they are heading towards each other at the rate of 80 + 120 = 200 mph. At 60 minutes they are 200 miles, so at 1 minute they will be 200/60= 10/3= 3 1/3 miles apart. Hope that helps!!",
"Solution_3": "Yes, it does help."
}
{
"Tag": [
"probability",
"trigonometry",
"floor function",
"geometry",
"search",
"greatest common divisor"
],
"Problem": "Please read each part of the hidden text.\r\n\r\n[hide=\"Round 1\"]\nWelcome to the COP the 4th Round 1. I welcome you and here are few things you have to know.\n\n[hide=\"rules\"]\nPlease note that:\n\n1. You must send your answers (not your solutions. If you want to, you can, however)\n2. You have a week to do these 5 problems.\n3. To go to Round 2, you must get all five problems right.\n4. Check your answer carefully so you get 4 and miss one.[/hide]\n\n[hide=\"Problems\"]\n1. Laurence is choosing one card from a standard deck of 52 playing cards. Let probability of choosing face or spade card is $\\frac{n}{m}$ with $(m,n) = 1$. Compute $mn$.\n\n2. Let $y_1 = \\frac{1}{1 + \\cos x}$ and $y_2$ is another way to write $y_1$ without using cosine. Find $y_2$.\n\n3. Let $f(x) = \\lfloor x \\rfloor$. Find $S$ if:\n\n$S = f(\\frac{1}{3}) + f(\\frac{2}{3}) + f(\\frac{4}{3}) + f(\\frac{7}{3}) + f(\\frac{11}{3}) + f(\\frac{16}{3}) +... f(\\frac{211}{3})$\n\n4. Phil walks from the center of the circle $(x-5)^2+(y+5)^2 = 100$. From that point, he walks seven units up and nine units east. Call this point $z_2$ and the original point he started $z_1$. The midpoint of segment connecting $z_1$ and $z_2$ is $z_3$. Write the equation of the circle with radius 17 that has $z_3$ as its center. Express the equation with the fractions.\n\n5. Given that two points, $c_1$ and $c_2$, are on the circumeference of the circle $O$ that arc ${c_1 O c_2} = 60^\\circ$. If $\\overline{c_1 O} = \\overline {c_2 O} = \\overline {c_1 c_2} = 9$, find the difference between the area of the sector $c_1 O c_2$ and the area of $\\Delta {c_1 O c_2}$. Express your answer in fraction with $\\pi$.[/hide]\n\nPDF version of this test is also available at this post.[/hide]",
"Solution_1": "We send them to you by PM, right? And on the fourth rule, I assume you mean so you [b]don't[/b] get 4 and miss one. :P",
"Solution_2": "PM is correct.\r\n\r\nThanks for pointing that out.",
"Solution_3": "for #2, there could be multiple ways too answer.... and do you want it simplified or not?",
"Solution_4": "True.. That is only problem.\r\n\r\nSo, for #2, send me the steps you did to get the answer.\r\n\r\nThen I'll later post all the answers of #2. :)",
"Solution_5": "In what form do you want #4 expressed?\r\nWith the specifications given, it could be anything with fractions.",
"Solution_6": "For question 1, are face cards just Jacks, Queens or Kings, or do they also include Aces?",
"Solution_7": "#1: It's a regular 52 playing cards.\r\n\r\n#4:\r\n\r\nIn form of $(x - h)^2 + (y - k)^2 = r^2$",
"Solution_8": "Hmm.. my question better phrased:\r\n\r\nAre aces included as face cards?",
"Solution_9": "No.",
"Solution_10": "Do you mean angle instead of arc on problem 5?",
"Solution_11": "It is arc.",
"Solution_12": "I'm not used to math in english, so excuse me, but I would like a few things clarified:\r\n\r\n#1: (m,n)=1 means that m and n are mutually prime, right. Or generally, (a,b)=c means that the greatest common divisor of a and b is c?\r\n\r\n#2: I guess sine is not allowed either? :)\r\n\r\n#5: \"Arc $c_1Oc_2$\" means the curve you get if you trace along the circle from $c_1$ to $c_2$, correct? And what is with the lines? Does it mean the distance between these two points? And in that case, which point is O? Actually, come to think of it, can't you just generally clarify this one? :)",
"Solution_13": "[quote=\"TheBB\"]I'm not used to math in english, so excuse me, but I would like a few things clarified:\n\n#1: (m,n)=1 means that m and n are mutually prime, right. Or generally, (a,b)=c means that the greatest common divisor of a and b is c?\n\n#2: I guess sine is not allowed either? :)\n\n#5: \"Arc $c_1Oc_2$\" means the curve you get if you trace along the circle from $c_1$ to $c_2$, correct? And what is with the lines? Does it mean the distance between these two points? And in that case, which point is O? Actually, come to think of it, can't you just generally clarify this one? :)[/quote]\r\n\r\nNumber one: Yes\r\nNumber two: There is no restriction on using sine. :) \r\nNumber five: Curve, yes. The distance between the two points (straight line). Which point is O? Circle P has center P. :)",
"Solution_14": "Can we enter this if we haven't participated in the previous rounds?",
"Solution_15": "This is the first round so you may enter.\r\n\r\nAll you have to do is solve these five problems and send the answers by PM.",
"Solution_16": "Ok, thanks. Now let's see... =)",
"Solution_17": "Sorry if this is really obvious to you guys, but...\r\n\r\nOn #3, what are the absolute-value (| |) like symbols around $x$?",
"Solution_18": "Greatest integer function.\r\n\r\n$\\lfloor \\frac{3}{4} \\rfloor$ is equal to 0 while $\\lfloor \\frac{5}{4} \\rfloor$ is equal to 1.\r\n\r\nThis basically takes the greatest integer of $x$.\r\n\r\nFor more info, look here:\r\n\r\n\r\n [url=http://mathworld.wolfram.com/FloorFunction.html]From Mathworld[/url]",
"Solution_19": "at #2 $tan$ is alowed?",
"Solution_20": "My english is not so good and I have an \"unknown-word\" question. At #1 I understood tha face cards=Queens, Kings & Jacks, but Spade cards means?! ... (in romanian :D )",
"Solution_21": "From what I know, any expression other than cosine is allowed in your answer.\r\n\r\nDo you know which ones clubs are? Spades are the other black cards. The have the point at the top, and if you put your thumb and index finger of each hand together (index finger touching index finger, thumb touching thumb), it looks kind of like a spade... I hope that last part didn't confuse you.",
"Solution_22": "Go to this site (from Google search engine)\r\n\r\nhttp://images.google.com/images?q=Spade&hl=en",
"Solution_23": "Spades = Small shovels.",
"Solution_24": "[quote=\"tarquin\"]Spades = Small shovels.[/quote]\r\n\r\nWhat? Spades = big shovels",
"Solution_25": ":) thanx. And $\\lfloor 8,9 \\rfloor=8$?",
"Solution_26": "IF that's supposed to be fraction then no it's not equal ti 8.\r\n\r\n$\\left\\lfloor \\frac 89 \\right\\rfloor = 0$",
"Solution_27": "Hm, he might have meant the comma to be a decimal point, which in that case it would be $8$.",
"Solution_28": "Yes. The comma was a decimal point. Thanx :P",
"Solution_29": "if x is an integer and\r\nif x <= n < x+1\r\nthen [n]=x",
"Solution_30": "When is this due?",
"Solution_31": "Can we use $\\sin$ in #2?",
"Solution_32": "What is a sector? Is the area between a arc and a chord?",
"Solution_33": "Answers must be turned by this Friday. 5/6/05.\r\n\r\nPlease send them in! :)",
"Solution_34": "Yes, sin can be used in #2. A sector is the area inside a circle bounded by two radii.",
"Solution_35": "Where can we discuss the solutions and when?",
"Solution_36": "Solution in PDF Format will be posted in tomorrow afternoon 1:00 EST.\r\n\r\nPlease do not talk about the solutions until I have posted PDF solution document.\r\n\r\nThanks and Regards,\r\n\r\nSilverfalcon",
"Solution_37": "All right. And we discuss the problems in this thread?",
"Solution_38": "I'll post the separate thread for solutions.\r\n\r\nIt's easier to have two different threads for problems and solutions and that's how I did on the past ones.",
"Solution_39": "Thats great. Thank Silverfalcon for all the efforts made to this competition. :)"
}
{
"Tag": [
"Euler",
"limit",
"logarithms",
"inequalities proposed",
"inequalities"
],
"Problem": "Let $\\gamma=$ Euler's constant,i.e.\r\n$ \\gamma:=\\lim\\limits_{n\\to \\infty}\\left(1+\\frac{1}{2}+\\frac{1}{3}+...+\\frac{1}{n} - \\ln{n}\\right) .$\r\n[b] (Without using a PC), prove or disprove that [/b] $ \\pi^2 < 8e^{\\gamma} . $",
"Solution_1": "I wonder if we can prove this using the fact that :\r\n$\\gamma =\\lim\\limits_{n\\to \\infty}\\left(1+\\frac{1}{2}+\\frac{1}{3}+...+\\frac{1}{n} - \\ln{n}\\right) .$\r\nand\r\n$\\frac{\\pi^2}6 = \\lim_{n \\to \\infty} \\left(1 + \\frac 1{2^2} + \\frac 1{3^2} + \\ldots + \\frac 1{n^2}\\right) .$\r\n\r\n :?:"
}
{
"Tag": [
"inequalities",
"inequalities unsolved"
],
"Problem": "Let $ \\operatorname{frac}(x): \\equal{}x\\minus{}\\operatorname{floor}(x)$ where $ \\operatorname{floor}(x)$ is the largest integer $ \\leq x$. Prove that $ \\operatorname{frac}((3/2)^n)<1\\minus{}(3/4)^n$ as $ n>1$.",
"Solution_1": "Maybe you mean $ frac(x)\\equal{}\\{x\\}$?"
}
{
"Tag": [
"Putnam"
],
"Problem": "Im down in san diego and one of my teachers up north wants problems to present for a putnam masterclass. I would like to present the first paul erdos problem in this book, so if anyone could tell me what it was id be so grateful. Thanks a bunch,\r\nJ.D.",
"Solution_1": "The first problem from Paul Erdos is as follows:\r\n\r\nProve that every infinite sequence of distinct positive integers contains either\r\n(a)an infinite subsequence such that for every pair of terms, neither term ever divides the other,or\r\n\r\n(b)an infinite subsequence such that in every pair of terms one term always divides the other."
}
{
"Tag": [
"function",
"calculus",
"integration",
"complex analysis",
"complex analysis solved"
],
"Problem": "let f be an analytic function in the once_punctured disc G={ z: 0<|z-zo|0 \r\nthen what can u say about the nature of the singularity at zo?\r\n\r\nappreciate ur help",
"Solution_1": "[url=http://www.mathlinks.ro/viewtopic.php?t=162883]here[/url]"
}
{
"Tag": [
"analytic geometry",
"graphing lines",
"slope"
],
"Problem": "The points $ (x, y)$ represented in this table lie on a straight line. When the equation of this line is written in the form $ y \\equal{} Ax \\plus{} B$, what is the value of $ A \\plus{} B$?\\\\\n\n\\begin{tabular}{c|c}\n $ x$ & $ y$ \\\\ \\hline\n $ 2$ & $ 7$ \\\\\n $ t\\minus{}2$ & $ v$ \\\\\n $ t$ & $ v\\plus{}6$\n\\end{tabular}",
"Solution_1": "The points (x,y) represented in this table lie on a straight line. When the equation of this line is written in the form y=Ax+B, what is the value of A+B?\r\n\r\n$ \\\\\r\n\\begin{tabular}{c|c} \r\nx & y \\\\ \\hline \r\n2 & 7 \\\\ \r\nt\\minus{}2 & v \\\\ \r\nt & v\\plus{}6 \r\n\\end{tabular}$\r\n\r\n...I LaTeXified it.\r\n\r\nslope : $ \\frac{v\\plus{}6\\minus{}v}{t\\minus{}(t\\minus{}2)}\\equal{}3$\r\n\r\n(x,y) = (2,7).\r\n\r\nso y=3x+1.\r\n\r\nA=3, B=1.\r\n\r\nA+B=4.\r\n\r\nanswer : 4",
"Solution_2": "Instead of using the slope formula, we see that from $t-6$ to $t$ (an increase in $x$ by 2), we have $v \\Rightarrow v+6$ (an increase in $y$ by 6). Slope, essentially, is the change in $y$ over the change in $x$, so we have the slope as $\\dfrac{6}{2}=3$."
}
{
"Tag": [
"number theory proposed",
"number theory"
],
"Problem": "(x+y) = a (mod c)\r\n(x*y) = b (mod c)\r\n\r\nspecifically:\r\n\r\n(x+y) = 17 (mod 35)\r\n(x*y) = 2 (mod 35)",
"Solution_1": "Nice source. ;) :( \r\n\r\nPierre.",
"Solution_2": "[quote=\"integer\"](x+y) = a (mod c)\n(x*y) = b (mod c)\n\nspecifically:\n\n(x+y) = 17 (mod 35)\n(x*y) = 2 (mod 35)[/quote]\r\n\r\nyes. Chinese remainder theorem says that you can assemble the solutions (if any exist) from solutions modulo 5 and 7. There can be up to 2 solutions in each case, so one expects (at most) 2x2=4 solutions in total."
}
{
"Tag": [
"group theory",
"abstract algebra",
"LaTeX",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "Let N_1 and N_2 be normal subgroups of a group G\r\n\r\nis the following true? \r\n(N_1N_2)/(N_1 intersection N_2) = N_1/(N_1 intersection N_2) x N_2/(N_1 intersection N_2) where / denotes quotient group, x denotes direct product and intersection as in setwise sense.",
"Solution_1": "Does this make it simpler?\r\n\r\n$N_{1}N_{2}/(N_{1}\\cap N_{2}) \\simeq N_{1}N_{2}/N_{1}\\times N_{1}N_{2}/N_{2}$",
"Solution_2": "are you saying that is true? (I believe you if you say so...)\r\nwhat about my expression?",
"Solution_3": "Actually ignore what I said, I was thinking about some stupid approach. Just define an isomorphism between them.\r\n\r\n$\\phi : N_{1}N_{2}/(N_{1}\\cap N_{2}) \\mapsto N_{1}/(N_{1}\\cap N_{2}) \\times N_{2}/(N_{1}\\cap N_{2})$\r\n\r\nAs,\r\n\r\n$\\phi (n_{1}n_{2}(N_{1}\\cap N_{2})) = (n_{1}(N_{1}\\cap N_{2}), n_{2}(N_{1}\\cap N_{2}))$\r\n\r\nI am sure this maps is well-defined, not in the mood to show that. And it seems like this is a homormophism. So it seems like it is true.\r\n\r\nIn fact, we never need the fact that $H_{1}\\cap H_{2}$. Rather we can say the following, a stronger statement.\r\n\r\n\"If $N$ is a normal subgroup of both subgroups $H_{1}$ and $H_{2}$ then,\r\n$H_{1}H_{2}/N \\simeq H_{1}/N \\times H_{2}/N$\"",
"Solution_4": "thanks for that info. \r\n\r\nhow do you put math symbols as you did? are the codes same as that in latex? and do i just type it in? \r\n\r\n\r\nlet me try:\r\n\r\n$a+b$",
"Solution_5": "[quote=\"sit.think.solve\"]thanks for that info. \n\nhow do you put math symbols as you did? are the codes same as that in latex? and do i just type it in? \n\n\nlet me try:\n\n$a+b$[/quote]\r\n\r\nYes same as LaTeX just use \"dollar symbol\" to start and end."
}
{
"Tag": [],
"Problem": "There are 6 differently coloured balls placed in a row. A new arrangement cannot be made if two balls that were together in the last arrangement, are together again. How many new arrangements can be made?",
"Solution_1": "[hide]only the numbers like 123456 and 135246 are possible. we get trapped on any other number after this. [/hide]"
}
{
"Tag": [],
"Problem": "If $a,b,$ and $c$ are digits for which \\[ \\begin{tabular}{cccc} & 7 & a & 2 \\\\ - & 4 & 8 & b\\\\ \\hline & c & 7 & 3 \\end{tabular} \\] then $a+b+c =$\n\n$\\text{(A)} \\ 14 \\qquad \\text{(B)} \\ 15 \\qquad \\text{(C)} \\ 16 \\qquad \\text{(D)} \\ 17 \\qquad \\text{(E)} \\ 18$",
"Solution_1": "[hide]The answer should be D) 17 because the only number fit for c is 9 and the only number fit for b is 6 and a is left with 2.[/hide]",
"Solution_2": "[hide]\n12-b=3, b=9\n(10+(a-1))-8=7, a=6\n(7-1)-4=2\n\na+b+c=15\n\n(B)\n\n[/hide]",
"Solution_3": "[hide]This problem is easier to do one column at a time. For the ones column, 3 is greater than 2, so we know that 2 must have borrowed 10 from a, leaving 12-b=3. [b]b=9[/b]. a is now 1 digit less than before, so whatever digit we get for a we must add one to. In the tens column, 7+8=15, so we find that a's current digit is 5 and it borrowed from the seven to its left. Adding 1 to 5 leaves [b]a=6[/b]. In the hundreds column, 7 became 6 after a borrowed from it. 6-4=[b]c=2[/b]. a+b+c=9+6+2=[b]17[/b].[/hide]",
"Solution_4": "The answer is c 16. I think",
"Solution_5": "Oh no I messed up :( It supposed to be d) 17",
"Solution_6": "[quote=\"akimatsu\"][hide]\n12-b=3, b=9\n(10+(a-1))-8=7, a=6\n(7-1)-4=2\n\na+b+c=15\n\n(B)\n\n[/hide][/quote]\r\n\r\nyou screwed up on your last steps.",
"Solution_7": "a=6 b=9 c=2 so 6+9+2=17(D)",
"Solution_8": "About half of these people really don't know how to add. 9+6+2=17 not 15 people. the answer is 17",
"Solution_9": "[hide=\"check ur additions, guys\"]\nusing common mathematical sense, you're gonna realize that $a+b+c=17$, which is $D$.[/hide]",
"Solution_10": "[quote=\"dogseatcheese\"][hide]The answer should be D) 17 because the only number fit for c is 9 and the only number fit for b is 6 and a is left with 2.[/hide][/quote]\r\n\r\no and by the way, ur answer is correct, but the numbers correspondent to the variables are wrong :huh:",
"Solution_11": "[hide]\nObviously b=9 because 12 - 9 = 3 and then we get 7A2 - 489 = C73 so a=6 and c=2. a+b+c=17. [b]D[/b]\n[/hide]"
}
{
"Tag": [
"probability",
"complementary counting"
],
"Problem": "There are 20 kids in your class and each is randomly assigned a country to report on by random picking from a hat. It goes like this: the teacher puts the names of 20 countries into a hat and goes to each student one-by-one to randomly choose a country. Once a country is chosen, it cannot be chosen again. You really want to pick United States because that means no extra research is required :D\r\n\r\nThe question is, is it better to go first, last, or somewhere in the middle?",
"Solution_1": "I kind of did like the first three, and you can see that all of them has a chance of $ \\frac{1}{20}$ But, I prefer to go third. :D",
"Solution_2": "How will this change if you have more than one country you want to report on? How about if you have a country that you [b]definitely[/b] do not want to report on?",
"Solution_3": "First one uh depends if they have the country I want. :P Jk, lets see, experiment again but this time there is casework.(Use complementary counting) you find that the probability that you don't pick it is 9/10, so 1-9/10=1/10.\r\nSecond one same thing as first one.",
"Solution_4": "should be the same.... but... i prefer middle."
}
{
"Tag": [
"percent"
],
"Problem": "I'm having trouble getting past level 11. I put tons of stone towers in the earlier levels (well like 3) because otherwise I can't get past the land creeps but then on level 11 I don't have enough to buy air towers to kill the spaceships. I think the only way is to get several Stick towers in the beginning and sell them off later when you have better towers...",
"Solution_1": "umm you didnt post a link...\r\nSILLY ALBERT TRIX ARE FOR KIDS",
"Solution_2": "Yeah oops... It's called Shock Defense.\r\n\r\nhttp://www.addictinggames.com/shockdefence.html",
"Solution_3": "hmm, pretty good, but not very hard\r\nplayed it a few times, beat it the second time i actually knew what was going on\r\ni only had 1 life left though :D \r\ni had kinda messed up earlier\r\nmy score was 2629\r\ni can describe my strategy if you want",
"Solution_4": "Yeah I just suck at stuff like this... describe your strategy please.\r\n\r\nSo I put 3 stone towers in the beginning and upgrade them ASAP for level 6 (the fast legions). I also put a few stick towers cuz they attack land + air and I need them for level 1-10 and then for level 11. But I don't have enough money for more anti-air units...",
"Solution_5": "here's how i did it:\r\nfirst i started with about 5 upgraded stones in the middle of the loopy section. i then got a few lightning towers in that same place\r\ni sold a few of the stones when the flyings started coming, but kept 2 (this is where i lost some life, i dont think i kept enough stones)\r\nagainst the flyings i used flying towers and lightning towers upgraded twice (the third was too expensive, dont know how good it is)\r\nthen, i put a poison in the left loop\r\neventually, i added some fires to the mix, though they werent a big part of my evil plan\r\nonce i started getting to the 30s i got an ultimate tower\r\nby the end of the last level i had two ultimate towers, one upgraded (those things are awesome)\r\ni also put a few more poisons around the place to slow the people down in the final levels (cuz the ultimates have a huge range)\r\n\r\noh yeah there are 35 levels total",
"Solution_6": "I just beat level 33 as of right now... and I have 13 lives left.\r\n\r\nI forgot there were ultimate towers... :blush:\r\n\r\nEDIT: Darnit... I invested 2 of my tributes into +percent so now I can't get ultimates :( . And also, DO NOT get the thunder 3rd upgrade... it increases damage from 360 to 400 :( .",
"Solution_7": "okay, a few more things i've learned\r\n1: DO NOT KEEP YOUR OLD TOWERS\r\n2: YOU MUST UPGRADE THE ULTIMATE TOWER OR YOU WILL NOT WIN",
"Solution_8": "I'm on level 30 right now, and I have an upgraded rocket tower thingy.\r\n\r\nI think they should be banned.\r\n\r\nLet's see how far people get without a Shock Rocket Tower. I got to level 34 :) .\r\n\r\nOh yeah, and for the bonus levels there are 40 enemies each.",
"Solution_9": "Yay. I beat the game.\r\n\r\nScore: 3450.\r\n\r\nGeez, they should give like lives * 100 because lives left are way more important than money or whatever.",
"Solution_10": "Stone then lightning then rocket\r\nQuite easy to beat! :D \r\nScore: 3155",
"Solution_11": "ehhh\r\ni got a 3317\r\n\r\npretty cool that 3317 is an anagram of 1337",
"Solution_12": "I finished the game with 20 HP, 18000 points and 14000 money\r\n\r\neasy strategy\r\n\r\n1. levels 1-7 \r\nbuild 1 stone tower ,upgrade twice, dont spend any more money, use 1st tribute on lightning tower\r\n\r\n2. levels 8-14 kill stone tower, build lightning tower, upgrade once\r\nuse tribute on interest rate, dont spend any more money\r\n\r\n3. levels 15-21 upgrade lightning tower 2nd time, dont spend any money, use tribute on interest rate\r\n\r\n4. levels 22-28 , build 1 more lightning tower, upgrade twice, use tribute on interest rate\r\n\r\n5. levels 29-34 , build adjacent lightning towers so they all shoot at the target at the same time , upgrade twice (not 3 times) \r\n3rd upgrade is a waste of money, no range diff, and damage diff is only 40 points, whilst you spend 720\\$ on it \r\n\r\n6. spend half your money on twice upgraded lightning towers to finish the last level \r\n\r\nimportant note\r\nput towers on corners\r\nso they shoot in circle and reach many places at once",
"Solution_13": "I fail to comprehend how it is possible to beat level 6 with only 1 stone tower.",
"Solution_14": "put another stone tower\r\n\r\nbut!! they must shoot at the same target at the same time",
"Solution_15": "I can only get 16039 by 4 rocket towers\r\nSeems upgrading to rocket tower is not a good strategy",
"Solution_16": "How do you get a lot of points? I only got 4300 or so, even though I had 19 lives left.",
"Solution_17": "your points are the same as your gold\r\ninfluinced by interest\r\n\r\nif you only upgrade between levels\r\n\r\nand after level 7 get rid of the small towers (ALL OF THEM)\r\n\r\nand put only lightning towers\r\n\r\nyou can win \r\n\r\nalso until certain level, you can keep on only 2 towers, also dont upgrade the towers more than twice, its a waste of money",
"Solution_18": "how do u beat the level with the 20000 hp ships",
"Solution_19": "Have lots of lightning towers :) .\r\n\r\nAnd the level after that is 30000 hp ships.",
"Solution_20": "Easy.\r\n\r\nScore: 31356\r\n\r\nI would have higher, but I started selling off my towers at the end too late. I estimate I could have gotten several thousand more if I had started selling them sooner (and, I built a few too many, but whatever).",
"Solution_21": "I beat it. After you get the advanced towere then the things get easy. after you have a super tower, sell all of youre non advanced towers. Use this monny to buy supertowers and upgrade them."
}
{
"Tag": [
"geometry",
"trigonometry",
"trig identities",
"Law of Sines",
"Law of Cosines",
"area of a triangle"
],
"Problem": "$\\triangle{ABC}$. $\\angle{A}=60^{o}$, $AC=16$. The area of the triangle is $220 \\sqrt{3}$. \r\nFind $BC$.",
"Solution_1": "By the Law of Sines, \r\n$[\\triangle ABC] =\\frac{AC \\cdot AB \\cdot \\sin \\angle A}{2} = 220 \\sqrt3$\r\nSo $AB = \\frac{55}{4}$\r\nUsing Law of Cosines, \r\n$BC = \\frac{\\sqrt{3601}}{4}$\r\n\r\nSince the answer seems so weird, I am pretty sure that I did something wrong :?",
"Solution_2": "[hide=\"bornformath\"]\ntry recalculating $AB$\n[/hide]\n\n[hide=\"answer\"]\nBornForMath's method is correct, but calculating gives us $AB = 54$\n\nThen using law of cosines, $BC=2\\sqrt{577}$\n[/hide]",
"Solution_3": "really?\r\n[hide]i got $BC=49$ :D[/hide]",
"Solution_4": "[quote=\"Palytoxin\"]really?\n[hide]i got $BC=49$ :D[/hide][/quote]\r\nBingo! :D",
"Solution_5": "i dunno how, i got ab=55",
"Solution_6": "[quote=\"Priapus\"]i dunno how, i got ab=55[/quote]\r\n[hide]Yep. AB=55, BC=49.[/hide]",
"Solution_7": "[quote=\"deej21\"][hide=\"bornformath\"]\ntry recalculating $AB$\n[/hide]\n\n[hide=\"answer\"]\nBornForMath's method is correct, but calculating gives us $AB = 54$\n\nThen using law of cosines, $BC=2\\sqrt{577}$\n[/hide][/quote]\r\n\r\nThanks for pointing that out. :) But after I redid it, I got $AB = 55$",
"Solution_8": "Weird. I even redid that two or three times to make sure I did it right. For some reason, I didn't catch that $\\frac{20}{4} = 5$\r\n\r\nMeh, I guess that's what I get for doing math late.",
"Solution_9": "sorry, this is a little off topic, but is the area of a triangle $(\\frac{1}{2})b)(c)(\\sin{A})$?\r\nI used this formula and the law of cosines to obtain 49 for BC also, so I assume it's correct, but it could be luck.",
"Solution_10": "I have never expected that this problem will recieve such a result. \r\nIt seems that lots of you misread the problem or made calculation mistakes......\r\nBut after all, it is not a difficult problem and all you need is to be careful."
}
{
"Tag": [
"search",
"projective geometry",
"geometry unsolved",
"geometry"
],
"Problem": "A circle meets three sides BC, CA, AB of triangle ABC at points D_1,D_2;E_1,E_2 and F_1, F_2 in turn.Then, line segments D_1E_1 and D_2F_2 intersect at point L, line segments E_1F_1 and E_2D_2 intersect at point M, line segments F_1D_1 and F_2E_2 intersect at point N.Prove that three lines AL, BM, CN are concurrent.",
"Solution_1": "This is an old ( > 50 years ) well known problem, but I don't remember the proof. I must search my archive about it and I will post here a solution, if will be not any reply with references.\r\n\r\nBy the way, I would like to present another interesting and no so well known ( I think ) problem in this configuration.\r\n\r\n[b][size=100][color=DarkBlue]We denote as $A',$ the intersection point of $BE_{2},$ $CF_{1}$ and as $A'',$ the intersection point of $BE_{1},$ $CF_{2}.$\n\nSimilarly we define the points $B',$ $B''$ and $C',$ $C''.$ Prove that:\n\n1). - The segment lines $AA',$ $BB',$ $CC',$ are concurrent at one point, so be it, $P.$ \n\n2). - The segment lines $AA'',$ $BB'',$ $CC'',$ are concurrent at one point, so be it, $Q.$ \n\n3). - The segment lines $A'A'',$ $B'B'',$ $C'C'',$ are concurrent at one point, so be it, $R.$ \n\n4). - The points $P,$ $Q,$ $R,$ are collinear.[/color][/size][/b]\r\n\r\n\r\nKostas Vittas.\r\n\r\nPS. In my drawing, the points $D_{1},$ $E_{1},$ $F_{1},$ are nearer the vertices $B,$ $C,$ $A,$ respectively.",
"Solution_2": "[quote=\"AkiM\"]A circle meets three sides BC, CA, AB of triangle ABC at points D_1,D_2;E_1,E_2 and F_1, F_2 in turn.Then, line segments D_1E_1 and D_2F_2 intersect at point L, line segments E_1F_1 and E_2D_2 intersect at point M, line segments F_1D_1 and F_2E_2 intersect at point N.Prove that three lines AL, BM, CN are concurrent.[/quote]\r\n\r\nReally well-know!\r\n\r\nI have a solution with Pascal Theorem and Desargues Theorem, but is basically the same of Darij!\r\n\r\nSee here: http://www.mathlinks.ro/Forum/viewtopic.php?search_id=1926634524&t=24378"
}
{
"Tag": [
"Euler",
"AMC",
"AIME",
"articles",
"Gauss",
"downwithcloning"
],
"Problem": "what do you guys think about cloning?",
"Solution_1": "I think clonning is a wonderful breakthough for science but it needs to be kept in check for now and as experiments only. Like the sheep \"Dolly\". I do not think it should be avaliable for a long time.",
"Solution_2": "Cloning is absolutely vital to making a new breakthrough in science, but this power must be kept in a TIGHT check system. I personally am all for cloning human organs... However, I am opposed to cloning human beings because there are tons of problems that come with that...",
"Solution_3": "Exacly what I meant ragging.",
"Solution_4": "Just to play devil's advocate, what are those \"tons of problems\"?",
"Solution_5": "Think about it. What would happen if someone tryed to clone Einstein. Its completely possible. We have his DNA. What would happen if we did clone him. Think of the problems it would cause. Those are part of the \"tons\" of problems with clonning if there is not tight messures suronding it.",
"Solution_6": "ah...forgot to put what I thought...sorry about that....I think cloning has its good parts and its bad parts. I'm pretty much in the middle. It has the potential to help people, but it could ultimately turn out to be something that has serious consequences....",
"Solution_7": "[quote]Think about it. What would happen if someone tryed to clone Einstein. Its completely possible. We have his DNA. What would happen if we did clone him. Think of the problems it would cause.[/quote]\r\n\r\nYou haven't actually supplied a single problem. As far as I am concerned, lets clone Einstein. The only problem I see is possible fraud due to having the same retinal scan or fingerprint or dna. But Einstein is dead, so thats not the case. I think the reason that human cloning is banned now is that cloning is sucessful in only a small percentage of cases, and in the process of cloning a human many fetuses would die. However, if the sucess rate was 100%, I am all for cloning a human. You mention Einstein. First, I have to say that the environment this clone will live in is so different from that the real Einstein lived in that they probably will mentally be completely different. If not, whats wrong with having another genius?",
"Solution_8": "I don't see any problem if we clone Einstein, but I see a lot of problem if we clone Hitler.",
"Solution_9": "[quote]I don't see any problem if we clone Einstein, but I see a lot of problem if we clone Hitler.[/quote]\r\n\r\nI dont. What if the young Hitler had been born in, say, the US. Do you think that he would have become the same evil person? I dont. It all boils down to a nature vs. nurture debate, but what you need to note is that it would take 20 years before the clone Hitler reached maturity. Worst case scenario, he is the same evil person as the real Hitler, he could be prevented from doing anything due to the scientific attention on him. From a purely scientific point of view, I think cloning Hitler would be quite interesting.",
"Solution_10": "Same here, but a lot of people would protest if you clone Hitler.",
"Solution_11": "I don't think its so much as who you clone, but what the clone can be used for. I doubt cloning Einstein would be much different than cloning Hitler; Einstein was a scientific genius, and Hitler was a military genius. The only true difference was that Hitler was extremely addicted to many narcotics and extremely racist, but if we cloned him, I doubt he would be as bad as we think him to be :p. \r\n\r\nI think the biggest problem with cloning is the whole imposter idea. You can easily create another Koffi or GWB and have both people be genetically identical. Now, in the hands of these crazy terrorists, that would be a bad thing, yes? Cloning is more of a weapon than anything else.\r\n\r\nI'm sure there are other things wrong with cloning too...perhaps clones, although genetically human, would not be treated as humans. And then this whole civil rights issue with clones would be a pain to deal with as well. There's also this whole \"playing God\" issue, but I stand neutral on that.",
"Solution_12": "The entire judicial system could be tricked.... Say I created 13 identical tetrahedr0ns all at once. Then one of them commits a serious crime, how do the police know which tetrahedr0n to get? Everyone of them has the same DNA....\r\n\r\nAm I right?",
"Solution_13": "yes, you're right.\r\n i dont think cloning humans should be allowed unless it's absolutely essential as a test or something; even then it should be very heavily regulated",
"Solution_14": "[quote=\"Ragingg\"]The entire judicial system could be tricked.... Say I created 13 identical tetrahedr0ns all at once. Then one of them commits a serious crime, how do the police know which tetrahedr0n to get? Everyone of them has the same DNA....\n\nAm I right?[/quote]\r\n\r\nBy the time we are able to clone people, we would also have some really really really advanced science that's totally beyond DNA.",
"Solution_15": "While Euler may not have been doing olympiad problems, I am sure he could have the potential to easily do them.",
"Solution_16": "Genes have to have something to do with smartness. Everyone is probably born with some degree of math ptential. Whether or not they reach that potential is decided by the environment. \r\nCloning would be very beneficial to scientific experiments. For example, if scientists wanted to test a new drug theyve created, they could test it on clones, which allows the scientists to reduce the amount of variables they have to be concerned with.",
"Solution_17": "The genes definitely help, but they aren't the only thing that is important. My father has a degree in mathematics, and I don't remember him ever forcing me to do math or even encouraging it very much (or discouraging it either). It may have helped save me from the fate of many of my friends, who are forced to study engineering or something boring like that rather than math.\r\n\r\nOn the other hand, I don't think Gauss's family contained any mathematicians at all, and still he was undoubtably one of the greatest mathematicians ever. Apparently his father couldn't even add properly (he had Gauss do his arithmetic for him when Gauss was maybe 4 years old, or so the story goes). His mother apparently didn't believe that he was that exceptional at math until one of his friends or teachers (I forget which) told her that he would be the greatest mathematician in Europe when he was older.\r\n\r\nI should add that the Bernoulli family wasn't exactly known for being an especially close-knit one. Jakob and Johann (who were brothers) fought bitterly and eventually were not on speaking terms. Johann kicked his son out of the house when Daniel was co-awarded a prize with Johann. So I don't know how much the members of that family wanted to imitate each other.",
"Solution_18": "What if some scientists found DNA of some religious person in the Bible, like Moses or Jesus? What would people think about cloning them? I dont know if this would offend anyone, i do on go to church regularly.",
"Solution_19": "[quote=SirErnest]What if some scientists found DNA of some religious person in the Bible, like Moses or Jesus? What would people think about cloning them? I dont know if this would offend anyone, i do on go to church regularly.[/quote]\n\nIf we could clone Jesus back from the dead, that would be amazing. However, there are ethical issues with this. ",
"Solution_20": "Unless Jesus was thrown into a peat bog or glacier, they're not going to get any DNA from something two thousand years old. ",
"Solution_21": "Are we bumping an 18 year old thread?!",
"Solution_22": "I personally believe that a clone of one's self is inferior until its value surpasses the original donor. ",
"Solution_23": "[quote=Robomania_534]I personally believe that a clone of one's self is inferior until its value surpasses the original donor.[/quote]\n\nwell how exactly do you define value?",
"Solution_24": "[quote=Clemmie]Are we bumping an 18 year old thread?![/quote]\n\nsigma penguin\n\nclonign is scary in the case of consciousness",
"Solution_25": "[quote=pezzeepenguinstalker][quote=SirErnest]What if some scientists found DNA of some religious person in the Bible, like Moses or Jesus? What would people think about cloning them? I dont know if this would offend anyone, i do on go to church regularly.[/quote]\n\nIf we could clone Jesus back from the dead, that would be amazing. However, there are ethical issues with this.[/quote]\n\nNot really, because jesus would just be a regular guy (probably) seeing as they are technically different people with the same dna.",
"Solution_26": "[quote=Ragingg]What if I grew all the tetrahedr0ns to be exactly alike, and besides that, even if they were different what if everyone denied it?\n\nSecondly, I am not quite sure, but I don't think identical twins have the exact same genetical code.... They don't come out exactly alike all the time (not sure).\n\nIf I grew 1,000 Eulers all at once, would that be fair to other people? How are the normal people supposed to get into the AIME or MOSP? What would happen if other countries raised 1,000 Eulers and we competed at the IMO?[/quote]\n\nAgain, these \"tetrahedr0ns\" would not all be born at once. The original is guaranteed to be older, and that difference would likely be noticeable. Furthermore, maybe one of the clones got a scar on its knee when it was ten from a bad fall. Maybe one of them was raised differently and became very distinguishable mentally. The list goes on and on. These clones, even if born the same, will\n1) not be the same age\n2) will develop differently",
"Solution_27": "[quote=callie313][quote=Ragingg]What if I grew all the tetrahedr0ns to be exactly alike, and besides that, even if they were different what if everyone denied it?\n\nSecondly, I am not quite sure, but I don't think identical twins have the exact same genetical code.... They don't come out exactly alike all the time (not sure).\n\nIf I grew 1,000 Eulers all at once, would that be fair to other people? How are the normal people supposed to get into the AIME or MOSP? What would happen if other countries raised 1,000 Eulers and we competed at the IMO?[/quote]\n\nAgain, these \"tetrahedr0ns\" would not all be born at once. The original is guaranteed to be older, and that difference would likely be noticeable. Furthermore, maybe one of the clones got a scar on its knee when it was ten from a bad fall. Maybe one of them was raised differently and became very distinguishable mentally. The list goes on and on. These clones, even if born the same, will\n1) not be the same age\n2) will develop differently[/quote]\n\nvalid points!\n\nbtw long time no see!",
"Solution_28": "Cloning would be amazing!!!\n\nHowever, it should only be done with close security, so that people don't abuse their power.\n\neg. clone a bunch of children to work on a farm.",
"Solution_29": "@2above somehow this is strangely addicting?!\n\n@above, well, if somebody had enough money to buy clones I doubt they would use it for menial labor or work.... but I do think abuse of power would be a definite outcome of cloning"
}
{
"Tag": [
"Harvard",
"college",
"search",
"geometry",
"geometric transformation",
"reflection",
"AMC"
],
"Problem": "I'm just wondering what kind of person gets accepted to Harvard. Like what kind of extra curricular activities do you do, how high are your ACT and SAT scores, how high is your GPA, what are your advance classes, and that type of stuff. Just wondering...\r\n :)",
"Solution_1": "[quote=\"mulan2890\"]I'm just wondering what kind of person gets accepted to Harvard. Like what kind of extra curricular activities do you do, how high are your ACT and SAT scores, how high is your GPA, what are your advance classes, and that type of stuff. Just wondering...\n :)[/quote]\r\n\r\nHello mulan2890,\r\n\r\nFirst than all, someone just discussed this a few days ago, you can look for the conversation with the search tools. \r\n\r\nOn the other hand, you don't have to post three times the exact same post to get someones attention, probably with just one is enough. \r\n\r\nFinally, Harvard in general is (as far as I have seen) is pretty worried about extracurricular activities (I am not there), and most of the people that I know who got in, were first perfect or almost perfect scores in SAT, IB, high GPA (the four people that I know got GPA of around 3.9 or so) and were part of the school goverment or some other clubs.\r\n\r\n[b]Moderators:[/b] Could you please delete the other two posts!\r\n\r\nBest,",
"Solution_2": "Since I got my letter today, I guess I can go ahead and tell you.\r\n\r\nSAT I- 750 - Verbal, 800 - Math\r\nSAT II - \r\n800 - Chem \r\n720 - Physics \r\n800 - Math II\r\n770 - Math I \r\n760 - Writing\r\n780 - Literature\r\n(As far as Harvard Applicants go, I think that scores above 700 and even 750 are pretty standard for everything though).\r\nUS Hist, Chem, and Calc AB exams - 5s\r\nand currently taking US Gov, Phys B, Eng Lit, Euro Hist, and Stat AP.\r\nGPA of 4.0 unweighted, 2nd in a class of 700.\r\n\r\nAs you said though, relatively high numbers are standard at places like Harvard.\r\nMy real hook was in the extracurricular stuff, like\r\n\r\nMusic - \r\nPercussion - 4 years All-state band, 2 years National Honor Band, World Percussion Championship and youngest member of the Cadets of Bergen County Drum and Bugle Corps + 11 Years of Piano (Guild Member)\r\nEagle Scout\r\nMembership in Model UN - (Which is a blast by the way, and not too taxing)\r\nNHS VP, Math Team VP, 7th in National French Contest, State Winner of Reflections Music Composition/photography Award\r\nand Various Academic teams/awards\r\nUSNCO Finalist, USABO semifinalist\r\nAIME Qualifier, placed at various math comps.\r\n\r\nAs you probably can tell, my record in pure [b]mathematics[/b] does not compare with many of the others here (got a 4 on the AIME this year :? ). However, I think I was able to get because of my jack-of-all-trades persona. \r\n\r\nSo I would highly recommend beefing up your extracurricular resume as much as possible (try to find things that you enjoy), and listing [i][b]everything[/b][/i] that could possibly set you apart from the average applicant (which for Harvard University, is probably hardly average). In addition, get an interview and all that jazz. Third, write a crazy essay that is real far out, to catch their attention (I wrote mine on the importance of sanitation, making it sound like war propaganda).\r\n\r\nI would also recommend that you start applying for scholarships as early as possible so that you can accumulate enough money to attend, as I might not be able to this year for financial reasons (though I believe that they offer a good deal of financial aid).",
"Solution_3": "miraculouspostmaster: did you have any sports? i've heard that you really have to have a sport to get into harvard unless youre extremely lopsided on something like westinghouse 1st place or IMO. people are telling me to just join a team, even if i won't be any good at it. is this at all true?",
"Solution_4": "The general word on the street is that athletics is enormously helpful for Ivy League admissions--after all, the Ivy League is a sports conference, and Harvard has more Division I NCAA sports than any other college. Most of the local Harvard alumni who attended the regional information session in my town a year and a half ago were college athletes--the only one who was not has a physical disability, and he was a member of the marching band. One book about college admission advice seriously suggests taking up a sport to strengthen the application for a bright, well-rounded kid without a superimpressive extracurricular activity. \r\n\r\nFor all that, there are counterexamples of people with no sports background who get into Harvard. Those who do are frequently \"lopsided\" up to the international level in some kind of nonsport extracurricular activity, for example the International Mathematical Olympiad.",
"Solution_5": "There are lots and lots of different types of people who get into Harvard. Sure, some atheletes get into Harvard, and there are also plenty of people here who are unathletic. I certainly didn't participate in organized sports in high school. There isn't any reason to think that being atheletic will help you get into college (any college) more than anything else you do, if you aren't being recruited. Certainly, don't join a sports team that you won't enjoy being a member of -- spend your time doing something that interests you.\r\n\r\nThe fundamental truths of college admissions: there is no formula for getting in, and there are more qualified individuals than there are spots at any one college. Their corollaries: do the things that you will enjoy, not just the things that will pad your resume, and don't base your life around getting in to Harvard -- if you're good, you'll get in somewhere good.",
"Solution_6": "Someone from Collegeconfidential got into Harvard with a 1300(He was an Asian male too!!). Thus high SAT isn't necessary; Harvard is rather arbitrary looking for people who would fit in best with the campus.\r\n\r\nPlus Nathan Pflueger got a Harvard reject and he was pretty damn qualified.\r\n\r\nI think it's much easier to find a formula for Caltech admission rather than Harvard admission because Caltech looks purely at merit rather than arbitrary factors.",
"Solution_7": "I think Harvard would characterize that as that it looks for a broader variety of forms of \"merit\" than does Caltech. I agree with the basic point that it is probably easier to figure out a plan for a high probability of getting into Caltech than a plan for a high probability of getting into Harvard--although it might be HARDER to find a plan for getting out of Caltech with an undergraduate degree (in view of its rather high drop-out rate) than for completing a degree program at Harvard.",
"Solution_8": "No school organized sports, but anyone who participates in drum corps knows that the rigors of a summer on tour are just as physically demanding as any high school sport (I lost about 60 pounds over a combined six months). The man who interviewed me seemed to be interested in physical activity though.",
"Solution_9": "I've never heard that you need a sport to get into Harvard...I think what the person who told you this means is that unless you're well-lopsided, you need to be very well-rounded and thus participate in a variety of activities (but not necessarily sports).",
"Solution_10": "I know a few people who got in as an undergrad, and none of them was particularly good on any sport (actually, at least one of them was REALLY BAD in any sport)."
}
{
"Tag": [
"real analysis",
"real analysis solved"
],
"Problem": "let $f$ be a maximal solution of the (pendulum) equation:\r\n$y''+ay'+bsin(y)=0$ where $a>0$. Prove that $f$ is defined over $R$.",
"Solution_1": "What is a maximal solution in your opinion?",
"Solution_2": "a maximal solution $f$ is a solution defined on an intervalle $I$ such that there does not exist a greater intervalle $J$ and a solution $g$ defined on $J$ with $f=g$ on $I$.",
"Solution_3": "It is an immediate consequence of the following general theorem: if $X'=F(X)$ is a system of differential equations ($X(t)\\in \\mathbb R^n$, $F:\\mathbb R^n\\to\\mathbb R^n$ and $F$ is Lipschitz in the entire space, the solution of any initial value problem exists and is unique on the entire real line.",
"Solution_4": "why do you mean by Lipschitz ? locally Lipschitz or globally ?",
"Solution_5": "He wrote \"$F$ is Lipschitz in the entire space\". I think it means globally, i.e. with the same constant.",
"Solution_6": "[quote=\"Myth\"]He wrote \"$F$ is Lipschitz in the entire space\". I think it means globally, i.e. with the same constant.[/quote]\r\nYeah, that's what I meant. Sorry for not making it clear the first time..."
}
{
"Tag": [
"search"
],
"Problem": "I'm sorry if this isn't the right forum, for this, but I need an answer for my homework.\r\n\r\nWhat eating habits do people from Venezuela have?\r\n\r\n--Eating Times\r\n--Breakfast foods\r\n--Lunch foods\r\n--Dinner foods\r\n\r\nThanks in advance!",
"Solution_1": "Um... use the internet.... search engines are your friend.",
"Solution_2": "I don't think this there is an appropriate forum on AoPS for this, AoPS is not a place to seek answers to homework...\r\n\r\nI assumed this was a discussion Venezuela's political status and Hugo Chavez, but, I found a request for eating habits of Venezuelans. :o"
}
{
"Tag": [
"calculus",
"integration",
"trigonometry",
"function",
"limit",
"complex analysis"
],
"Problem": "Evaluate ${ \\int_{0}^{\\infty}\\frac{1}{x^{n}+1}dx}$ if $ n$ is an integer with $ n \\ge 2$.\r\n\r\n(In particular, try to find a nice form)\r\n\r\nIf anyone can find a proof that does not use complex analysis, I would be interested to hear.",
"Solution_1": "Since you would actually prefer to see a proof that does not use complex analysis, I can give you a one-liner:\r\n\r\nLet $ u=x^{n}\\rightarrow dx=\\frac{du}{n u^{\\frac{n-1}{n}}}$. So we have:\r\n\r\n$ \\int_{0}^{\\infty}\\frac{1}{x^{n}+1}\\hspace{1 mm}dx = \\frac{1}{n}\\int_{0}^{\\infty}\\frac{u^{\\frac{1}{n}-1}}{u+1}du=\\frac{1}{n}B(\\frac{1}{n},\\frac{n-1}{n})=\\frac{1}{n}\\Gamma(\\frac{1}{n})\\Gamma(\\frac{n-1}{n})=\\frac{\\pi}{n \\sin(\\frac{\\pi}{n})}$\r\n\r\nOkay, it would be a one-liner if AoPS did not wrap text so early in its buffer :lol: But in the above I just used an alternative definition of the Beta function, the relation between Beta and Gamma, and one of the Gamma identities.\r\n\r\nUsing complex analysis, one would consider a contour in the shape of a wedge containing the positive real axis from $ 0$ to $ R$, the arc of the circle of angle $ \\frac{2 \\pi}{n}$, and the radius of the circle from the end of the arc back to $ 0$ oriented counterclockwise when $ n$ is odd, and change the angle to $ \\frac{\\pi}{n}$ when $ n$ is even. \r\n\r\nYou only have one residue here at $ z=e^{\\frac{\\pi i}{n}}$.\r\n\r\n$ \\text{Res}(\\frac{1}{1+z^{n}},e^{\\frac{\\pi i}{n}}) = \\lim_{z\\rightarrow e^{\\frac{\\pi i}{n}}}\\frac{z-e^{{\\pi i}{n}}}{1+z^{n}}=\\frac{1}{n e^{\\frac{(n-1) \\pi i}{n}}}$\r\n\r\nManipulation of the integrals and limiting estimates will give:\r\n\r\n$ \\int_{0}^{\\infty}\\frac{1}{x^{n}+1}\\hspace{1 mm}dx = 2 \\pi i \\frac{1}{n e^{\\frac{(n-1) \\pi i}{n}}}\\frac{1}{1-e^{\\frac{2 \\pi i}{n}}}= \\frac{\\pi}{n}\\frac{2 i}{e^{\\frac{\\pi i}{n}}-e^{-\\frac{\\pi i}{n}}}= \\frac{\\pi}{n \\sin(\\frac{\\pi}{n})}$",
"Solution_2": "I don't consider that a proof without complex analysis- unless you have a proof of those identities that also uses complex analysis.\r\n\r\nSee also [url=http://www.artofproblemsolving.com/Forum/viewtopic.php?&t=21316]this thread[/url], which contains a full version of the wedge argument."
}
{
"Tag": [
"number theory unsolved",
"number theory"
],
"Problem": "find x,y,z in this equation that x,y,z are positive integers.\r\n(x+y)(1+xy)=2^z",
"Solution_1": "Because $ x\\plus{}y$ is power $ 2^t$ we can write $ x\\equal{}2^{t\\minus{}1}\\plus{}a,y\\equal{}2^{t\\minus{}1}\\minus{}a$, then $ xy\\equal{}2^{2t\\minus{}2}\\minus{}a^2$.\r\nTherefore $ 2^{2t\\minus{}2}\\plus{}1\\minus{}2^{z\\minus{}t}\\equal{}a^2$. It give $ z\\minus{}t\\equal{}t,a\\equal{}\\pm (2^{t\\minus{}1}\\minus{}1)$ or $ (x,y,z)\\equal{}(2^t\\minus{}1,1,2t),(1,2^t\\minus{}1,2t).$",
"Solution_2": "Posted before, take searches next time. But especially do never use such useless titles again! (see http://www.mathlinks.ro/viewtopic.php?t=135914 )"
}
{
"Tag": [
"geometry",
"rhombus",
"geometry proposed"
],
"Problem": "Let $ ABCD$ be a rhombus with the length of each side is $ 1$. Denote $ M\\in BC$ and $ N\\in CD$ such that $ MC\\plus{}CN\\plus{}NM\\equal{}2$ and $ 2\\angle {MAN}\\equal{}\\angle {BAD}$. Find the the values of $ \\angle ABC$, $ \\angle BAD$.",
"Solution_1": "Construct ray $ Ax$ such that $ AM$ is the bisector of $ \\angle BAx$. Let $ G$ be a point on $ Ax$ such that $ AG\\equal{}AB\\equal{}AD$. \r\nBecause $ \\angle BAD\\equal{}2\\angle MAN$ then $ \\Delta ABM\\equal{}\\Delta AGM, \\Delta ADN\\equal{}\\Delta AGN$\r\n$ \\Rightarrow MG\\equal{}MB, NG\\equal{}ND$.\r\nSince $ MC\\plus{}NC\\plus{}MN\\equal{}2\\equal{}BC\\plus{}BD$ we get $ MN\\equal{}BM\\plus{}DN\\equal{}MG\\plus{}NG \\Rightarrow M,G,N$ are collinear.\r\n$ 2\\angle ABC\\equal{}\\angle ABM\\plus{}\\angle ADN\\equal{}\\angle AGM\\plus{}\\angle AGN\\equal{}180^o$ therefore $ \\angle ABC\\equal{}90^o\\equal{}\\angle BAD$."
}
{
"Tag": [],
"Problem": "\u0394\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03b7 \u03c3\u03c5\u03bd\u03ac\u03c1\u03c4\u03b7\u03c3\u03b7 $ f(x)\\equal{}x^2\\minus{}1$ \u03c3\u03c4\u03bf \u03b4\u03b9\u03ac\u03c3\u03c4\u03b7\u03bc\u03b1 [-3,3]\r\n\r\n\u039d\u03b1 \u03ba\u03b1\u03c4\u03b1\u03c3\u03ba\u03b5\u03c5\u03b1\u03c3\u03c4\u03b5\u03af \u03bc\u03b9\u03b1 \u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03af\u03b1 \u03ba\u03bb\u03b9\u03bc\u03b1\u03ba\u03c9\u03c4\u03ce\u03bd \u03c3\u03c5\u03bd\u03b1\u03c1\u03c4\u03ae\u03c3\u03b5\u03c9\u03bd \u03c0\u03bf\u03c5 \u03c3\u03c5\u03b3\u03ba\u03bb\u03af\u03bd\u03b5\u03b9 \u03bf\u03bc\u03bf\u03b9\u03cc\u03bc\u03bf\u03c1\u03c6\u03b1 \u03c3\u03c4\u03b7\u03bd f\r\n\r\n\r\n\r\n(\u039c\u03b9\u03b1 \u03c3\u03c5\u03bd\u03ac\u03c1\u03c4\u03b7\u03c3\u03b7 s \u03bf\u03c1\u03b9\u03c3\u03bc\u03ad\u03bd\u03b7 \u03c3\u03c4\u03bf [\u03b1,\u03b2] \u03bb\u03ad\u03b3\u03b5\u03c4\u03b1\u03b9 \u03ba\u03bb\u03b9\u03bc\u03b1\u03ba\u03c9\u03c4\u03ae \u03b1\u03bd \u03c5\u03c0\u03ac\u03c1\u03c7\u03b5\u03b9 \u03bc\u03b9\u03b1 \u03b4\u03b9\u03b1\u03bc\u03ad\u03c1\u03b9\u03c3\u03b7 {lo, l1, l2,...ln} \u03c4\u03bf\u03c5 [\u03b1,\u03b2]\r\n\r\n\u03c4\u03ad\u03c4\u03bf\u03b9\u03b1 \u03ce\u03c3\u03c4\u03b5 \u03b7 s \u03bd\u03b1 \u03b5\u03af\u03bd\u03b1\u03b9 \u03c3\u03c4\u03b1\u03b8\u03b5\u03c1\u03b7 \u03c3\u03b5 \u03ba\u03ac\u03b8\u03b5 \u03b4\u03b9\u03ac\u03c3\u03c4\u03b7\u03bc\u03b1 (li-1, li) \u03b3\u03b9\u03b1 \u03ba\u03ac\u03b8\u03b5 i \u03b1\u03bd\u03ae\u03ba\u03b5\u03b9 \u03c3\u03c4\u03bf {1,2,3,...n}.)",
"Solution_1": "Ummm... \u03c5\u03c0\u03ac\u03c1\u03c7\u03bf\u03c5\u03bd \u03c0\u03bf\u03bb\u03bb\u03bf\u03af \u03c4\u03c1\u03cc\u03c0\u03bf\u03b9 \u03bd\u03b1 \u03c4\u03bf \u03ba\u03ac\u03bd\u03b5\u03b9 \u03ba\u03b1\u03bd\u03b5\u03af\u03c2. \u0391\u03bd \u03c3\u03b5 \u03b4\u03c5\u03c3\u03ba\u03bf\u03bb\u03b5\u03cd\u03b5\u03b9 \u03ba\u03b1\u03b9 \u03b2\u03b1\u03c1\u03b9\u03ad\u03c3\u03b1\u03b9 \u03bd\u03b1 \u03b3\u03c1\u03ac\u03c8\u03b5\u03b9\u03c2, \u03ba\u03bf\u03af\u03c4\u03b1 \u03c4\u03b7\u03bd \u03b1\u03c0\u03cc\u03b4\u03b5\u03b9\u03be\u03b7 \u03c4\u03bf\u03c5 \u03c0\u03c5\u03ba\u03bd\u03bf\u03cd \u03c4\u03c9\u03bd \u03ba\u03bb\u03b9\u03bc\u03b1\u03ba\u03c9\u03c4\u03ce\u03bd \u03c3\u03c5\u03bd\u03b1\u03c1\u03c4\u03ae\u03c3\u03b5\u03c9\u03bd :)\r\n\r\nCheerio,\r\n\r\nDurandal 1707",
"Solution_2": "\u03a0\u03bf\u03c5 \u03b2\u03c1\u03b9\u03c3\u03ba\u03b5\u03c4\u03b1\u03b9 \u03b1\u03c5\u03c4\u03ae \u03b7 \u03b1\u03c0\u03cc\u03b4\u03b5\u03b9\u03be\u03b7 \u03c6\u03af\u03bb\u03b5 Durandal?\r\n\u0394\u03b5\u03bd \u03c4\u03b7\u03bd \u03be\u03ad\u03c1\u03c9 \u03bc\u03c0\u03bf\u03c1\u03b5\u03af\u03c2 \u03bd\u03b1 \u03bc\u03bf\u03c5 \u03c0\u03b5\u03b9\u03c2 \u03bb\u03af\u03b3\u03b1 \u03c0\u03c1\u03ac\u03b3\u03bc\u03b1\u03c4\u03b1 \u03b3\u03b9\u0384\u03b1\u03c5\u03c4\u03cc \u03b3\u03b9\u03b1\u03c4\u03af \u03ad\u03c8\u03b1\u03be\u03b1 \u03c3\u03b5 \u03b2\u03b9\u03b2\u03bb\u03af\u03b1 \u03ba\u03b1\u03b9 \u03b4\u03b5\u03bd \u03c4\u03bf \u03b2\u03c1\u03ae\u03ba\u03b1\r\n\u03c0\u03c9\u03c2 \u03c4\u03b7\u03bd \u03b2\u03c1\u03af\u03c3\u03ba\u03bf\u03c5\u03bc\u03b5 \u03b1\u03c5\u03c4\u03ae\u03bd \u03c4\u03b7\u03bd \u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03af\u03b1?",
"Solution_3": "[quote=\"spinos\"]\u0394\u03af\u03bd\u03b5\u03c4\u03b1\u03b9 \u03b7 \u03c3\u03c5\u03bd\u03ac\u03c1\u03c4\u03b7\u03c3\u03b7 $ f(x) \\equal{} x^2 \\minus{} 1$ \u03c3\u03c4\u03bf \u03b4\u03b9\u03ac\u03c3\u03c4\u03b7\u03bc\u03b1 [-3,3]\n\n\u039d\u03b1 \u03ba\u03b1\u03c4\u03b1\u03c3\u03ba\u03b5\u03c5\u03b1\u03c3\u03c4\u03b5\u03af \u03bc\u03b9\u03b1 \u03b1\u03ba\u03bf\u03bb\u03bf\u03c5\u03b8\u03af\u03b1 \u03ba\u03bb\u03b9\u03bc\u03b1\u03ba\u03c9\u03c4\u03ce\u03bd \u03c3\u03c5\u03bd\u03b1\u03c1\u03c4\u03ae\u03c3\u03b5\u03c9\u03bd \u03c0\u03bf\u03c5 \u03c3\u03c5\u03b3\u03ba\u03bb\u03af\u03bd\u03b5\u03b9 \u03bf\u03bc\u03bf\u03b9\u03cc\u03bc\u03bf\u03c1\u03c6\u03b1 \u03c3\u03c4\u03b7\u03bd f.[/quote]\r\n\r\n\u039d\u03b1 \u03b2\u03ac\u03bb\u03c9 \u03bc\u03af\u03b1 \u03b5\u03cd\u03ba\u03bf\u03bb\u03b7 \u03bb\u03cd\u03c3\u03b7 \u03bc\u03b9\u03b1\u03c2 \u03ba\u03b1\u03b9 \u03ba\u03b1\u03bd\u03b5\u03af\u03c2 \u03b4\u03b5\u03bd \u03ad\u03c7\u03b5\u03b9 \u03b3\u03c1\u03ac\u03c8\u03b5\u03b9 \u03c4\u03af\u03c0\u03bf\u03c4\u03b1 \u03ba\u03b1\u03b9 \u03b5\u03af\u03bd\u03b1\u03b9 \u03ba\u03c1\u03af\u03bc\u03b1 \u03bd\u03b1 \u03bc\u03ad\u03bd\u03bf\u03c5\u03bd \u03b5\u03cd\u03ba\u03bf\u03bb\u03b5\u03c2 \u03b1\u03c3\u03ba\u03ae\u03c3\u03b5\u03b9\u03c2 \u03b1\u03bd\u03b1\u03c0\u03ac\u03bd\u03c4\u03b7\u03c4\u03b5\u03c2...\r\n\r\n\u0395\u03c0\u03b5\u03b9\u03b4\u03ae \u03b2\u03b1\u03c1\u03b9\u03ad\u03bc\u03b1\u03b9 \u03bd\u03b1 \u03b3\u03c1\u03ac\u03c6\u03c9 \u03c0\u03bf\u03bb\u03bb\u03ac, \u03b8\u03b1 \u03b4\u03bf\u03c5\u03bb\u03ad\u03c8\u03bf\u03c5\u03bc\u03b5 \u03bc\u03b5 \u03c4\u03b7\u03bd $ f(x) \\equal{} x^2$ \u03c3\u03c4\u03bf $ [0,1]$.\r\n\r\n\u0398\u03b5\u03ce\u03c1\u03b7\u03c3\u03b5 \u03bc\u03af\u03b1 \u03b4\u03b9\u03b1\u03bc\u03ad\u03c1\u03b9\u03c3\u03b7 $ a_0,a_1\\ldots a_n$ \u03c4\u03bf\u03c5 \u03b4\u03b9\u03b1\u03c3\u03c4\u03ae\u03bc\u03b1\u03c4\u03bf\u03c2 $ [0,1]$ \u03cc\u03c0\u03bf\u03c5 \u03c4\u03bf $ a_k$ \u03b5\u03af\u03bd\u03b1\u03b9 \u03bb\u03cd\u03c3\u03b7 \u03c4\u03b7\u03c2 \u03b5\u03be\u03af\u03c3\u03c9\u03c3\u03b7\u03c2 $ f(x) \\equal{} \\frac{k}{n}$ (\u03b4\u03b7\u03bb. $ a_k \\equal{} \\sqrt{\\frac{k}{n}}$).\r\n\r\nT\u03cc\u03c4\u03b5, \u03b1\u03bd $ f_n(x) \\equal{} \\frac{k}{n}$ \u03b3\u03b9\u03b1 $ x\\in[a_k, a_{k\\plus{}1})$, $ k\\equal{}0,1\\ldots n\\minus{}1$, \u03b8\u03b1 \u03ad\u03c7\u03b5\u03b9\u03c2 $ \\|f_n\\minus{}f\\|\\leq \\frac{1}{n}$ \u03ba\u03b1\u03b9, \u03c3\u03c5\u03bd\u03b5\u03c0\u03ce\u03c2 $ f_n\\to f$ \u03bf\u03bc\u03bf\u03b9\u03cc\u03bc\u03bf\u03c1\u03c6\u03b1.\r\n\r\nCheerio,\r\n\r\nDurandal 1707"
}
{
"Tag": [
"function"
],
"Problem": "I have a TI-84 Silver...and when I push the y= button on the top left of my calculator, I get the y1, y2, y3... and when i put, for example, one linear function in y1, and one linear function in y2, only one of the graph's actually graph when I push graph!! I tried pressing the = button to highlight it but only one \"=\" will be highlighted at a time...I also tried pushing enter on the little figure on the left of the y1, y2, y3's....but they won't return to the figure of \" \\ \"....it's an arrow with a vertical line behind it right now...I don't know how to change it back to be able to graph two graphs at a time!!\r\n\r\nPlease help!\r\nThanks!",
"Solution_1": "nvmm...i just resetted all the RAM...and it fixed it at the cost of my programs...but oh well!",
"Solution_2": "You probably had an application running that changed some of the settings.\r\n\r\nI think the symbol of an arrow with a vertical line belongs to the Transformation Graphing application."
}
{
"Tag": [
"analytic geometry",
"Asymptote",
"function"
],
"Problem": "A square is inscribed in a right triangle with legs of 8 units and 15 units. If two of\r\nthe vertices of the square lie on the hypotenuse and the other two vertices of the\r\nsquare lie on the legs of the triangle, what is the length of a side of the square?\r\nExpress your answer as a common fraction. \r\n\r\n\r\nPlease tell me how to draw that so that I can answer another person's question. Thanks!",
"Solution_1": "Once you know the answer, you can just \"compute\" the coordinates of all vertices:\r\n[asy]size(150);\ndraw((0,0)--(15,0)--(0,8)--cycle);\nreal a=960/409, b=15*a/8;\ndraw((0,a)--(b,0)--(b+a,b)--(a,a+b)--cycle);[/asy]\r\nNote that I didn't try to divide 960 by 409 on a calculator and write in the resulting long decimal but just asked Asymptote to do every relevant numerical computation. \r\n\r\nJust for your fun, here is a function that, given three pairs $ A,B,C$ inscribes a square into the triangle with 2 vertices lying on $ BC$ and draws the result (I used the foot command from the Olympiad package in the code; it is automatically imported on the Forum anyway but you'll need to add [i]import olympiad;[/i] to your code if you want it to work on your home computer).\r\n[code]\nvoid f(pair A,pair B, pair C)\n{\npair P=foot(A,B,C);\nreal a=abs(B-C), h=abs(P-A), t=h/(a+h);\ndraw(A--B--C--cycle);\npair P=A+t*(B-A), Q=A+t*(C-A), R=foot(Q,B,C), S=foot(P,B,C);\ndraw(P--Q--R--S--cycle);\n}\n[/code]\r\nHere is a picture (for some other triangle):\r\n[asy]size(150);\nvoid f(pair A,pair B, pair C)\n{\npair P=foot(A,B,C);\nreal a=abs(B-C), h=abs(P-A), t=h/(a+h);\ndraw(A--B--C--cycle);\npair P=A+t*(B-A), Q=A+t*(C-A), R=foot(Q,B,C), S=foot(P,B,C);\ndraw(P--Q--R--S--cycle);\n}\nf((0,0),(3,-1),(1,5));[/asy]"
}
{
"Tag": [],
"Problem": "For how many digits C is the positive three-digit number 1C3 a\nmultiple of 3?",
"Solution_1": "Because 1+c+3 must be divisible by 3, c=2,5,8. Thus there are 3 digits C can be."
}
{
"Tag": [
"linear algebra",
"matrix",
"induction",
"algebra",
"polynomial",
"linear algebra unsolved"
],
"Problem": "How can i prove that \\[ A^{n}\\equal{}O_{n}\r\n\\] if \\[ A\\in M_{n}(R)\r\n\\] is a nilpotent matrix ?",
"Solution_1": "Cayley-Hamilton offers some possibilities for the proof, but let's try this in a more elementary style.\r\n\r\nTheorem: a nilpotent matrix (over any field) is similar to an upper triangular matrix with zeros on the main diagonal.\r\n\r\nSince $ A$ is nilpotent, there must be some $ v\\ne 0$ such that $ Av \\equal{} 0.$ (Just take any nonzero $ x$ and look at the sequence $ x,Ax,A^2x,\\dots.$ The last nonzero term in that sequence can be taken as $ v.$) Extend $ v$ to a basis, meaning find an invertible $ P$ with $ v$ as the first column. Then we have the following block structured result:\r\n\r\n$ B \\equal{} P^{ \\minus{} 1}AP \\equal{} \\begin{pmatrix}0 & * \\\\\r\n0 & A'\\end{pmatrix}.$\r\n\r\nBy matrix multiplication, we get that $ B^k \\equal{} \\begin{pmatrix}0 & * \\\\\r\n0 & (A')^k\\end{pmatrix}.$\r\n\r\nHence, $ (A')^k \\equal{} 0$ for large enough $ k$ and $ A'$ is itself nilpotent. Continue the process. Eventually (essentially this is an induction on the size of the matrix) we get that $ A$ is similar to an upper triangular matrix with zeros on the main diagonal.\r\n\r\nThe rest is just matrix multiplication.\r\n\r\nIf $ A \\equal{} \\begin{bmatrix}0 & * & * & & \\\\\r\n0 & 0 & * & * & \\\\\r\n0 & 0 & 0 & * & * \\\\\r\n& & & \\ddots & \\\\\r\n0 & & & & 0\\end{bmatrix},$ then\r\n\r\n$ A^2 \\equal{} \\begin{bmatrix}0 & 0 & * & & \\\\\r\n0 & 0 & 0 & * \\\\\r\n0 & 0 & 0 & 0 & * \\\\\r\n& & & \\ddots & \\\\\r\n0 & & & & 0\\end{bmatrix}.$\r\n\r\nContinuing, we get a new diagonal of zeros with each power, until we run out of room, so $ A^n \\equal{} 0.$",
"Solution_2": "This is one of the possibilities:\r\n$ A^k\\equal{}O_{n}\\Rightarrow\\lambda_{i}^k\\equal{}0$, where $ \\lambda_{i}$ is an eigenvalue of A;$ i\\equal{}\\overline{1,n}$. So all the eigenvalues are equal to 0. Then the characteristic polynomial $ p\\equal{}(X\\minus{}\\lambda_{1})\\cdot(X\\minus{}\\lambda_{2})\\cdot...\\cdot(X\\minus{}\\lambda_{n})\\equal{}X^{n}$.\r\nIt is known that $ p(A)\\equal{}O_{n}$(Cayley-Hamilton Thorem), so you get $ A^n\\equal{}O_{n}$.",
"Solution_3": "thanks for the solution...i didn't tried to solve it with eignvalues, it's a good idea :D"
}
{
"Tag": [
"search",
"logarithms",
"trigonometry",
"geometry",
"calculus",
"AMC",
"AIME"
],
"Problem": "I stumbled upon this website while doing a google search for some mathy stuff... I'm a math major at a pretty big state school here in the.... I just have to say, I cannot believe that those high school guys are solving such complex stuff... I read up on the students' solutions to:\r\n\r\nHow many solutions are there to\r\nx + 2y + 3z = 100\r\nedit: x,y,z in N\r\n\r\nAnd was shocked at the complexity of their answers.\r\n\r\nSo do you guys study algebra on your own? I certainly did not learn any abstract algebra in high school - I went to one of the best high schools in my state and the highest level math we offered was multivariable calc for college credit...\r\n\r\nAnyways, it's awesome to see high school students solving problems that I am just now encountering in college!\r\n\r\nedit for clairty on that example pblm above..\r\n\r\nIn response to a post below, I'm pretty competent in quantum mechanics, so I guess that does make me feel a little better :)",
"Solution_1": "I doubt many, if any, of us learn much math as part of the high school curriculum. However, it is true that most of us study further topics on our own by getting more advanced books or using this forum as a resource.",
"Solution_2": "Many people in advanced mathematics in high school like to take classes outside of their own curriculm (not classes at school, but maybe looking at stuff they are not learning yet). For example, I'm only in 8th grade but I am reading up on calculus. My 10th grade friend left high school and started doing linear algebra. \r\n\r\nFor your question, many of us do study math on our own. For some reason, many people at my school would rather go see movies instead of doing logarithms or trig :roll: . Those silly people. \r\n\r\nActually, school math curriculms are notoriously slow and wasteful. As proof, chess64 and I skipped about 2-3 years worth of math in less than one school year because we were bored of the way we were taught. Our teachers would focus more on the people who needed help or wouldn't pay attention rather than the people who are looking for a challenge (in my opinion they should treat both groups of people equally). \r\n\r\nBecause of the sluggish math at school, many of us are obligated to feel that we need to take our own education in our own hands.",
"Solution_3": "For the question, x can be anything, y can be anything, and z is just the result when you plug in an x and a y.\r\n\r\n\r\n\r\nDo you mean natural numbers?\r\n\r\nThen that's much harder.\r\n\r\n\r\nTry going to the basics forum. If you can solve everything there, you are stuck; like me! :(",
"Solution_4": "Welcome to the world of mathlinks, jgb!\r\n\r\nI am also a college freshman, but many fourteen year olds on this site know more of number theory and graph theory (let alone geometry) than me. Now I've got used to it!\r\n\r\nSomething similar happened in college some days ago. I read up the official statement of the Yang Mills theory problem, and next day in college, told my friends that the mass and velocity of a particle are noncommuting operators operating on a Hilbert space! :rotfl: That day, during lunch, when two seniors heard two of my friends discussing this, their mouths dropped open! :D \r\n\r\nWanna stay ahead of these kids? Read up stuff like quaternions and lambda calculus and complex analysis and quantum mechanics and game theory, as I am now trying to do. :(",
"Solution_5": "[quote=\"bubka\"]Wanna stay ahead of these kids? Read up stuff like quaternions and lambda calculus and complex analysis and quantum mechanics and game theory, as I am now trying to do. :([/quote]\r\n\r\nSorry. I'm studying game theory this summer.",
"Solution_6": "Why? I thought it was for statisticians alone. :maybe: \r\nAnyway, either you are way above fourteen, or you have missed at least one of those I mentioned. :)",
"Solution_7": "I'm curious: What's the solution to the posted problem? Could someone post a link?\r\n\r\nCurrently I'm trying to study quantum mechanics as much as I can :harhar:",
"Solution_8": "[quote=\"bubka\"]Why? I thought it was for statisticians alone. :maybe: \nAnyway, either you are way above fourteen, or you have missed at least one of those I mentioned. :)[/quote]\r\n\r\nNeither. :D \r\n\r\n[hide=\"partial solution\"]For a natural number x, we need to find the number of solutions to the diophantine equation 2y+3z=100-x.\n\nI could find a pattern, but I'm too tired right now.[/hide]",
"Solution_9": "That is not a diophantine equation.",
"Solution_10": "When you plug in x, it is. :wink: \r\n\r\n[hide=\"solution\"]For x=1, you get (x,y,z)=\n\n(1,0,33), (1,3,31)....(1,48,1)\n\n17 pairs\n\nFor x=2,\n\n(2,1,32),(2,4,30).......(2,49,0)\n\n17 pairs\n\nFor x=3,\n\n17 pairs.\n\n\n17+17+17+16+16+16+15+15+15+14+14+14+13+13+13+12+12+12+11+11+11+10+10+10+9+9+9+8+8+8+7+7+7+6+6+6+5+5+5+4+4+4+3+3+3+2+2+2+1+1+1=459[/hide]",
"Solution_11": "[quote=\"7h3.D3m0n.117\"]Many people in advanced mathematics in high school like to take classes outside of their own curriculm (not classes at school, but maybe looking at stuff they are not learning yet). For example, I'm only in 8th grade but I am reading up on calculus. My 10th grade friend left high school and started doing linear algebra. \n\nFor your question, many of us do study math on our own. For some reason, many people at my school would rather go see movies instead of doing logarithms or trig :roll: . Those silly people. \n\nActually, school math curriculms are notoriously slow and wasteful. As proof, chess64 and I skipped about 2-3 years worth of math in less than one school year because we were bored of the way we were taught. Our teachers would focus more on the people who needed help or wouldn't pay attention rather than the people who are looking for a challenge (in my opinion they should treat both groups of people equally). \n\nBecause of the sluggish math at school, many of us are obligated to feel that we need to take our own education in our own hands.[/quote]\r\n\r\nYeah, that's pretty much what I did.\r\n\r\nMost people would rather see a movie than do math?! :o :D",
"Solution_12": "[quote=\"mathnerd314\"]Most people would rather see a movie than do math?! :o :D[/quote]\r\n\r\nIt's hard to believe buts its true. I don't understand why either. :huh:",
"Solution_13": "LOL, what about doing math WHILE watching a movie? (Numb3rs, etc.)",
"Solution_14": "[quote=\"tjhance\"][quote=\"mathnerd314\"]Most people would rather see a movie than do math?! :o :D[/quote]\n\nIt's hard to believe buts its true. I don't understand why either. :huh:[/quote]\r\n\r\nYeah. My friend asked me if I could go with him with some girls to the movies to watch something and I said no cuz I had to read my calculus book :P .",
"Solution_15": ":huh: \r\n\r\nBad move...",
"Solution_16": "[quote=\"7h3.D3m0n.117\"]Actually, school math curriculms are notoriously slow and wasteful. As proof, chess64 and I skipped about 2-3 years worth of math in less than one school year because we were bored of the way we were taught. Our teachers would focus more on the people who needed help or wouldn't pay attention rather than the people who are looking for a challenge (in my opinion they should treat both groups of people equally).[/quote]\r\n\r\nObviously you didn't skip English and seeing as neither you nor chess64 received anything higher than a 4 on the AIME, I hardly consider that proof.",
"Solution_17": "...btw guys, not everybody is interested in doing math for fun - do you guys go to the weight room for fun?\r\n\r\n\r\n\r\n[quote=\"7h3.D3m0n.117\"]Yeah. My friend asked me if I could go with him with some girls to the movies to watch something and I said no cuz I had to read my calculus book :P .[/quote]\r\n\r\n\r\nman you fail at life\r\n\r\nyou shoulda told them to go to http://www.senhorcarteiro.blogspot.com/\r\n\r\nunless it was 300, then you should just plain go",
"Solution_18": "[quote=\"Ignite168\"][quote=\"7h3.D3m0n.117\"]Actually, school math curriculms are notoriously slow and wasteful. As proof, chess64 and I skipped about 2-3 years worth of math in less than one school year because we were bored of the way we were taught. Our teachers would focus more on the people who needed help or wouldn't pay attention rather than the people who are looking for a challenge (in my opinion they should treat both groups of people equally).[/quote]\n\nObviously you didn't skip English and seeing as neither you nor chess64 received anything higher than a 4 on the AIME, I hardly consider that proof.[/quote]\r\ni thought chess64 was a genius? he has 5 stars. :lol:",
"Solution_19": "[quote=\"Ignite168\"][quote=\"7h3.D3m0n.117\"]Actually, school math curriculms are notoriously slow and wasteful. As proof, chess64 and I skipped about 2-3 years worth of math in less than one school year because we were bored of the way we were taught. Our teachers would focus more on the people who needed help or wouldn't pay attention rather than the people who are looking for a challenge (in my opinion they should treat both groups of people equally).[/quote]\n\nObviously you didn't skip English and seeing as neither you nor chess64 received anything higher than a 4 on the AIME, I hardly consider that proof.[/quote]\r\n\r\nCompetition math is different from book math, and I thought that you, of all people would know. \r\n\r\n1) My teacher won't let me take AMC, so I wonder why I didn't take the AIME hmm? \r\n2) English is different. The curriculm is VERY flexible. You could be prepping for what you think you will be learning in 12th grade English but be doing something else. Math, however, is very strict.",
"Solution_20": "Yeah competition maths totally sucks (partly because they involve problem solving, which is no maths at all, but mostly because i am no good at competition maths).\r\n\r\nBut as I now sadly realize, so does analysis. And algebra."
}
{
"Tag": [
"Princeton",
"college",
"HMMT",
"Princeton Math Comp."
],
"Problem": "Welcome to the Princeton Math Competition forum. Princeton University will be hosting a high school math competition for the first time in the 2006-2007 school year. More information will be posted soon!",
"Solution_1": "You're amazing, Nathan. Congratulations on getting the competition set up and best of luck in the future.",
"Solution_2": "Looking forward to hearing more about it! I wonder if it's going to run concurrently with the HMMT...being able to go to both would be nice.",
"Solution_3": "sounds good. Good luck!",
"Solution_4": "We haven't fixed the date yet, but we're going to make sure it doesn't conflict with HMMT or any other big competitions in this region. We're also trying to pick a good date that avoids SAT weekends and other such tomfoolery. We should have the date, schedule, and test format available for your viewing pleasure by early next week."
}
{
"Tag": [],
"Problem": "I thought that it would be a good idea to start a topic like this one, to ask questions about this book.\r\nI will start.\r\n\r\nOn page 14, Engel says on the last line that $ 2^k(mod)n \\equal{} 0(mod)n$, schouldn't that be $ [2^k\\cdot a](mod)n \\equal{} 0(mod)n$ ?",
"Solution_1": "I couldn't edit, but PSS stands for Problem solving strategies.",
"Solution_2": "Can someone please take a look in the book, I will be very thankful."
}
{
"Tag": [
"abstract algebra",
"group theory",
"linear algebra",
"linear algebra unsolved"
],
"Problem": "Prove that a finite abelian subgroup in $GL_n(Z)$ has at most $ 6^n$ elements.",
"Solution_1": "harazi,\r\n\r\nThe biggest abelian subgroup of $\\mbox{GL}_{n}(\\mathbb{Z})$ I found has $2^{n}3^{n/3}$ elements.\r\n\r\nDo you have an explicit abelian subgroup whose order is $6^{n}$?",
"Solution_2": "No, actually one can prove that the biggest has $2^{[(n+1)/2]}\\cdot 3^{[n/2]}$ elements."
}
{
"Tag": [
"inequalities",
"real analysis",
"advanced fields",
"advanced fields unsolved"
],
"Problem": "Suppose $X$ is an $\\mathbb{R}^n$-valued random variable with finite second moments. (In real analysis terms, $X\\in L^2.)$ Notationally, treat $X$ as a (random) column vector.\r\n\r\nDefine $C=\\mathrm{Cov}(X)=E(XX^T)-E(X)E(X^T).$\r\n\r\n$C$ is obviously symmetric. It's a fairly easy standard argument to show that it is positive semidefinite, and positive definite if the components of $X$ are [i]linearly[/i] independent, which is a far, far, far weaker condition than probabilistic independence.\r\n\r\nIn what follows, assume that $C$ is invertible ([i]i.e.[/i] positive definite.)\r\n\r\nDefine $Q(X)=(X-E(X))^TC^{-1}(X-E(X)).$\r\n\r\nProve that for all $\\alpha>0$, $P(Q(X)\\ge \\alpha^2)\\le \\frac{n}{\\alpha^2}.$\r\n\r\n(Obviously, this is of interest only if $\\alpha>\\sqrt{n}.$)",
"Solution_1": "Isn't this the Hotelling T-squared?"
}
{
"Tag": [
"MATHCOUNTS",
"AMC",
"USA(J)MO",
"USAMO",
"AIME",
"videos"
],
"Problem": "You and Bob are at the same level in math competitions. During the summer, you study math competitions on your own and Bob goes to a math camp like Awesome Math... Is Bob going to beat you next year?\r\n\r\nHow many people would have the motivation and discipline to study on his own as hard as someone who is surrounded by others interested in math and gets lectures and problem sets planned? Is there a limit to what you can do on your own?[hide=\"Why I Ask\"]\n\nI'm wondering because in my middle school there was an enormous ability difference between the people who went to Awesome or other math camps and the next highest level students ... I might be overgeneralizing but from what I remember last year in middle school was that there were several \"tiers\" of ability which became apparent during mathcounts practices. \n\n-Highest level students (2 or 3 of them) \nAll went to math camps and all (I think) qualified for USAMO in middle school. Scores in mathcounts practices were roughly 100 points or more higher than everyone else's out of only 400 or 500 points. \n\n-Next to highest level (around 5 of them)\nAll had extensive tutoring for several years. Some of them qualified for AIME, close to perfect scores on the SAT math section\n\n-The rest (my group)\nBecame interested in math competitions in middle school, no significant training in elementary school.\n[/hide]",
"Solution_1": "I'm just going to do stuff myself this year, but I might go next year to a camp.\r\nThe reason I'm choosing to stay home is because I can do a lot of practice stuff, and have days off and breaks and stuff, while at camp, you have a schedule to follow.\r\n\r\nI'd probably fit in the last group.",
"Solution_2": "On one hand, teaching yourself allows you to analyze and examine your own weaknesses and strengths. On the other, being taught by others can allow you to see things in a different view and you can potentially learn a lot. \r\n\r\nThere are many advantages as well as disadvantages to both. I believe you need to analyze your own personality as to which you should chose. If you are lazy and procrastinating, then teaching yourself isn't a good idea. If you prefer open discussions about problems to benefit your learning, then attending a math camp or something along those lines might help.",
"Solution_3": "When improving in mathematics, there are three things I work on: raising ceiling performance, raising floor performance, and raising dexterity/reliability. These are general ideas relating to each:\r\n\r\n[color=darkred][b]Raising ceiling performance:[/b][/color] For this, going to a summer camp can be very effective. You will gain new perspectives, learn new methods/approaches, and perhaps new material altogether. However, there is plenty to be learned on the AoPS forum and such. Self study and playing around help plenty. Also, trying to make your own problems can allow you to feel the mindset of a problem writer, which is good to be familiar with. While trying to write problems, you yourself will gain a better understanding of the concepts as you try to make them fit together nicely.\r\n\r\n[color=darkred][b]Raising floor performance:[/b][/color] For this, you need to improve clarity of thought and health. If you keep your body in good condition, you will be more alert and conscious, and less likely to make a stupid mistake or to blank out. You will also have more energy to last through long tests. State of mind comes into play here, too. Exercise, meditation, playing music, art, etc, or genuine relaxation (not video games) have helped me here.\r\n\r\n[color=darkred][b]Raising reliability/dexterity:[/b][/color] While practice and problem writing can't be trumped, there are many other ways to improve your mental dexterity that have little to do with math competitions. These include playing chess, art, writing, debate/mock trial/etc, and more. To put things into perspective, this year, I barely learned any contest math material, and barely practiced for contests, but thanks to very involved mock trial work, have seen huge increases in almost all my contest scores.\r\n\r\n\r\nSo now, (if you believe my theory), it's up to you to decide where you want to put your attention :) \r\n\r\nAlso influencing your decision should be opportunity cost -- summers are very valuable."
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "a,b,c are positive real numbers.prove that:\r\nbc/a^2 +2bc +ca/b^2+2ac +ab/c^2+2ab<=1",
"Solution_1": "from cauchy $ (\\sum a^2\\plus{}2bc) \\sum \\frac{a^2}{a^2\\plus{}2bc}\\geq (a\\plus{}b\\plus{}c)^2$ so $ \\sum \\frac{a^2}{a^2\\plus{}2bc}\\geq 1$"
}
{
"Tag": [
"calculus",
"integration",
"logarithms",
"search",
"real analysis",
"real analysis unsolved"
],
"Problem": "show :D \r\n\r\n$ \\int_0^1\\;\\int_0^1\\;\\;\\dfrac{\\textbf dx\\;\\textbf dy}{\\left(\\minus{}\\,\\ln\\,(x\\,y)\\right)^{\\frac{3}{2}}}\\;\\;\\equal{}\\;\\;\\boxed{\\sqrt\\pi}$",
"Solution_1": "use the substitution $ u\\equal{}xy,v\\equal{}{\\frac {x}{y}}$we have that Jacobian $ J\\equal{}\\minus{}\\frac{1}{2v}$:\r\n$ I\\equal{}\\minus{}\\int _{0}^{1}\\! {du} \\int _{u}^{\\frac{1}{u}}\\! {dv}\\frac{1}{2v}\\frac{1}{(\\minus{}\\ln{u})^\\frac{3}{2}}\\equal{}\r\n\\minus{}\\int _{0}^{1}\\! \\frac{du}{(\\minus{}\\ln{u})^\\frac{3}{2}}\\int _{u}^{\\frac{1}{u}}\\! \\frac{dv}{2v}\\equal{}\r\n\\minus{}\\int _{0}^{1}\\! \\frac{du}{(\\minus{}\\ln{u})^\\frac{3}{2}}\\ln{u}\\equal{}\\int _{0}^{1}\\! \\frac{du}{(\\minus{}\\ln{u})^\\frac{1}{2}}$\r\nThen we use substitution $ t\\equal{}\\minus{}\\ln{u}$ and:\r\n$ I\\equal{}\\int _{0}^{1}\\! \\frac{du}{(\\minus{}\\ln{u})^\\frac{1}{2}}\\equal{}\\int _{0}^{\\infty}\\! \\frac{\\exp{t}}{t^\\frac{1}{2}}dt\\equal{}\\Gamma \\left( \\frac{1}{2} \\right)\\equal{}\\sqrt{\\pi}$",
"Solution_2": "See also [url=http://www.mathlinks.ro/viewtopic.php?search_id=1368209970&t=164877]here[/url]."
}
{
"Tag": [],
"Problem": "Let $S = \\{1, 2, \\ldots , 10\\}$. Find the number of unordered pairs A, B, where A and B are disjoint non-empty subsets of S. (http://www.kalva.demon.co.uk/aime/aime02b.html)",
"Solution_1": "[hide]\n\nFor each of the $10$ members of S, there are $3$ categories it can fall under: Set A, Set B, or Neither. So there are $3^{10}$ ways to assign the distinctions. Order does not matter in this case, so every set except the one where every number is part of Neither is double counted.\n\n$\\frac{3^{10}-1}{2}=29524$. We threw out the empty set (meaning A and B are empty, Neither has all 10 values) which we didn't want, and accounted for the double counting. But this includes the cases where either Set A or Set B is empty, while the other letter Set is not empty. To find the number of such cases, think of each value of S being assigned one of $2$ distinctions: part of Set A or part of Neither (we do not have to do this again for B since we are not concerned with order). \n\nSo there are $2^{10}-1$ such cases (we subtracted the case where all $10$ values were part of Neither since that case was thrown out in the first step).\n\nFinal answer equals: $29524-1023=28501$\n\n[/hide]"
}
{
"Tag": [
"inequalities",
"inequalities proposed"
],
"Problem": "Let $ a,b,c$ be positive real numbers. Prove that\r\n\r\n$ \\frac{a^2}{b^2}\\plus{}\\frac{b^2}{c^2}\\plus{}\\frac{c^2}{a^2}\\geq\\frac{2}{3}(a\\plus{}b\\plus{}c)(\\frac{1}{a}\\plus{}\\frac{1}{b}\\plus{}\\frac{1}{c})\\minus{}\\sqrt{(a^2\\plus{}b^2\\plus{}c^2)(\\frac{1}{a^2}\\plus{}\\frac{1}{b^2}\\plus{}\\frac{1}{c^2})}$",
"Solution_1": "[quote=\"Ligouras\"]Let $ a,b,c$ be positive real numbers. Prove that\n\n$ \\frac {a^2}{b^2} \\plus{} \\frac {b^2}{c^2} \\plus{} \\frac {c^2}{a^2}\\geq\\frac {2}{3}(a \\plus{} b \\plus{} c)(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c}) \\minus{} \\sqrt {(a^2 \\plus{} b^2 \\plus{} c^2)(\\frac {1}{a^2} \\plus{} \\frac {1}{b^2} \\plus{} \\frac {1}{c^2})}$[/quote]\r\n\r\n\r\neasy proven \r\n\r\n$ \\sqrt {(a^2 \\plus{} b^2 \\plus{} c^2)(\\frac {1}{a^2} \\plus{} \\frac {1}{b^2} \\plus{} \\frac {1}{c^2})}\\geq3$\r\n\r\nonly prove \r\n$ \\frac {a^2}{b^2} \\plus{} \\frac {b^2}{c^2} \\plus{} \\frac {c^2}{a^2} \\plus{} 3\\geq\\frac {2}{3}(a \\plus{} b \\plus{} c)(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c})$\r\n\r\nor \r\n$ (3a^2c^2 \\plus{} (bc^2 \\minus{} 2b^2c \\plus{} 3c^3)a \\plus{} 2b^2c^2 \\plus{} 3b^4 \\minus{} 4b^3c \\minus{} bc^3 \\plus{} 3c^4)(a \\minus{} b)(a \\minus{} c) \\plus{} ((b \\plus{} c)(3b^2 \\plus{} 2bc \\plus{} 3c^2)a \\minus{} bc(3b^2 \\plus{} 2bc \\plus{} 3c^2))(b \\minus{} c)^2\\geq 0$\r\nfrom the inequality assume $ a \\equal{} \\max{(a,b,c)}$\r\n\r\nthen\r\nonly prove while $ a \\equal{} \\max{(a,b,c)}$\r\n\r\n$ 3a^2c^2 \\plus{} (bc^2 \\minus{} 2b^2c \\plus{} 3c^3)a \\plus{} 2b^2c^2 \\plus{} 3b^4 \\minus{} 4b^3c \\minus{} bc^3 \\plus{} 3c^4\\geq 0$\r\nand \r\n$ (b \\plus{} c)(3b^2 \\plus{} 2bc \\plus{} 3c^2)a \\minus{} bc(3b^2 \\plus{} 2bc \\plus{} 3c^2)\\geq 0$",
"Solution_2": "$ (b\\plus{}c)(3b^2\\plus{}2bc\\plus{}3c^2)a\\minus{}bc(3b^2\\plus{}2bc\\plus{}3c^2)\\equal{}(3b^2\\plus{}2bc\\plus{}3c^2)(ba\\minus{}bc\\plus{}ac)\\geq 0$\r\n$ f(a)\\equal{}3c^2a^2\\plus{}(bc^2\\minus{}2b^2c\\plus{}3c^3)a\\plus{}2b^2c^2\\plus{}3b^4\\minus{}4b^3c\\minus{}bc^3\\plus{}3c^4$\r\n$ \\bigtriangleup\\equal{}(bc^2\\minus{}2b^2c\\plus{}3c^3)^2\\minus{}12c^2(2b^2c^2\\plus{}3b^4\\minus{}4b^3c\\minus{}bc^3\\plus{}3c^4)\\equal{}\\minus{}c^2(32b^4\\plus{}35b^2c^2\\minus{}44b^3c\\minus{}18bc^3\\plus{}27c^4)$\r\n\r\n$ 32b^4\\plus{}35b^2c^2\\minus{}44b^3c\\minus{}18bc^3\\plus{}27c^4\\geq 0$",
"Solution_3": "[quote=\"Ligouras\"]Let $ a,b,c$ be positive real numbers. Prove that\n\n$ \\frac {a^2}{b^2} \\plus{} \\frac {b^2}{c^2} \\plus{} \\frac {c^2}{a^2}\\geq\\frac {2}{3}(a \\plus{} b \\plus{} c)(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c}) \\minus{} \\sqrt {(a^2 \\plus{} b^2 \\plus{} c^2)(\\frac {1}{a^2} \\plus{} \\frac {1}{b^2} \\plus{} \\frac {1}{c^2})}$[/quote]\r\nneed to prove\r\n$ \\frac {a^2}{b^2} \\plus{} \\frac {b^2}{c^2} \\plus{} \\frac {c^2}{a^2} \\plus{} 3\\geq\\frac {2}{3}(a \\plus{} b \\plus{} c)(\\frac {1}{a} \\plus{} \\frac {1}{b} \\plus{} \\frac {1}{c})$\r\n$ \\Longleftarrow 3\\plus{}3(\\frac{a^2}{b^2}\\plus{}\\frac{b^2}{c^2}\\plus{}\\frac{c^2}{a^2}) \\geq 2(\\frac{b\\plus{}c}{a}\\plus{}\\frac{c\\plus{}a}{b}\\plus{}\\frac{a\\plus{}b}{c})$\r\n$ \\Longleftarrow 3\\plus{}(\\frac{a}{b}\\plus{}\\frac{b}{c}\\plus{}\\frac{c}{a})^2 \\geq 2(\\frac{b\\plus{}c}{a}\\plus{}\\frac{c\\plus{}a}{b}\\plus{}\\frac{a\\plus{}b}{c})$\r\n$ \\Longleftrightarrow \\sum(\\frac{a}{b}\\minus{}1)^2 \\geq 0$"
}
{
"Tag": [
"linear algebra",
"matrix"
],
"Problem": "va mai propun o problema interesanta inainte sa plec :)\r\nfie triunghiul $ABC$ si $M,N,P$ sunt centrele cercurilor exinscrise, opuse lui $A,B,C$ respectiv. perpendicularele din $N$ pe $AB$ si din $P$ pe $AC$ se intersecteaza in $A_{1}$.\r\naratati ca $AA_{1}\\cap BB_{1}\\cap CC_{1}\\neq \\emptyset$.",
"Solution_1": "O s\u0103 rezolv urm\u0103toarea problem\u0103:\r\n\r\n[color=blue]Fie $ABC$ un triunghi, $O$ centrul cercului circumscris, $U,V,W$ simetricele lui $O$ fa\u0163\u0103 de $BC,CA,AB$ \u015fi $D,E,F$ picioarele \u00een\u0103l\u0163imilor din $A,B,C$ pe $BC,CA,AB$.\n\nDemonstra\u0163i c\u0103 $DU, EV, FW$ sunt concurente.\n[/color]\r\n\r\n[hide=\"Rezolvare cu numere complexe\"]\nConsider\u0103m cercul circumscris $\\triangle ABC$ drept cercul-unitate din planul complex. Vom nota cu litere mici afixele corespunz\u0103toare punctelor.\nAvem $|a|=|b|=|c|=1$, $d = \\frac{a(a+b+c)-bc}{2a}$, $e = \\frac{b(a+b+c)-ca}{2b}$, $f = \\frac{c(a+b+c)-ab}{2c}$, $u = b+c$, $v = c+a$, $w = a+b$ ($d,e,f$ necesit\u0103 destul de multe calcule). De asemenea, $\\overline d = \\frac{-a^{2}+\\sum bc}{2abc}, \\, \\overline u = \\frac{b+c}{bc}$ etc.\n\nEcua\u0163ia dreptei $DU$ este \\[| \\begin{matrix}d & \\overline d & 1 \\\\ u & \\overline u & 1 \\\\ z & \\overline z & 1 \\end{matrix}| = 0\\] \\[\\, \\Longleftrightarrow \\, \\, ( 2bc-a^{2}-\\sum bc ) z+bc (-a^{2}+\\sum bc ) \\overline z = 2 ( bc-a^{2}) ( b+c ) .\\] Celelalte dou\u0103 ecua\u0163ii ies prin permut\u0103ri ciclice.\nFaptul c\u0103 cele trei drepte sunt concurente e echivalent cu faptul c\u0103 cele trei ecua\u0163ii formeaz\u0103 un sistem ce are solu\u0163ie nenul\u0103.\nAltfel zis, noi trebuie s\u0103 demonstr\u0103m c\u0103\n\n$| \\begin{matrix}2bc-a^{2}-( ab+bc+ca ) & bc (-a^{2}+ab+bc+ca ) & 2 ( b+c ) ( bc-a^{2}) \\\\ 2ca-b^{2}-( ab+bc+ca) & ca (-b^{2}+ab+bc+ca ) & 2 ( c+a ) ( ca-b^{2}) \\\\ 2ab-c^{2}-( ab+bc+ca) & ab (-c^{2}+ab+bc+ca ) & 2 ( a+b ) ( ab-c^{2}) \\end{matrix}| = 0$\n\n(sc\u0103dem prima linie din celelalte dou\u0103)\n\n$\\, \\Longleftrightarrow \\, \\, 2 | \\begin{matrix}2bc-a^{2}-( ab+bc+ca ) & bc (-a^{2}+ab+bc+ca ) & ( b+c ) ( bc-a^{2}) \\\\ 2c(a-b)+(a-b)(a+b) & abc(a-b)+c (ab+bc+ca)(a-b) & (a-b) ( c^{2}+2 \\sum bc-ab ) \\\\ 2b(a-c)+(a-c)(a+c) & abc(a-c)+b (ab+bc+ca)(a-c) & (a-c) ( b^{2}+2 \\sum bc-ac ) \\end{matrix}| = 0 $\n\n$\\, \\Longleftrightarrow \\, \\, 2(a-b)(a-c) | \\begin{matrix}2bc-a^{2}-( ab+bc+ca ) & bc (-a^{2}+ab+bc+ca ) & ( b+c ) ( bc-a^{2}) \\\\ a+b+2c & abc+c (ab+bc+ca) & c^{2}+2 \\sum bc-ab \\\\ a+2b+c & abc+b (ab+bc+ca) & b^{2}+2 \\sum bc-ac \\end{matrix}| = 0 $\n\n(sc\u0103dem a doua linie din a treia)\n\n$\\, \\Longleftrightarrow \\, \\, 2(a-b)(a-c) | \\begin{matrix}2bc-a^{2}-( ab+bc+ca ) & bc (-a^{2}+ab+bc+ca ) & ( b+c ) ( bc-a^{2}) \\\\ a+b+2c & abc+c (ab+bc+ca) & c^{2}+2 \\sum bc-ab \\\\ b-c & (b-c) (ab+bc+ca) & (b-c)(b+c)+a(b-c) \\end{matrix}| = 0 $\n\n$\\, \\Longleftrightarrow \\, \\, 2(a-b)(a-c)(b-c) | \\begin{matrix}2bc-a^{2}-( ab+bc+ca ) & bc (-a^{2}+ab+bc+ca ) & ( b+c ) ( bc-a^{2}) \\\\ a+b+2c & abc+c (ab+bc+ca) & c^{2}+2 \\sum bc-ab \\\\ 1 & ab+bc+ca & a+b+c \\end{matrix}| = 0 $\n\n(\u00eenmul\u0163im ultima linie cu $-bc$ \u015fi o adun\u0103m la prima, apoi o \u00eenmul\u0163im cu $-c$ \u015fi o adun\u0103m la a doua)\n\n$\\, \\Longleftrightarrow \\, \\, 2(a-b)(a-c)(b-c) | \\begin{matrix}bc-a^{2}-( ab+bc+ca ) &-a^{2}bc &-abc-( b+c ) a^{2}) \\\\ a+b+c & abc & ab+bc+ca \\\\ 1 & ab+bc+ca & a+b+c \\end{matrix}| = 0 $\n\n(\u00eenmul\u0163im a doua linie cu $a$ \u015fi o adun\u0103m la prima)\n\n$\\, \\Longleftrightarrow \\, \\, 2(a-b)(a-c)(b-c) | \\begin{matrix}0 & 0 & 0 \\\\ a+b+c & abc & ab+bc+ca \\\\ 1 & ab+bc+ca & a+b+c \\end{matrix}| = 0, $\n\nrela\u0163ie care este evident adev\u0103rat\u0103.\n\nAcum mai r\u0103m\u00e2ne de verificat faptul c\u0103 punctul de intersec\u0163ie este finit. Minorul din st\u00e2nga sus iese $-(a-b) (a+b)^{2}c (a+c)(b+c) \\neq 0$ (pe \u0103sta nu l-am mai calculat de m\u00e2n\u0103, dar am \u00eencredere \u00een soft-ul \u0103sta).\n\n$\\blacksquare$\n[/hide]\r\n\r\n\r\n$ABC$ este triunghiul ortic al lui $MNP$, iar $A_{1},B_{1},C_{1}$ sunt simetricele centrului cercului circumscris triunghiului $MNP$ fa\u0163\u0103 de $NP,PM,MN$. Aceste afirmatii reies dintr-un [url=http://www.mathlinks.ro/Forum/viewtopic.php?highlight=long&t=50503]exerci\u0163iu pe care l-am postat mai demult[/url].\r\n\r\nAcum trebuie doar s\u0103 aplic\u0103m ce-am demonstrat mai sus triunghiului $MNP$ \u015fi iese problema."
}
{
"Tag": [
"modular arithmetic"
],
"Problem": "Suppose $ p$ is a prime gretaer than 3. Find all pairs $ (a,b)$ of integers satisfying the equation $ a^2\\plus{}3ab\\plus{}2p(a\\plus{}b)\\plus{}p^2\\equal{}0$",
"Solution_1": "We have $ p \\equal{} \\minus{} (a \\plus{} b)\\pm\\sqrt {b(b \\minus{} a)}$. So $ b(b \\minus{} a)$ must be a perfect square.\r\n\r\nIf $ b \\equal{} 0,\\ a \\equal{} \\minus{} p$.\r\n\r\nSuppose $ b\\neq0$. Then* any prime that divides $ b$ must divide $ a$; hence $ b\\mid a$, say $ bc \\equal{} a$. $ \\therefore\\ b(b \\minus{} a) \\equal{} b^2(1 \\minus{} c)\\ \\Rightarrow\\ 1 \\minus{} c$ is a perfect square $ \\Rightarrow\\ 1 \\minus{} c \\equal{} d^2\\ \\Rightarrow\\ a \\equal{} b(1 \\minus{} d^2)$.\r\n\r\nSo we have $ p \\equal{} \\minus{} [b(1 \\minus{} d^2) \\plus{} b]\\pm bd \\equal{} b(d^2\\pm d \\minus{} 2)\\ \\Rightarrow\\ b\\mid p\\ \\Rightarrow\\ b \\equal{} \\pm1$ or $ b \\equal{} \\pm p$. In any case, $ d^2\\pm d \\minus{} 2 \\equal{} d(d\\pm1) \\minus{} 2$ is an even number, contradicting the fact that $ p$ is an odd prime.\r\n\r\n*See my next post below.",
"Solution_2": "Why is $ b|a$ ? For instance, $ 4(4\\minus{}3)$ is a perfect square, but $ 4\\not|3$.",
"Solution_3": "[quote=\"bertram\"]Why is $ b|a$ ? For instance, $ 4(4 \\minus{} 3)$ is a perfect square, but $ 4\\not|3$.[/quote]\r\nIt seems I have carelessly missed some things there. :blush: What I said would be true if both $ b$ and $ b \\minus{} a$ are greater than 1 (in absolute value). I\u2019ve checked the cases $ b \\equal{} \\pm1$ and found that they are impossible. However, $ a \\minus{} b \\equal{} \\pm1$ is possible: e.g. $ ( \\minus{} 3, \\minus{} 4)$ for $ p \\equal{} 5$.",
"Solution_4": "[quote=\"Jane_Bennet\"]...So it appears that $ ( \\minus{} p,0)$ is the only solution.[/quote]\r\n\r\n$ p\\equal{}7$, $ a\\equal{}\\minus{}5$ and $ b\\equal{}4$ works. So you're wrong.\r\n\r\nThe equation may be written $ (a\\plus{}b\\plus{}p)^2\\equal{}b(b\\minus{}a)$\r\n\r\nSo $ b(b\\minus{}a)$ is a perfect square and we have $ b\\equal{}ku^2$ and $ a\\equal{}ku^2\\minus{}kv^2$ fore some $ k,u,v$.\r\n\r\nThen $ a\\plus{}b\\plus{}p\\equal{}\\epsilon_0kuv$, which may be written $ 2ku^2\\minus{}kv^2\\plus{}p\\equal{}\\epsilon_0kuv$ (where $ \\epsilon_0$ is $ \\minus{}1$ or $ \\plus{}1$)\r\n\r\nSo $ p\\equal{}kv^2\\minus{}2ku^2\\plus{}\\epsilon_0kuv$ $ \\equal{}k(v\\minus{}\\epsilon_0u)(v\\plus{}2\\epsilon_0u)$\r\n\r\nSince $ p$ is prime, we have three cases :\r\n$ 1)$ \r\n$ k\\equal{}\\epsilon_1$\r\n$ v\\minus{}\\epsilon_0u\\equal{}\\epsilon_2$\r\n$ p\\equal{}\\epsilon_0\\epsilon_1\\epsilon_2(3u\\plus{}\\epsilon_0\\epsilon_2)$\r\nThis is equivalent to :\r\n$ k\\equal{}\\epsilon_1$\r\n$ v\\minus{}\\epsilon_0u\\equal{}\\epsilon_0\\epsilon_1$\r\n$ p\\equal{}3u\\plus{}\\epsilon_1$\r\nAnd so :\r\n$ p\\equal{}3u\\plus{}\\epsilon_1$\r\n$ a\\equal{}\\minus{}2u\\minus{}\\epsilon_1$\r\n$ b\\equal{}\\epsilon_1u^2$\r\n\r\n$ 2)$\r\n$ k\\equal{}\\epsilon_1$\r\n$ v\\plus{}2\\epsilon_0u\\equal{}\\epsilon_2$\r\n$ p\\equal{}\\minus{}\\epsilon_0\\epsilon_1\\epsilon_2(3u\\minus{}\\epsilon_0\\epsilon_2)$\r\nThis is equivalent to :\r\n$ k\\equal{}\\epsilon_1$\r\n$ v\\equal{}2\\epsilon_1\\epsilon_2u\\plus{}\\epsilon_2$\r\n$ p\\equal{}3u\\plus{}\\epsilon_1$\r\n\r\nAnd so :\r\n$ p\\equal{}3u\\plus{}\\epsilon_1$\r\n$ a\\equal{}\\minus{}\\epsilon_1(3u^2\\plus{}1)\\minus{}4u$\r\n$ b\\equal{}\\epsilon_1u^2$\r\n\r\n$ 3)$\r\n$ v\\minus{}\\epsilon_0u\\equal{}\\epsilon_1$\r\n$ v\\plus{}2\\epsilon_0u\\equal{}\\epsilon_2$\r\n$ p\\equal{}\\epsilon_1\\epsilon_2k$\r\n\r\nWhich gives : \r\n$ p\\equal{}k$\r\n$ a\\equal{}\\minus{}k$\r\n$ b\\equal{}0$\r\n\r\n*******************************\r\nAs a conclusion \r\n\r\nFor any prime $ p$ we have the solution $ (\\minus{}p,0)$\r\n\r\nFor any prime $ p\\equal{}1\\pmod{3}$, we also have the two solutions :\r\n\r\n$ (\\minus{}\\frac{2p\\plus{}1}{3},\\frac{(p\\minus{}1)^2}{9})$\r\n\r\n$ (\\minus{}\\frac{p(p\\plus{}2)}{3},\\frac{(p\\minus{}1)^2}{9})$\r\n\r\nFor any prime $ p\\equal{}2\\pmod{3}$, we also have the two solutions :\r\n\r\n$ (\\minus{}\\frac{2p\\minus{}1}{3},\\minus{}\\frac{(p\\plus{}1)^2}{9})$\r\n\r\n$ (\\frac{p(p\\minus{}2)}{3},\\minus{}\\frac{(p\\plus{}1)^2}{9})$"
}
{
"Tag": [
"videos"
],
"Problem": "In countdown round (which i'm pretty sure that im going to make), what is the format? they choose top ten (right?) and then what? how many coundtowns do they do? is it elimination or is it something else? :huh:",
"Solution_1": "[quote=\"Arvind_sn\"]In countdown round (which i'm pretty sure that im going to make), what is the format? they choose top ten (right?) and then what? how many coundtowns do they do? is it elimination or is it something else? :huh:[/quote]\r\nIt depends on your state.\r\nI know that in Georgia they have all of the top ten people come down, and the put the questions up. The whoever buzzes in first and gets the question right, gets a point. It's best out of 80.\r\n\r\nIn some states, like mine, it's official.",
"Solution_2": "They choose the top 10, then they put #9 against #10, and the winner of that faces #8, and the winner of that faces #7, and so on.\r\nThey ask questions, and the first person to get 2/3 right goes on.",
"Solution_3": "ok, so now i have a 6/17 chance of making nationals...not good. :lol:",
"Solution_4": "whatever you have i have 1/10 of not good...\r\nsee you at state\r\n\r\nEDIT: since it isn't official if someone \"had a coughing contest\" or started say umm to much they wouldn't be dqed or something would they?",
"Solution_5": "If you get first on the written test, you are guarenteed nationals right? Because CountDown is really not my thing... :(",
"Solution_6": "if it's not official.",
"Solution_7": "If the countdown is official, it works exactly the way the [b]old[/b] countdown at nationals worked. Top ten come up, pair 10 vs. 9, winner vs. 8, and so on. It's best of 3 with a sudden-death tiebreaker until the last three matches, which are first to three points. Official countdown rounds everywhere except nationals are required to follow this format. Also, if you're top 3 on written, you're guaranteed a spot at nats.\r\n\r\nIf the countdown is unofficial, anything goes. I think Florida, or at least some Florida chapters, have a system in which every student participates, and teams get a score. California, or some chapters here, have a giant tournament for all the students. Minnesota used to do a single-elim tournament for the top 16, but changed it a few years back to top 12--except the four students who get byes are picked randomly to avoid divulging the results before the awards ceremony later. The format of an unofficial countdown is limited only by the availability of questions and the imagination of the organizers.",
"Solution_8": "[quote=\"Sly Si\"]I think Florida, or at least some Florida chapters, have a system in which every student participates, and teams get a score.[/quote]\r\nNo, in all the chapters in Florida it's official. (unless I'm mistaken)\r\n\r\nThe round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.",
"Solution_9": "In Georgia, the countdown round is only for a trophy. The state team is chosen based on the points from the Sprint and Target round combined.\r\n\r\nSee you tomorrow!\r\nFulton",
"Solution_10": "[quote=\"diophantient\"][quote=\"Sly Si\"]I think Florida, or at least some Florida chapters, have a system in which every student participates, and teams get a score.[/quote]\nNo, in all the chapters in Florida it's official. (unless I'm mistaken)\n\nThe round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.[/quote]\r\n\r\nIt's official in everything, including state. The problem is the questions are not that great, but oh well. It does not affect team standings or anything of that nature. It's just an extra, fun round where the winning team gets a trophy.",
"Solution_11": "[quote=\"gtf\"]In Georgia, the countdown round is only for a trophy. The state team is chosen based on the points from the Sprint and Target round combined.\n\nSee you tomorrow!\nFulton[/quote]\r\ndo you know if you can be dqed; and if so can that mean if you did make it to nats you can't go?",
"Solution_12": "I guess that it would depend on what you are DQed for. If it was for cheating, for instance, I would assume that you would be DQed from the whole competition and thus any nationals spot would be forfeited. If was for something stupid during the Countdown round, I would hope that you would still have your spot.\r\n\r\nWhat types of DQs are possible?\r\n\r\nThis is all my opinion only. Any official decision would need to come from the organizers.\r\n\r\nFulton",
"Solution_13": "[quote=\"Arvind_sn\"]ok, so now i have a 6/17 chance of making nationals...not good. :lol:[/quote]\r\n\r\nWhere did you get that number from?",
"Solution_14": "at the actual nationals competition, the format is top twelve people.\r\ntop four get a bye, and the other 8 are paired such that the sum of their ranks is 17. Next round is like an eight-person single elimination",
"Solution_15": "I rock at the countdown round, but I never do good enough on the written round to qualify. :)",
"Solution_16": "[quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\r\nHow does the ciphering round work?",
"Solution_17": "[quote=\"gtf\"]I guess that it would depend on what you are DQed for. If it was for cheating, for instance, I would assume that you would be DQed from the whole competition and thus any nationals spot would be forfeited. If was for something stupid during the Countdown round, I would hope that you would still have your spot.\n\nWhat types of DQs are possible?\n\nThis is all my opinion only. Any official decision would need to come from the organizers.\n\nFulton[/quote]\r\n\r\nI am not certain about this, but I remember hearing a while back that if a student has to take the state competition on an alternate day (usually for religious reasons) and can't attend on the day the countdown is held, they forfeit any countdown round matches that they would have participated in. Thus if countdown is official, they drop one spot automatically, but no more. This is why official countdowns are required to follow that format, so a student unable to attend could still make nationals by finishing in the top 3 in the written round.",
"Solution_18": "at r state comp we have a tourney bracket, but its unofficial, so we really just dont care",
"Solution_19": "[quote=\"magixter\"][quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\nHow does the ciphering round work?[/quote]\r\nOne person from each school comes up at a time. A question comes up and everyone tries to answer it under a minute. You have this kind of machine that you can type in your answer. After a minute has passed, they see who got it in first. The first person that got it right gets 5 points. The next person that gets it right gets 4 points and so on. Theres four rounds for each of your team mates to do. And each round has 4 questions.",
"Solution_20": "Is the round official, for trophies, or just for fun?",
"Solution_21": "[quote=\"ckck\"][quote=\"magixter\"][quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\nHow does the ciphering round work?[/quote]\nOne person from each school comes up at a time. A question comes up and everyone tries to answer it under a minute. You have this kind of machine that you can type in your answer. After a minute has passed, they see who got it in first. The first person that got it right gets 5 points. The next person that gets it right gets 4 points and so on. Theres four rounds for each of your team mates to do. And each round has 4 questions.[/quote]\r\nWhen do they do the ciphering round?\r\n\r\nIn response to magxiter's sig: NEVER!!!!",
"Solution_22": "[quote=\"erdogankerem123\"][/quote][quote=\"ckck\"][quote=\"magixter\"][quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\nHow does the ciphering round work?[/quote]\nOne person from each school comes up at a time. A question comes up and everyone tries to answer it under a minute. You have this kind of machine that you can type in your answer. After a minute has passed, they see who got it in first. The first person that got it right gets 5 points. The next person that gets it right gets 4 points and so on. Theres four rounds for each of your team mates to do. And each round has 4 questions.[/quote][quote=\"erdogankerem123\"]\nWhen do they do the ciphering round?\n\nIn response to magxiter's sig: NEVER!!!![/quote]\r\n\r\nThe ciphering round is only done in some states, like Florida.",
"Solution_23": "[quote=\"erdogankerem123\"][/quote][quote=\"ckck\"][quote=\"magixter\"][quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\nHow does the ciphering round work?[/quote]\nOne person from each school comes up at a time. A question comes up and everyone tries to answer it under a minute. You have this kind of machine that you can type in your answer. After a minute has passed, they see who got it in first. The first person that got it right gets 5 points. The next person that gets it right gets 4 points and so on. Theres four rounds for each of your team mates to do. And each round has 4 questions.[/quote][quote=\"erdogankerem123\"]\nWhen do they do the ciphering round?\n\nIn response to magxiter's sig: NEVER!!!![/quote]\r\nYea, I think Florida is the only one that does it because they let parents video tape the whole thing. But, its right before countdown.",
"Solution_24": "[quote=\"ckck\"][/quote][quote=\"erdogankerem123\"][/quote][quote=\"ckck\"][quote=\"magixter\"][quote]The round you are thinking of, is ciphering, which has questions similar to countdown. It's very very cool.\n[/quote]\nHow does the ciphering round work?[/quote]\nOne person from each school comes up at a time. A question comes up and everyone tries to answer it under a minute. You have this kind of machine that you can type in your answer. After a minute has passed, they see who got it in first. The first person that got it right gets 5 points. The next person that gets it right gets 4 points and so on. Theres four rounds for each of your team mates to do. And each round has 4 questions.[/quote][quote=\"erdogankerem123\"]\nWhen do they do the ciphering round?\n\nIn response to magxiter's sig: NEVER!!!![/quote][quote=\"ckck\"]\nYea, I think Florida is the only one that does it because they let parents video tape the whole thing. But, its right before countdown.[/quote]\r\nLouisiana and this one other state have similar rounds, but the problems are different."
}
{
"Tag": [
"function",
"trigonometry"
],
"Problem": "[size=150]The horizontal axis is marked in degrees. Find a+b+c where one period of the sinusoid \ny=a+bsinc(x-45^0 )begins at (45^0 ,11) and ends at (135^0 ,11) with horizontal tangent lines at (67.5^0 ,16) and (112.5^0 , 6) \n\n\na) 21.5 \nb)21 \nc)20.5 \nD) 20 \ne)none of these [/size]\r\n\r\n\r\nThanks!",
"Solution_1": "[hide]\n$y=a+b\\sin(c(x-45))$\n\nsubstituting $(45,11)$, we get\n$11=a+b\\sin(c(45-45))=a\\\\ a=11$\n\nthe period of the function is 90 degrees, so c must be 4, because 4*90=360.\n$y=11+b\\sin(4(x-45))$\n\nsubstituting $(\\frac{135}{2},16)$, we get\n$16=11+b\\sin{(4(\\frac{135}{2}-45))}\\\\ 5=b\\sin{90}\\\\ b=5\\\\ a+b+c=11+4+5=20$\n\n[/hide]",
"Solution_2": ":lol: :lol:"
}
{
"Tag": [
"Euler",
"number theory",
"prime numbers",
"number theory unsolved"
],
"Problem": "We define $\\delta(n)$ is the sum of divisors of $n\\in N$.Prove that $n$ is producting of twin prime numbers if and only if $\\varphi(n) \\delta(n)=(n-3)(n+1)$.($\\varphi$ is the Euler function).\r\n [hide]when $p,q$ are prime and $|p-q|<2$ then we say $p,q$ are twin prime numbers.[/hide]",
"Solution_1": "I think you meant p,q are twin primes iff |p-q|=2, right? :)",
"Solution_2": "This seems hard at first, but when you really tackle it, it's not that bad:\r\n\r\nLet P(n) denote the number of distinct primes that divide n. We try to prove that that equation is true iff P(n)=2 and that the two primes which divide n are twin primes.\r\n\r\nThe \"if\" part is trivial (just write n=pq, and calculate). So I'll only prove the \"only if\" part:\r\n\r\nFirst when P(n)=1, it's easy to see that (n-3)(n-1) > Phi(n)*Epsilon?(n). \r\nNow if P(n)>=3, we write n=abk, where a,b are primes and k a positive integer. Then write k=Product (i=1 to s) p_i^q_i. Then let A = Epsilon(n)*Phi(n)= Epsilon(ab)*Epsilon(k) * Phi(ab)*Phi(k)= (a^2-1)(b^2-1)*Product (i=1 to s) (p_i^q_i+1 - 1) < (a^2-1)(b^2-1)*k*Product (i=1 to s). Now you can show that A < n^2 - 2n - 3 for k > 1. And we are done. :D"
}
{
"Tag": [
"function"
],
"Problem": "Can anyone help me with these two problems:\r\n\r\nConsider the function defined on 1..10 by the following: S(10) = 1; S(n) = n + S(n + 1) for 0 < n < 10 What is the value of S(7)? \r\n\r\nConsider the language on the alphabet {a,b} defined by the grammar: \r\n ->ab | a \r\na. What is the shortest string in this language? \r\nb. List all strings of length 7.",
"Solution_1": "Er... are they really that simple or is there something I've missed here?\r\n\r\n1:\r\nS(10) = 1\r\nS(9) = 9 + S(10) = 10\r\nS(8) = 8 + 10 = 18\r\nS(7) = 7 + 18 = 25\r\n\r\n2:\r\nYour language quite obviously looks like this:\r\n{a, aba, ababa, abababa, ...}\r\n\r\n(a) The shortest string is \"a\".\r\n(b) The only 7-letter string is \"abababa\"."
}
{
"Tag": [
"geometry proposed",
"geometry"
],
"Problem": "prove that for every triangle $ABC$,we have:\r\n$\\frac{3\\sqrt{3}}{2cos\\frac{A}{2}cos\\frac{B}{2}cos\\frac{C}{2}}+8sin\\frac{A}{2}sin\\frac{B}{2}sin\\frac{C}{2}\\geq 5$",
"Solution_1": "it's equivalent with : $p^{2}\\leq \\frac{108R^{4}}{(5R-2r)^{2}}$, and if you really want to end it use $p^{2}\\leq 4R^{2}+4Rr+3r^{2}\\leq \\frac{108R^{4}}{(5R-2r)^{2}}$.",
"Solution_2": "$4R^{2}+4Rr+3r^{2}\\leq \\frac{108R^{4}}{(5R-2r)^{2}}$ $\\Longleftrightarrow$ $(5t-2)^{2}\\left(4t^{2}+4t+3\\right)\\le 108t^{4}$, where $t=\\frac{R}{r}\\ge 2$ $\\Longleftrightarrow$\r\n\r\n$(t-2)(8t^{3}-4t^{2}-19t+6)\\ge 0$ $\\Longleftrightarrow$ $(t-2)\\left[2t^{2}(t-2)+5t\\left(t^{2}-4\\right)+t^{3}+t+6\\right]\\ge 0$, what is truly for any $t\\ge 2$.\r\n\r\n[b]Remark.[/b] $8t^{3}-4t^{2}-19t+6=(t-2)(8t^{2}+12t+5)+16\\ge 16>0$ for any $t\\ge 2$.",
"Solution_3": "I don't understand,Can you post full solution?"
}
{
"Tag": [
"group theory",
"abstract algebra",
"superior algebra",
"superior algebra unsolved"
],
"Problem": "It's sort of nice:\r\n\r\nLet K be a field (commutative) and $t\\in K\\setminus \\{0,1\\}$. We write K as $A \\cup B$, where $A \\cap B= \\emptyset$ . Show that there are distinct $a, b, c$ in the same set (all in A or all in B) s.t. $a=tb+(1-t)c$.\r\n\r\n[amfulger] :Some moderator edits. I could not have deleted the posts you can read downwards. They are much too funny to be deleted :D",
"Solution_1": "Are there other conditions ? because it's trivial for any a=b=c !!",
"Solution_2": "The problem may ask that the numbers are distinct, but then for t=0 or 1 we need a=b or a=c. So what is the complete problem? :D",
"Solution_3": "Sorry! I didn't think about that. Take t=/=0 or 1, and a, b, c distinct.",
"Solution_4": "Are you sure that's possible to write K as A \\cup B,where A \\cap B = \\emptyset ?I don't think so.((K,+)can't be written as a reunion of 2 distinct subsets).",
"Solution_5": "You're thinking about writing a group as the reunion of 2 proper SUBGROUPS, while I'm talking about writing it as a reunion of 2 SUBSETS. Nobody said anything about the 2 subsets being rings or groups or anything like that. \r\n\r\nFor example: K={0} \\cup K* and {0} \\cap K*= \\emptyset .",
"Solution_6": "They aren't subcorps, just subsets!",
"Solution_7": "Are you sure that's possible to write K as A \\cup B,where A \\cap B = \\emptyset ?I don't think so.((K,+)can't be written as a reunion of 2 distinct subsets).",
"Solution_8": "Don't you think that we can partition any set in 2 subsets?",
"Solution_9": "I think some posts need to be erased :).",
"Solution_10": "Very interesting problem. Visualizing this just for $ \\mathbb{C}$ would give you the solution. I'm not going into details.\r\n\r\nLet $ a_{1},a_{2},a_{3}\\in A$ be different points. If the problem's claim was't right then $ b_{1}\\equal{}ta_{1}\\plus{}(1\\minus{}t)a_{2}\\in B, b_{2}\\equal{}ta_{3}\\plus{}(1\\minus{}t)a_{2}\\in B, b_{3}\\equal{}ta_{1}\\plus{}(1\\minus{}t)a_{3}\\in B$. applying this process again:\r\n\r\n$ a_{5}\\equal{}\\dfrac{b_{3}\\minus{}(1\\minus{}t)b_{2}}{t}, a_{4}\\equal{}\\dfrac{b_{1}\\minus{}(1\\minus{}t)b_{2}}{t}$ where $ a_{4}, a_{5}$ are the solutions of $ tx\\plus{}(1\\minus{}t)b_{2}\\equal{}b_{1},\\: tx\\plus{}(1\\minus{}t)b_{2}\\equal{}b_{3}$ and therefore they must lie in $ A$ otherwise the problem would be true again. But now it's easy to check that \r\n\r\n$ ta_{5}\\plus{}(1\\minus{}t)a_{4}\\equal{}a_{1}$ where $ a_{1},a_{4},a_{5}\\in A$ also they are different ( easy to check ).\r\n\r\nsome little details must be filled out here. How would we guarantee that $ b_{1},b_{2},b_{3}$ are different? how would we know if either of $ A$ or $ B$ has three elements ( and in fact the problem for fields with cardinality less than 6 is not true....actually I am not sure, I haven't checked it. ) and once you know one of these sets has more than three elements choosing those intial elements is very cruial so that $ b_{1},b_{2},b_{3}$ be different. by our simple selection ( $ a_{1},a_{2},a_{3}$ are different ) it is obvious that $ b_{1}\\neq b_{3}, b_{1}\\neq b_{2}$ but for having $ b_{2}\\neq b_{3}$ we should have $ t(2a_{3}\\minus{}a_{2}\\minus{}a_{1})\\neq a_{2}\\minus{}a_{3}$ and we can have this if those intial elements have this property that $ 2a_{3}\\minus{}a_{2}\\minus{}a_{1}\\equal{}0$ which would be again this problem for $ t\\equal{}1/2$ ( when char($ K$)$ \\neq 2$ ). but this time when $ t\\equal{}1/2$ you can check that in fact $ b_{1},b_{2},b_{3}$ are different and you can find such $ a_{1},a_{2},a_{3}\\in A$ that $ 2a_{3}\\minus{}a_{2}\\minus{}a_{1}\\equal{}0$ and then starting with these elements for each $ t$, $ b_{i}$s would be different."
}
{
"Tag": [
"complementary counting"
],
"Problem": "The dial on the lock of a school locker has 40 numbers. If no combination uses the same number all three times, how many three-number combinations are possible on the lock?",
"Solution_1": "Since it is only the 3rd number that has to be different, the number of combinations is $ 40\\times40\\times39\\equal{}\\fbox{62400}$.",
"Solution_2": "No. Your way does not allow combinations of 1-2-x, where x is any number. It only allows 39 of them, when all 40 should be allowed.\r\n\r\nInstead, we use complementary counting to obtain $ 40^3\\minus{}40\\equal{}\\boxed{63960}$."
}
{
"Tag": [
"trigonometry",
"geometry proposed",
"geometry"
],
"Problem": "[color=indigo]Given $\\triangle ABC$, point $D$ lies on $\\overline{AC}$ such that $\\overline{AB}=\\overline{CD}$. If $\\measuredangle{\\text{ }}A = 2x,{\\text{ }}\\measuredangle{\\text{ }}C = 3x,{\\text{ }}\\measuredangle{\\text{ }}ABD = x$, find $\\measuredangle\\ x$[/color]",
"Solution_1": "You won't receive from me more than a trigonometric solution: :) \r\nFrom the sin law we have:\r\n1) $\\frac{BC}{\\sin{3x}}=\\frac{CD}{\\sin{6x}}$\r\n2) $\\frac{BC}{\\sin{2x}}=\\frac{AB}{\\sin{3x}}$.\r\nBy dividing them we obtain: $\\sin{2x}\\cdot \\sin{6x}=\\sin^{2}{3x}$\r\n\r\n$\\iff 2\\sin{2x}\\cos{3x}=\\sin{3x}\\iff \\sin{5x}-\\sin{3x}=\\sin{x}$\r\n\r\n$\\iff 2\\sin{x}\\cos{4x}=\\sin{x}\\iff \\cos{4x}=\\frac{1}{2}$\r\n\r\n$\\iff x = \\frac{\\pi}{12}$.",
"Solution_2": ":rotfl:\r\n\r\nIs not bad proving different ways, but at least, each problem that I've posted, has an elementary solution.\r\n\r\nGreetings :lol: \r\n\r\nAnd the answer is correct.",
"Solution_3": "[color=darkblue][b][u]Proof I (synthetical).[/u][/b] Observe that $m(\\widehat{BDC})=3x$, i.e. $BD=BC$. Denote the middlepoints $E$, $F$ of the segments $[CD]$, $[AB]$ respectively, i.e. $FA=FB=ED=EC$. Thus, $EB\\perp EA$, i.e. $EF=\\frac{1}{2}\\cdot AB$. Therefore, $EF=\\frac{1}{2}\\cdot CD$, i.e. $FD\\perp FC$. From $FE=FA$ obtain $m(\\widehat{FEA})=2x$ and from $EF=EC$ obtain $m(\\widehat{DCF})=x$. In conclusion, $\\widehat{DCF}\\equiv\\widehat{DBF}$, i.e. the quadrilateral $FBCD$ is cyclically. Thus, $\\widehat{DBC}\\equiv\\widehat{DFC}$, i.e. $BD\\perp BC$. From $BD=BC$ obtain $3x=\\frac{\\pi}{4}$, i.e. $x=\\frac{\\pi}{12}$.\n\n[b]Remark.[/b] $FA=FB=FE=ED=EC=EB$. The triangle $ABC$ is \"[u][b]sliced[/b][/u]\" in three triangles :\n\none $F$- isosceles $\\triangle AFE$, one $E$- right isosceles $\\triangle BEC$ and one equilateral $\\triangle BEF$.\n\n[b][u]Proof II (trigonometrical).[/u][/b] $\\sin 2x=\\sin \\widehat{BAE}=\\frac{EB}{AB}=\\frac{EB}{DC}=\\frac{1}{2}\\cdot\\frac{EB}{ED}=\\frac{1}{2}\\cdot\\tan\\widehat{BDE}=\\frac{1}{2}\\cdot\\tan 3x$ $\\implies$ $2\\sin 2x\\cos 3x=\\sin 3x$ $\\implies$\n\n$\\sin 5x-\\sin x=\\sin 3x$ $\\implies$ $\\sin 5x-\\sin 3x=\\sin x$ $\\implies$ $2\\sin x\\cos 4x=\\sin x$ $\\implies$ $\\cos 4x=\\frac{1}{2}$ $\\implies$ $x=\\frac{\\pi}{12}$.[/color]"
}
{
"Tag": [
"linear algebra",
"linear algebra unsolved"
],
"Problem": "Let $ A$ be an $ n\\times n$ real symmetric matrix. Show that\r\n\r\n\\[ \\mathbb{R}^{n}\\equal{}\\ker\\left(A\\right) \\plus{}\\textrm{Im}\\left(A\\right).\\]",
"Solution_1": "Hint: you can make use of the fact that $ A\\equal{}U^TDU$ for orthogonal $ U$ and diagonal $ D.$ Can you solve the original problem for $ A$ diagonal?",
"Solution_2": "$ \\equal{}\\equal{}0$ if $ x\\in{ker(A)}$."
}
{
"Tag": [
"function",
"inequalities unsolved",
"inequalities"
],
"Problem": "Let $ x \\geq 1$. Find the minimum value :\r\n$ A \\equal{}$ $ \\sqrt {x \\minus{} 1}$ + $ \\sqrt {2x^{2} \\minus{} 5x \\plus{} 7}$",
"Solution_1": "hello, our function has no maximum since\r\n$ A'(x)\\equal{}\\frac{1}{2\\sqrt{x\\minus{}1}}\\plus{}\\frac{4x\\minus{}5}{2\\sqrt{2x^2\\minus{}5x\\plus{}7}}$ is allways positive for all real $ x$ with $ x>1$.\r\nSonnhard.",
"Solution_2": "[quote=\"Nam Luu\"]Let $ x \\geq 1$. Find the maximum value :\n$ A \\equal{}$ $ \\sqrt {x \\minus{} 1}$ + $ \\sqrt {2x^{2} \\minus{} 5x \\plus{} 7}$[/quote]\r\n\r\n\r\n\r\nmaximum value???",
"Solution_3": "sorry I have edited the problem"
}
{
"Tag": [],
"Problem": "The mean of two positive single-digit integers is obtained by placing a decimal point between the two numbers. What is the sum of the two numbers?",
"Solution_1": "$ \\frac{x\\plus{}y}{2} \\equal{} \\frac{10x \\plus{} y}{10}$\r\n\r\n$ 20x \\plus{} 2y \\equal{} 10x \\plus{} 10 y$\r\n\r\n$ 5x \\equal{} 4y$\r\n\r\nSo $ x \\equal{} 4$ and $ y \\equal{} 5$, therefore $ 4 \\plus{} 5 \\equal{} \\boxed{9}$."
}
{
"Tag": [
"geometry unsolved",
"geometry"
],
"Problem": "On my printscreen I have this image. Does anyone has a solution? Thanks a lot !",
"Solution_1": "Solution:\r\nLet $ PY$ be second tangent line to the given circle,$ Y$ is the point of tangency.\r\nThen $ \\widehat{YAB} \\equal{} \\widehat{PYB} \\equal{} \\frac {\\widehat{BOY}}{2} \\equal{} \\widehat{POB}$,hence $ \\boxed{AY\\|PO}$ and quadrilateral $ CYDB$ is [b]harmonic[/b].\r\nConsider pencil $ A(C,Y,D,B)$ and its intersection with $ PO$,we obtain new harmonic four:$ (EF,O\\infty)$(since $ AY\\|PO$),hence $ OE \\equal{} OF$.\r\n$ \\blacksquare$",
"Solution_2": "Let M be the orthogonal projections of O to CD, N is the point that MN \u2225 AD,MN meet AB at N,CN meet DA at K.\r\nwe can aquire N,B,M,C are concircle O,M,B,P are concircle CN \u2225 OP CN/OE=AN/AO=NK/OF but we have CN=NK\r\nThus we are done"
}
{
"Tag": [
"calculus",
"derivative",
"integration",
"real analysis",
"calculus computations"
],
"Problem": "I have Kaplan's Prep book but I\"m wondering how I can find a good book for AP Calc BC that can guarantee me five. I'll have to study on my own so.. Thank you!",
"Solution_1": "Speaking from general terms, I'm going to assume you know the material well and just want to know it *really well*.\r\n\r\nIn that case, [url=http://www.artofproblemsolving.com/Books/AoPS_B_Item.php?page_id=8]you need a book of tough problems.[/url]\r\n\r\nThat really should be plenty, but if you want more problems maybe the second book on [url=http://www.skylit.com/beprepared-calc.html]this[/url] page is the way to go.",
"Solution_2": "Well at my high school curriculm for math it says that students take two calculus courses for their junior and senior year. AP Calc AB and BC or whatever...whats actually the difference between AB and BC or whatever? Is one half about applications of the calculus and one about the basic concepts?\r\n\r\nsorry if this is off topic. i have a book that i would highly recommend but i don't think it falls under the \"BC\" category",
"Solution_3": "Calc AB covers differentiation and an introduction to integration; Calc BC also covers techniques/applications of integration and sequences/series.\r\n\r\nBack when I was learning calculus, I found Thomas/Finney and Larson/Hostetler to be both readable and precise texts. That was 20 years ago, though, so the current editions may be less good. I would avoid most of the currently-popular textbooks (such as Stewart) for self-study for the AP.",
"Solution_4": "Calculus BC covers more material than Calculus AB and also can involve problems that are harder versions of Calculus AB material. The exact differences should be covered (if vaguely and generally) on the College Board website.\r\n\r\nI took AP Calculus last year; I practically had to teach myself the BC material, but my teacher gave me previous multiple choice and free response questions. The best way to prepare for the AP Calculus Exam (AB or BC) (or any AP exam, really) is to do past problems.",
"Solution_5": "I agree with Zakary about doing past problems. Look into Barron's AP. Barron's always gives challenging problems.",
"Solution_6": "Do the prep books use ACTUAL past problems? Or do they make up ones that are supposed to be similar to them? My advice is to get one's hands directly on problems from the exams...",
"Solution_7": "I remember that Barron's AP Physics used variants of past problems. I don't know about other books. But I can confidently state that if you master Barron's then you will get a 5 on the exam no sweat. However Barron's is much tougher than the exam but nonetheless prepares you for it in a way that learning Rudin does not prepare you to take the AP Calc exam. So it's still good."
}
{
"Tag": [],
"Problem": "What does the word $arbitrary$ mean as a mathematical term?",
"Solution_1": "random??",
"Solution_2": "is it supposed to be spelled \"arbitrary\"? Or not, since it is spelled as \"arbitraty\" in both the topic title and your post. If it's arbitrary, shobber is right, it means random. If it actually is \"arbitraty\" then I have no idea, and it's probably a typo.",
"Solution_3": "Sorry, its arbitrary.",
"Solution_4": "e.g. take $arbitrary$ triangle $ABC$\r\n\r\ni.e. $arbitrary$ means $any$",
"Solution_5": "[quote=\"random_mathematician\"]e.g. take $arbitrary$ triangle $ABC$\n\ni.e. $arbitrary$ means $any$[/quote]\r\n\r\nSo your name would be arbitrary_mathematician?"
}
{
"Tag": [
"number theory solved",
"number theory"
],
"Problem": "PROVE OR DISPROVE\r\n\r\nFor all positive integer $k$ there exist prime $p$ such that $2,...,k$ are not perfect square modulo $p$",
"Solution_1": "[quote=\"vnmathboy\"]PROVE OR DISPROVE\n\nFor all positive integer $k$ there exist prime $p$ such that $1,2,...,k$ are not perfect square modulo $p$[/quote]\r\n\r\nwhat about $1$ man ? :D . If $a^2\\leq k\\leq (a+1)^2$ then $a^2$ is square modulo $p$ for may be all $p$ which $(p,a)=1$ :D",
"Solution_2": ":blush: sorry!"
}
{
"Tag": [
"inequalities unsolved",
"inequalities"
],
"Problem": "What number is the most biggest $ (1\\minus{}\\frac{1}{365})(1\\minus{}\\frac{2}{365})\\cdot...\\cdot(1\\minus{}\\frac{25}{365})$ and $ \\frac{1}{2}$",
"Solution_1": "[quote=\"maksmaks24\"]What number is the most biggest $ (1 \\minus{} \\frac {1}{365})(1 \\minus{} \\frac {2}{365})\\cdot...\\cdot(1 \\minus{} \\frac {25}{365})$ and $ \\frac {1}{2}$[/quote]\r\nUse $ \\ln(1\\minus{}x)\\leq \\minus{}x$ for all $ x\\in[0,1).$",
"Solution_2": "You know to what this calculation hints! It is the well known that $ 25$ people have a better chance than odds on for two of them to share a same birthday (day of the year only, not the year too)."
}
{
"Tag": [
"Harvard",
"college",
"Yale",
"search"
],
"Problem": "Are any of you guys eagle scouts?\r\n\r\nI'm probably going to be one my end of freshman year (or beginning of sophmore), and have a question\r\n\r\nI have asked this to many people, but here it goes\r\n\r\nHow much will eagle help you get into college?\r\n\r\nHeres the summary of answers I have...\r\n\r\nWell...it really depends. If you're going to Westpoint or Norwich, it matters a lot, while if it is Harvard or Yale, it doesn't matter to much.\r\n\r\nWhat are your ideas?\r\n\r\n-jorian",
"Solution_1": "I'm an eagle scout\r\n\r\nif you're getting eagle as early as sophomore year then you haven't really had the full experience (just saying!)\r\n\r\nno clue how it matters at all in college",
"Solution_2": "National figures on how many Eagle Scouts there are each year would help show how rare a distinction that is. My educated guess is that it probably hardly helps at all for any Ivy League college--so go for Eagle it it's personally important to you, but not so much for a college admission boost.",
"Solution_3": "ill probably feel bad in 40+ years if i don't get eagle...i'll say, \"why didn't i get it...i was soooo stupid\"\r\n\r\nand the reason the numbers for eagles are so low is because most people aren't in it :P\r\n\r\ni met someone at norwich university who got eagle at 13 :D\r\n\r\n-jorian",
"Solution_4": "It hasn't been established yet that the number of Eagle scouts is low. Maybe it's high: what kind of number of new Eagle scouts each year does Scouting see in the United States?",
"Solution_5": "[quote=\"tokenadult\"]It hasn't been established yet that the number of Eagle scouts is low. Maybe it's high: what kind of number of new Eagle scouts each year does Scouting see in the United States?[/quote]\r\n\r\nA random Google search turned up this:\r\n\r\nhttp://usscouts.org/usscouts/eagle/numbers.html",
"Solution_6": "Anything that 30,000 boys a year do is not an Ivy-meaningful distinction. It may still be a good idea to finish the Eagle requirements, but that wouldn't impress the admission officers you are trying to impress.",
"Solution_7": "no, i was just asking how much certain schools care\r\n\r\nthe eagle rate for kids who try in my troop is like 85% :)\r\n\r\n-jorian",
"Solution_8": "[quote=\"jhredsox\"]\ni met someone at norwich university who got eagle at 13 :D\n[/quote]\r\n\r\nthat's not anything noteworthy. The rate at which people get eagle varies wildly between troops; in my troop, no matter how good you are you won't get eagle before 16; even then, most don't get it until 17 or 18 (I got it at 18). \r\n\r\nThere's a good point to that, too; a 13 year old eagle scout is nothing compared to a 16 or 17 or 18 year old eagle scout. He just does NOT have the experience to justify getting the eagle distinction. Eagle is so much more than just doing the requirements in the book -- it's going through the entire scouting experience and actually being a part of it long enough to participate and enjoy all the stuff it has to offer. A guy who rushes through it and gets it at age 13 just does not have that at all.\r\n\r\nUnfortunately, there's no real way of determining who should and shouldn't be Eagle -- Of course, the Eagle board is fairly standardized, but they'll push anywhere with a decent plan through. The real barrier is the Scoutmaster, and there are plenty of troops even in my city that are eagle factories and will push out 13-14 year old eagle scouts ALL the time. It's unfortunate there isn't some sort of way to discourage pushing people through just for the badge.",
"Solution_9": "ill prob get it at 15\r\n\r\n-jorian",
"Solution_10": "[quote=\"jhredsox\"]ill prob get it at 15\n\n-jorian[/quote]\r\n\r\nok!!!!",
"Solution_11": "[quote=\"jhredsox\"]ill prob get it at 15\n\n-jorian[/quote]\r\n\r\nQuite impressive! I got it at 17.\r\n\r\nAnd as for college admissions, I don't think it matters that much in itself, but the fact that you remained in scouts that long would probably show dedication towards an activity, as one of my interviewers noted. Of course, it didn't hurt that he was an Eagle himself."
}
{
"Tag": [
"trigonometry",
"function"
],
"Problem": "Here is a problem my friend and I came up with. We had fun solving this, using two different methods. Care to try?\r\n\r\nAssume we have a block of mass $ m$ attached to a spring with spring constant $ k$. This system is attached to a point such that the system forms a pendulum. That is, we have two oscillations: the block on the spring, and the motion of the pendulum. Find an expression for the kinetic energy of the system.\r\n\r\nFeel free to introduce new variables for length of the pendulum, angle formed with the vertical,...",
"Solution_1": "I am depressed that no one can get this... :(",
"Solution_2": "Well here are the equations of motion:\r\n\r\n\\[ \\ddot r=\\dfrac{kl_0}{m}+g\\cos\\varphi -\\dfrac{k}{m}r\\]\r\n\\[ \\ddot \\varphi=-\\dfrac{g}{r}\\sin\\varphi\\]\r\n\r\nHere $ r$ is the distance of the mass from the point(of course this changes), and $ \\varphi$ is the angle with the vertical, $ l_0$ is the length of the spring, pendulum when the spring is loose.\r\n\r\nNow at this moment I dont see how you guys managed to get the kinetic energy of the system (The problem is finding $ \\dot r^2+r^2\\dot\\varphi^2$), since this system of coupled DE-s doesnt look solvable like this, in fact, we cant even solve the simple pendulum equation for the time function.\r\nEven if we try solving it for some other function $ E(\\varphi,r,t)$ I dont think we can. Maybe there is some simple method, which I cant see at the moment.\r\n\r\nIf we take the simplifing assumptions that, the angle is small, that is $ \\sin\\varphi \\approx \\varphi$, and $ \\cos\\varphi \\approx 1$, then we get the following:\r\n \r\n\\[ \\ddot r=\\dfrac{kl_0}{m}+g -\\dfrac{k}{m}r\\]\r\n\\[ \\ddot \\varphi=-\\dfrac{g}{r}\\varphi\\]\r\n\r\nNow this looks much simpler. in fact we can solve the first equation for $ r(t)$. Its just a simple nonhomogenous second order so:\r\n\\[ r(t)=A\\cdot \\cos\\left(t\\sqrt\\dfrac{k}{m}\\right)+B\\cdot \\sin\\left(t\\sqrt\\dfrac{k}{m}\\right)+l_0+\\dfrac{m^2g}{k}\\]\r\nplugging this into the other equation:\r\n\\[ \\ddot \\varphi +\\dfrac{g}{A\\cdot \\cos\\left(t\\sqrt\\dfrac{k}{m}\\right)+B\\cdot \\sin\\left(t\\sqrt\\dfrac{k}{m}\\right)+l_0+\\dfrac{m^2g}{k}}\\varphi=0\\]\r\n\r\nTheoretically this is solvable for $ \\varphi(t)$, and then we can get the kinetic energy, but just now I dont see if it can solved analiticaly...\r\n\r\nI would be very curious how you guys managed, to solve, please post your solution JRav :D"
}
{
"Tag": [
"function",
"absolute value",
"number theory",
"relatively prime",
"combinatorics unsolved",
"combinatorics"
],
"Problem": "A is a set of all integer sequence and \\emptyset :A--->Z is function satisfying these condition:\r\ni)for any s,t \\in A, we have \\emptyset (s+t)= \\emptyset (s)+ \\emptyset (t) {s+t is a sequence which ith member of it is ith member of s+ith member of t}\r\nii)for any n \\emptyset (0,0,....,0,1,0,0,0,...)=0 {1 is placed on nth place}\r\na)prove \\emptyset (1,2,4,8,16,....)=0\r\nB)PROVE \\emptyset =0",
"Solution_1": "Let {Ok} be the zero sequence (all terms are equal to zero). Since for any {ak} the sum {ak}+{Ok}={ak} holds,\r\n\\emptyset({ak}+{Ok})=\\emptyset({ak}) and \\emptyset({ak}+{Ok})=\\emptyset({ak})+\\emptyset ({Ok}) by property i, then\r\n\\emptyset({ak})=\\emptyset({ak})+\\emptyset ({Ok}) implies that \\emptyset({Ok})=0.\r\n\r\nLemma 1: For any sequence {sk} where every term is zero except at k, in which sk=m, then \\emptyset({sk})=0.\r\nProof: Suppose m\\neq0. \\emptyset({sk})=\\emptyset(0,...,0,m,0,...)=m*\\emptyset(0,...,0,1,0,...)=m*0=0 by properties i and ii, where m is at the nth place.\r\n\\emptyset(0,...,0,m,0,...)=m*\\emptyset(0,...,0,1,0,...) by property i, you might want to show this mentally, it's easy.\r\nIf m=0, {sk}={Ok} so that \\emptyset({sk})=\\emptyset({Ok})=0.\r\nEnd of lemma.\r\n\r\na) For the sequence {ak}={2k-1}, define the sequence {sk}=(0,...,0,2k-1,0,...) for all k\\geq1 so that {ak}=\\sumj=1\\infty {sj}\r\nThen \\emptyset({ak})=\\emptyset(\\sumj=1\\infty {sj})=\\sumj=1\\infty \\emptyset({sj})=0 by our lemma.\r\n\r\nb) As in part a, define the sequence {sk}=(0,...,0,ak,0,...} for all k\\geq1 so that {ak}=\\sumj=1\\infty {sj}\r\nThen \\emptyset({ak})=\\emptyset(\\sumj=1\\infty {sj})=\\sumj=1\\infty \\emptyset({sj})=0 by our lemma.\r\n\r\nEdit PS: Sorry guys, I am still finding a solution.",
"Solution_2": "[quote=\"3X.lich\"]\na) For the sequence {ak}={2k-1}, let the subsequence {skn}=(0,...,0,2n-1,0,...} for all n\\geq1. Then \n\\emptyset({ak})=\\sumj=1\\infty (2j-1)*(\\emptyset({skj}))=0 by our lemma.\n\nb) As in part a, define the subsequence {skn}=(0,...,0,an,0,...} for all n\\geq1. Then\n\\emptyset({ak})=\\sumj=1\\infty (aj)*(\\emptyset({skj}))=0 by our lemma.[/quote]\r\nIt is serious mistake!\r\nWe can't state that \\emptyset({ak})=\\sumj=1\\infty (aj)*(\\emptyset({skj})).\r\n3X.lich! Are you student?\r\n\r\nFor a).\r\n\\emptyset(1,2,4,8,16,...)= \\emptyset(1,0,0,...)+\\emptyset(0,2,4,8,...)= 0+2*\\emptyset(0,1,2,4,8,...)= 2(\\emptyset(0,1,0,0,...)+\\emptyset(0,0,2,4,8,...))= 4*\\emptyset(0,0,1,2,4,8,...)= and so on. Thus number \\emptyset(1,2,4,8,...) is divisible by any power of 2 ==> \\emptyset(1,2,4,8,...)=0 (!!!).",
"Solution_3": ":( My bad, looks like my post needs tons of corrections :( \r\nTo Myth: yeah I'm a student, so I can't make mistakes?\r\n\r\n\\emptyset(1,2,4,8,16,...)= \\emptyset(1,0,0,...)+\\emptyset(0,2,4,8,...)= \\emptyset(1,0,0,...)+2*\\emptyset(0,1,2,4,8,...)= \\emptyset(1,0,0,...)+2*\\emptyset(0,1,0,0,...)+2*\\emptyset(0,0,2,4,8,...)= \\emptyset(1,0,0,...)+2*\\emptyset(0,1,0,0,...)+4*\\emptyset(0,0,1,2,4,8,...)= \r\n=\\sumj=1n (2j-1)\\emptyset(0,...,0,1,0,...)+(2n-1)\\emptyset(0,...,0,1,2,4,8,...),\r\nwhere 1 is at jth place for the first \\emptyset and where 1,2,4,8,... starts at nth place for the second \\emptyset.\r\n\r\nPS: I think I understand where the mistake is. I have assumed in the first place that \\emptyset=0 and working with infinities are pretty dangerous. Anybody have any suggestions, like correcting my post, or an alternate approach? :D",
"Solution_4": "I am not sure that my solution is correct.\r\n\r\n\\emptyset (a(1),a(2),a(3),...)=\\emptyset(0,a(2),a(3),...)=\\emptyset (0,0,a(3),...)\r\nSo we have\r\n\r\n\\emptyset(1*a(1),2*a(2),3*a(3),...,n*a(n),...)=\r\n\\emptyset(a(1),a(2),a(3),...)+\\emptyset(0,a(2),a(3),...)+\\emptyset(0,0,a(3),...)+...\r\n\r\nIf we have \\emptyset(a(1),a(2),a(3),...) is not 0 so \\emptyset(1*a(1),2*a(2),3*a(3),...,n*a(n),...) is infinite and is not integer",
"Solution_5": "The same mistake as 3X.lich has done.",
"Solution_6": "Does anybody have solution?",
"Solution_7": "Myth's idea is the key:\r\n\r\nIf (a_n) is an unbounded sequence we have :\r\n \\emptyset ((a(1),a(2),..,a(n),..)) = a(k) * \\emptyset((0,0,(k-1 times),1,a(k+1),...) so a(k) divides \\emptyset(a) for each k. Therefore it is divisible by arbitrarily large numbers whose absolute value will exceed its absolute value if it's not zero. So \\emptyset(a)=0.\r\n\r\nIf (b_n) is a bounded sequence then c_n = b_n + 2^n and d_n=2^n are unbounded and thus \\emptyset(c_n-d_n)=\\emptyset(c_n)-\\emptyset(d_n)=\\emptyset(b_n)=0.",
"Solution_8": "What can I say? You are wrong, iandrei!",
"Solution_9": "What's wrong about it? I'm only using the fact that if a number is divisible by arbitrarily large numbers then it must be 0 (a divides b ===> |a|<=|b|). Please explain.",
"Solution_10": "[quote=\"iandrei\"]\n\\emptyset ((a(1),a(2),..,a(n),..)) = a(k) * \\emptyset((0,0,(k-1 times),1,a(k+1),...)[/quote]\r\nThis is wrong identity.",
"Solution_11": "Ooops, sorry :blush: My mistake!",
"Solution_12": "Here is an answer (thanks to VK ;))\r\nLet x be an element of Z^N. As 2 and 3 are relatively prime integers, there exists\r\ntwo sequences y and z s.t. $x(n)=2^ny(n)+3^nz(n).$\r\n$\\phi(2^ny(n))=\\phi(3^nz(n))=0$ with the same technique as above, so $\\phi(x)=0.$\r\nNice problem.",
"Solution_13": "[quote=\"Gillian_Seed\"]Here is an answer (thanks to VK ;))\nLet x be an element of Z^N. As 2 and 3 are relatively prime integers, there exists\ntwo sequences y and z s.t. $x(n)=2^ny(n)+3^nz(n).$\n$\\phi(2^ny(n))=\\phi(3^nz(n))=0$ with the same technique as above, so $\\phi(x)=0.$\nNice problem.[/quote]\r\nIndeed, very nice problem! Who is VK?"
}
{
"Tag": [
"inequalities",
"function",
"algebra unsolved",
"algebra"
],
"Problem": "Find all quadruples of real numbers [tex]x,y,z,t[/tex] which satisfy : [tex]x+y+z+t=x^7+y^7+z^7+t^7=0[/tex]",
"Solution_1": "i think the only two solutions are $(0,0,0,0)$ and $(-0,-0,-0,-0)$ ;)\r\n\r\n[hide=\"heres my proof\"]h\u00f6lders inequality says [by [url=http://mathworld.wolfram.com/HoeldersInequalities.html]wolfram[/url]]:\n$\\sum_{k=1}^n |a_kb_k| \\le \\left(\\sum_{k=1}^n |a_k|^p\\right)^{\\frac1p}\\cdot \\left(\\sum_{k=1}^n |b_k|^q\\right)^{\\frac1q}$\nwith $\\frac1p + \\frac1q = 1$\n\nnow setting $a= (x,y,z,t)$, $b = (1,1,1,1)$ and $p=7$, $q=\\frac76$\nwe get:\n$0=x+y+z+t \\le 4\\cdot \\sqrt[7]{x^7+y^7+z^7+t^7} = 0$\n\nnow wolfram says that equality is when $|b_k|=c|a_k|^{p-1}$\nthat would mean that they are all the same because $b_k=1$ for all $k$\nhence $0=4\\cdot x= 4\\cdot x^7$\n\nso $x=\\pm 0$ and we arrive at and only at the given solutions[/hide]",
"Solution_2": "Answer: $\\left( x,y,z,t\\right) =\\left( a,b,-a,-b\\right) $ with $a,b\\in R$ (and permutations).\r\n\r\nDiscarding the trivial solution $\\left( 0,0,0,0\\right) ,$ observe that $x,y,z,t$ cannot have the same sign. Thus, modulo some permutations, we have\r\nto analyze two cases:\r\n\r\n1. $x,y,z\\geq0\\geq t;$\r\n\r\n2. $x,y\\geq0\\geq z,t.$\r\n\r\nIn the first case, replacing $t$ by $-t,$ we have\r\n\\[\r\nx+y+z=t\r\n\\]\r\n\\[\r\nx^{7}+y^{7}+z^{7}=t^{7},\r\n\\]\r\nso\r\n\\[\r\n\\left( \\frac{x}{x+y+z}\\right) ^{7}+\\left( \\frac{y}{x+y+z}\\right)\r\n^{7}+\\left( \\frac{z}{x+y+z}\\right) ^{7}=1.\r\n\\]\r\nBut\r\n\\[\r\n\\left( \\frac{x}{x+y+z}\\right) ^{7}\\leq\\frac{x}{x+y+z},\r\n\\]\r\n\\[\r\n\\left( \\frac{y}{x+y+z}\\right) ^{7}\\leq\\frac{y}{x+y+z},\r\n\\]\r\n\\[\r\n\\left( \\frac{z}{x+y+z}\\right) ^{7}\\leq\\frac{z}{x+y+z},\r\n\\]\r\ntherefore the above equality holds only if one of the fractions equals $1$ and the other two equal $0.$ We obtain the solution $\\left( 1,0,0,-1\\right) $\r\n(and permutations).\r\n\r\nIn the second case, we have, for the nonnegative numbers $x,y,z,t:$\r\n\\[\r\nx+y=z+t,\r\n\\]\r\n\\[\r\nx^{7}+y^{7}=z^{7}+t^{7}.\r\n\\]\r\nSince the function $f:[0,+\\infty)\\rightarrow R,$ $f\\left( x\\right) =x^{7}$ is convex, the above equalities can hold only if $\\{x,y\\}=\\{z,t\\}$. We obtain the solution $\\left( a,b,-a,-b\\right) $ (and permutations).",
"Solution_3": "your solution quadruple seems obvious, and is without a doubt correct.\r\nbut then were is the mistake in my proof ... ?\r\n\r\npeeta",
"Solution_4": "[quote=\"peeta\"]but then were is the mistake in my proof ... ?[/quote]\r\nHolder's inequality holds only for positive numbers, that is your mistake.",
"Solution_5": "Indeed, you have:\r\n$0=|x+y+z+t| \\le 4\\cdot \\sqrt[7]{|x|^7+|y|^7+|z|^7+|t|^7} \\ne 0$\r\n\r\ntherefore..."
}
{
"Tag": [
"MATHCOUNTS"
],
"Problem": "Is it just more or did it say Greg Gauthier was an 8th grader in the 2003 Mathcounts ESPN airing? Was that just a mistake or did he actually stay back to redo mathcounts?",
"Solution_1": "I'm 99% sure that he wasn't in 8th that year. Probably a mistake.",
"Solution_2": "Greg won MATHCOUNTS in 2004 as an 8th grader, he was 7th in 7th grade in 2003"
}
{
"Tag": [],
"Problem": "Where can i find solutions to HC verma (concepts of physics)?is there any site where they can be found or on any p2p application?if yes pls tell me and if possible send the link.[/b]",
"Solution_1": "hey you use HC verma too?? thats a cool book , isnt it ???i use it too :lol: which standard are you in ??\r\n\r\n\r\ni advise you to try resnick halliday krane, its cooooooooooooooooool too!\r\n\r\nummm, i dont know about the solutions, i usually ask them to friends if i dont get some problems..........",
"Solution_2": "How about asking those problems here?there are many people here willing to help you :lol: Feel free.",
"Solution_3": "[quote=\"eminem\"]\ni advise you to try resnick halliday krane, its cooooooooooooooooool too![/quote]\r\n :o :mad: :| :ninja: :wallbash: :wallbash_red: :agent: :bomb: Forget it yes u can ask ur questions here!",
"Solution_4": "ya resmick halliday is a gud book but i prefer using hc verma.i was having a few problems in solving a few problems of laws of motion.i cud not solve the monkey problems as well as the last 5 problems.if u cud help i wud b great.thanx",
"Solution_5": "if u could mention the chapter then we could help :)",
"Solution_6": "[quote=\"riddler\"][quote=\"eminem\"]\ni advise you to try resnick halliday krane, its cooooooooooooooooool too![/quote]\n :o :mad: :| :ninja: :wallbash: :wallbash_red: :agent: :bomb: Forget it yes u can ask ur questions here![/quote]\r\n\r\n :D :rotfl: You ain't the only person in the universe who knows it riddler.",
"Solution_7": "[quote=\"shadysaysurspammed\"][quote=\"riddler\"][quote=\"eminem\"]\ni advise you to try resnick halliday krane, its cooooooooooooooooool too![/quote]\n :o :mad: :| :ninja: :wallbash: :wallbash_red: :agent: :bomb: Forget it yes u can ask ur questions here![/quote]\n\n :D :rotfl: You ain't the only person in the universe who knows it riddler.[/quote]\r\nI know",
"Solution_8": "hey cool dude , about the monkey ones as you are saying in the laws of motion are you talking about the TENSION ones?, did you get that one, the one in which 2 guys(me and Shady ) are pulling out their friend(riddler , ahahahahahahhaha) who's stuck in a ditch???",
"Solution_9": "I don't think HC Verma solutions will be on a P2P server. It's mostly used in Indian high schools and I even doubt whether most of indian universities even have any P2P servers. But as eminem and shady and riddler... said above, you can alwyas discuss your problems here!!! [hide]( hey guys, no new posts?!)[/hide]",
"Solution_10": "no i didnt get the ditch question and the monkey question is the one where one monkey is hanging on the tail of the other.and as far as the solutions story goe they are available in the market i guess.but if u get them online whats better :P",
"Solution_11": "Which problem are you refering to? Please be specific!!!",
"Solution_12": "I NEED HELP FOR SUMS 57,25-b,c,d,68,48 from the 31st chapter - capacitors.\r\nI would really appreciate ur help.\r\n(P.s: Does ne one the the exact name of the book containing the solns to hc verma co i can't get it newhere..???)[/size]",
"Solution_13": "Yup,good ol days you just revived.\r\n\r\nAnyway,They are available here in the market by the name \"Solutions of concepts of physics\"Depends where you live",
"Solution_14": "could u tell me the author's name or the publication bcoz i asked for it in 2 bookshops in b'mbay but couldn't find it...or could u give me the add. of some book shop from where i could order it??...",
"Solution_15": "H C verma solutions are availble at \r\nhttp://www.vidyadrishti.com/H%20C%20Verma%20Solutions.html\r\nThe most imp point is that problems r solved by iitians.",
"Solution_16": "guys, this thread was made THREE YEARS AGO!!! :mad:"
}
{
"Tag": [
"search",
"number theory unsolved",
"number theory"
],
"Problem": "If $a,b$ $(a^2+b^2)/(ab-1)=q$ are integers, then $q=5$",
"Solution_1": "Posted before...\r\nBut do you really think that \"Nt\" as title says something to the readers\u00bf Use topic names that describe (parts of) the problem!",
"Solution_2": "Please can anyone give that page ...",
"Solution_3": "Tried \"a^2+b^2 ab-1\", with additionally the search restricted to Number Theory\u00bf"
}
{
"Tag": [],
"Problem": "Bobby stands 10 feet from Sam, and Eric stands 8 feet from Bobby. What is the least number of feet that Eric could be from Sam?",
"Solution_1": "Let's think about everything in comparision to Bobby. Sam is somewhere in a 10 foott radius circle from Bobby. Eric can now be anywhere in an 8 foot radius circle from Bobby. For Sam and Eric to be as close together as possible, they will be in the same exact same direction, thus making their distance from each other $ 10 \\minus{} 8 \\equal{} 2$ feet."
}
{
"Tag": [
"geometry",
"trapezoid",
"angle bisector"
],
"Problem": "ABCD is a trapezoid. AB is parallel to CD. If DA = AB + CD, and AE is the angle bisector of angle A, show that BE = EC.",
"Solution_1": "[hide]Instead I'll prove that if M is the midpoint of BC, then AM bisects angle BAD. Continue DM to hit AB at F, and notice that MCD and MBF are congruent. So AF=AB+BF=AB+CD=AD, meaning triangle DAF is isosceles, and so the median AM of that triangle is also the angle bisector.\n\nSince there is only one point of intersection between the angle bisector of angle A and segment BC, point M must be this point.[/hide]"
}
{
"Tag": [
"inequalities"
],
"Problem": "Let be $ a,b,c\\in (0,\\infty )$ such that $ a\\plus{}b\\plus{}c\\equal{}1$ . Prove that :\r\n\r\n$ a\\sqrt{a}\\plus{}b\\sqrt{b}\\plus{}c\\sqrt{c}\\ge \\sqrt{a^4\\plus{}b^4\\plus{}c^4\\plus{}24abc\\sqrt[3]{abc}}$",
"Solution_1": "[hide=\"Ugly\"]\n\nBy Holder\n\n$ (a \\plus{} b \\plus{} c)(a\\sqrt {a} \\plus{} b\\sqrt {b} \\plus{} c\\sqrt {c})(a\\sqrt {a} \\plus{} b\\sqrt {b} \\plus{} c\\sqrt {c})\\ge (a^{\\frac {4}{3}} \\plus{} b^{\\frac {4}{3}} \\plus{} c^{\\frac {4}{3}})^3$\n\nSo the inequality is reduced to\n\n$ (a^{\\frac {4}{3}} \\plus{} b^{\\frac {4}{3}} \\plus{} c^{\\frac {4}{3}})^3\\ge a^4 \\plus{} b^4 \\plus{} c^4 \\plus{} 24a^{\\frac {4}{3}}b^{\\frac {4}{3}}c^{\\frac {4}{3}}$\n\nWhich is obvious[/hide]",
"Solution_2": "Any symmetric sum of degree 4 with 24 terms is $ \\ge 24abc \\sqrt [3] {abc}$ by Muirhead (or AM-GM) since all the exponents on a,b,c are equal.\r\nSo squaring, we have to show\r\n$ (a^3 \\plus{} b^3 \\plus{} c^3 \\plus{} 2 \\sum a^{3/2}b^{3/2})(a \\plus{} b \\plus{} c) \\ge a^4 \\plus{} b^4 \\plus{} c^4 \\plus{} 24abc \\sqrt [3] {abc}$\r\nBut the LHS contains an $ a^4 \\plus{} b^4 \\plus{} c^4$ and 24 other terms of degree 4, so it is obviously true"
}
{
"Tag": [
"graph theory",
"Ramsey Theory",
"combinatorics theorems",
"combinatorics"
],
"Problem": "Let $ G$ be a graph with $ n$ vertices that does not contain a $ K_k$ as subgraph ($ K_k$ is a complete graph with $ k$ vertices).\r\nThen the maximum number of edges in G is equal to\r\n\\[ \\displaystyle M(n,k) = \\frac {k - 2}{k - 1}\\cdot \\frac {(n - r)(n + r)}{2} + \\frac {r(r - 1)}{2}\r\n\\]\r\nif\r\n\\[ \\displaystyle n = (k - 1)t + r,0\\leq r \\leq k - 2.\r\n\\]",
"Solution_1": "You mean Kk (a k complete subgraph?)",
"Solution_2": "Yes, Pascual.\r\n\r\nPierre.",
"Solution_3": "I read tht there is a consecuence of this theorem that is a generalization, it is a fotmula such that f(n,k,r) is the maximun number of edges you can draw and color in r colors in such a way that there is no complete k monocromatic subgraph. I think that furthermore, if A= R(r1,r1,r3,.....,rt) is the minimun number such that in any coloraTion of a Complete graph with A vertices with t colors there is a Kri for some i, then if we chose a quantity of edeges in a graph with n vertices which depends on n and A, and color this edges in t colors, then there is a Kri.....\r\n\r\nThe question is, I remember this formula was very similar to turans, does anyone nows it, maybe perhaps Pierre?",
"Solution_4": "The problem you are referring to smells really like Ramsey's theorem, which states that :\r\n\r\nLet $k,r \\geq 1$ and $m_1,...,m_r \\geq k$ be integers.\r\nThere is a integer $n_0$ such that, for each $n \\geq n_0$, no matter how each $k$-elements subsets of a set $X$ with $|X|$ are colored by $r$ colors (say $c_1,...,c_r$) there exists a subset $X_i$ of $X$ such that $|X_i| = m_i$ and each $k$-subset of $X_i$ has color $c_i$.\r\n\r\nIt may be useful to get the sense of this theorem by checking small values of $k,r, \\cdots $.\r\n\r\nIn another hand, for $k=2$ the $2$-subsets of a set $X$ can be seen as the edges of the complete graph whose vertices are the elements of $X$. Thus, Ramsey's theorem states that, given $m_1,...,m_r \\geq 2$, then for $n$ sufficientely large no matter how you color the edges of $K_n$ each with one from $r$ colors, there will be a complete monochromatic $K_{m_i}$.\r\n\r\nThe least integer $n_0$ which have the property described in the theorem is called the Ramsey number $R(m_1,...,m_r, k)$.\r\nNote that, the exact value of the Ramsey numbers are unknown, except for some exceptional and small values of $r,k,m_i$ \r\n\r\nIs that what you were talking about?\r\n\r\nPierre.",
"Solution_5": ":? \r\ncan you tel me please somthings about the proof\r\nand thank you johan \r\n:roll:",
"Solution_6": "Turans Theorem talks about having a certain number of vertices, and no clique larger than some K_n. I have an article where it says tells you the amount of vertices n and edges, and it says you have averege degree no more than 4. Thus Turans theorem implies that G has independence number at least n/5. I have no idea how they could use turans theorem, to calculate the independence number of the graph. Do you have any idea? Maybe i should explain more, i dont know.",
"Solution_7": "The version of Turan's theorem I am familiar with is actually stronger: it gives the unique structure for a graph on $n$ vertices with no $K_k$ subgraph to have a maximal number of edges. When one considers this version of the theorem it implies much more than the version in the original post."
}
{
"Tag": [
"geometry",
"quadratics",
"area of a triangle",
"Heron\\u0027s formula"
],
"Problem": "1. In Triangle ABC, AB=2, BC=3, AC=4. Determine the length of the altitude of BD. \r\n\r\n2. Factor $ a^4c^2\\minus{}b^4c^2\\plus{}d^2b^4\\minus{}d^2a^4$\r\n\r\nSo I have I have gotten this problem down to:\r\n$ (c\\plus{}d)(c\\minus{}d)(a^4\\minus{}b^4)$. I know $ (a^4\\minus{}b^4)\\equal{}(a\\plus{}b)(a\\minus{}b)(a^2\\plus{}b^2)$ but can someone please post an explanation.\r\n\r\n3. Rewrite $ x^4\\plus{}4$ as the product of two quadratics. The answer is below but how would you do this. \r\n [hide=\"answer\"]$ (x^2\\plus{}2x\\plus{}2)(x^2\\minus{}2x\\plus{}2)$[/hide]",
"Solution_1": "[hide=\"1\"]\nBy Heron's ($ s \\equal{} \\frac {a \\plus{} b \\plus{} c}2 \\equal{} \\frac {2 \\plus{} 3 \\plus{} 4}2 \\equal{} \\frac92$), the area is\n\n$ \\sqrt {\\frac92\\left(\\frac92 \\minus{} 2\\right)\\left(\\frac92 \\minus{} 3\\right)\\left(\\frac92 \\minus{} 4\\right)} \\equal{} \\sqrt {\\frac92\\cdot\\frac52\\cdot\\frac32\\cdot\\frac12} \\equal{} \\sqrt {\\frac {135}{16}} \\equal{} \\frac {3\\sqrt {15}}4$\n\n$ \\implies\\frac {3\\sqrt {15}}4 \\equal{} \\frac12\\cdot4\\cdot BD\\implies BD \\equal{} \\boxed{\\frac {3\\sqrt {15}}8}$.\n\nAlternatively,...\n\n [geogebra]3ea30d445718d6dc1e9a01805a5fb4b37f243ab9[/geogebra] \n\nWe have\n\n$ \\sqrt {4 \\minus{} x^2} \\plus{} \\sqrt {9 \\minus{} x^2} \\equal{} 4\\implies\\sqrt {9 \\minus{} x^2} \\equal{} 4 \\minus{} \\sqrt {4 \\minus{} x^2}$\n\n$ \\implies9 \\minus{} x^2 \\equal{} 16 \\plus{} 4 \\minus{} x^2 \\minus{} 8\\sqrt {4 \\minus{} x^2}\\implies\\frac {11}8 \\equal{} \\sqrt {4 \\minus{} x^2}$\n\n$ \\implies\\frac {121}{64} \\equal{} 4 \\minus{} x^2\\implies x^2 \\equal{} \\frac {135}{64},x > 0$\n\n$ \\implies x \\equal{} \\boxed{\\frac {3\\sqrt {15}}8}$.\n[/hide]\n[hide=\"2\"]\nRepeatedly grouping and using differences of squares, we get\n\n$ a^4c^2\\minus{}b^4c^2\\plus{}d^2b^4\\minus{}d^2a^4$\n\n$ \\equal{}c^2(a^4\\minus{}b^4)\\minus{}d^2(a^4\\minus{}b^4)$\n\n$ \\equal{}(c^2\\minus{}d^2)(a^4\\minus{}b^4)$\n\n$ \\equal{}(c\\minus{}d)(c\\plus{}d)(a^2\\minus{}b^2)(a^2\\plus{}b^2)$\n\n$ \\equal{}\\boxed{(c\\minus{}d)(c\\plus{}d)(a\\minus{}b)(a\\plus{}b)(a^2\\plus{}b^2)}$.\n[/hide]\n[hide=\"3\"]\nI will demonstrate two techniques.\n\n[hide=\"One way\"]\nFrom the problem statement, we can set up $ x^4\\plus{}4\\equal{}(x^2\\plus{}ax\\plus{}b)(x^2\\plus{}cx\\plus{}d)\\equal{}x^4\\plus{}(a\\plus{}c)x^3\\plus{}(b\\plus{}d\\plus{}ac)x^2\\plus{}(ad\\plus{}bc)x\\plus{}bd$, and pairing up coefficients, we get\n\n$ a\\plus{}c\\equal{}0\\implies a\\equal{}\\minus{}c$ (paring $ x^3$ coefficients),\n\n$ b\\plus{}d\\plus{}ac\\equal{}0$ (paring $ x^2$ coefficients),\n\n$ ad\\plus{}bc\\equal{}0$ (paring $ x$ coefficients), and\n\n$ bd\\equal{}4$ (pairing the constants).\n\nFrom this, we know that it must be of the form $ (x^2\\plus{}ax\\plus{}b)(x^2\\minus{}ax\\plus{}d)\\equal{}x^4\\plus{}4$, and it is not hard to guess the rest of it.\n\nHowever, this technique doesn't work well when the coefficients of the quadratics aren't integral.\n[/hide]\n[hide=\"Another Way\"]\nThis is a special case of the [url=http://www.artofproblemsolving.com/Wiki/index.php/Sophie_Germain_Identity]Sophie Germain Identity[/url].\n\nWe have\n\n$ x^4\\plus{}4\\equal{}x^4\\plus{}4x^2\\plus{}4\\minus{}4x^2\\equal{}(x^2\\plus{}2)^2\\minus{}(2x)^2\\equal{}\\boxed{(x^2\\minus{}2x\\plus{}2)(x^2\\plus{}2x\\plus{}2)}$.\n[/hide]\n[/hide]",
"Solution_2": "[hide=\"1\"] Using Heron's formula For $ \\triangle ABC$ we get the area of $ \\triangle ABC$, take that thing multi by 2 and divide by $ AC\\equal{}4$ get the solution.\n[/hide]\n[hide=\"2\"]\nI don't know what things to explain for you $ (c\\plus{}d)(c\\minus{}d)(a\\minus{}b)(a\\plus{}b)(a^2\\plus{}b^2)$ and that's it, that's right.\n[/hide]\n[hide=\"3\"]\nLet $ x^4\\plus{}4\\equal{}(x^2\\plus{}ax\\plus{}b)(x^2\\plus{}cx\\plus{}d)$\nYou have $ x^4\\plus{}cx^3\\plus{}dx^2\\plus{}ax^3\\plus{}acx^2\\plus{}adx\\plus{}bx^2\\plus{}bcx\\plus{}bd$\n$ x^4\\plus{}(c\\plus{}a)x^3\\plus{}(b\\plus{}d\\plus{}ca)x^2\\plus{}(ad\\plus{}bc)x\\plus{}bd\\equal{}x^4\\plus{}4$\nSo \n$ c\\plus{}a\\equal{}0$\n$ b\\plus{}d\\plus{}ca\\equal{}0$\n$ ad\\plus{}bc\\equal{}0$\n$ bd\\equal{}4$\n$ \\equal{}>(a,b,c,d)\\equal{}(2,2,\\minus{}2,2)$\nand you have $ x^4\\plus{}4\\equal{}(x^2\\plus{}2x\\plus{}2)(x^2\\minus{}2x\\plus{}2)$\n[/hide]",
"Solution_3": "Thanks so much for the answers and explanation. Hit my self on the head for the Heron's formula one :)"
}
{
"Tag": [
"trigonometry",
"function",
"geometry",
"angle bisector",
"trig identities",
"Law of Sines"
],
"Problem": "Let ABC be a triangle ,angle 124$, then $\\sqrt[3]{n}+1 < \\frac{\\sqrt{n}-1}{2}$, so by Bertrand's Conjecture (Theorem, whatever), $\\exists$ a prime $p$ between $\\lceil \\sqrt[3]{n} \\rceil$ and $\\lfloor \\sqrt{n} \\rfloor$. Thus $p^2 \\le n$ but $n < p^3$, so if $L = \\text{lcm}{1, 2, \\cdots, n}$, we have $p^2 || L$, and thus $3 | \\tau(L) \\Rightarrow \\tau(L) \\ne 2^k$. Now, for $n \\le 124$, if we can find a prime $p$ between $\\sqrt[3]{n}$ and $\\sqrt{n}$, we can still apply the above argument. For $25 \\le n \\le 124, p = 5$ works. For $9 \\le n \\le 25, p = 3$ works. For $4 \\le n \\le 7, p = 2$ works. So now we just manually check $n = 1, 2, 3, 8$. For $n = 1$, $\\tau(1) = 1 = 2^0$. For $n = 2$, $\\tau(2) = 2 = 2^1$. For $n = 3, \\tau(6) = 4 = 2^2$. For $n = 8, \\tau(840) = 32 = 2^5$."
}
{
"Tag": [
"LaTeX"
],
"Problem": "How do you do vertical multiplication and division in LaTeX?\r\n\r\n\r\nThanks! :D",
"Solution_1": "The [url=http://tug.ctan.org/cgi-bin/ctanPackageInformation.py?id=xlop]xlop package[/url] will do this for you and will also do the calculations. If you prefer a different layout use a [url=http://www.artofproblemsolving.com/Wiki/index.php/LaTeX:Layout#Tables]table[/url].",
"Solution_2": "For tables, here is an example code:\r\n\r\n[code]\\begin{tabular}{ccc}\\\\\n & 1 & 2\\\\\n\\times& 4 & 3\\\\\n\\hline\n & 3 & 6\\\\\n 4 & 8 & \\\\\n\\hline\n 5 & 1 & 6\n\\end{tabular}[/code]\r\n\r\n\\begin{tabular}{ccc}\\\\\r\n & 1 & 2\\\\\r\n\\times& 4 & 3\\\\\r\n\\hline\r\n & 3 & 6\\\\\r\n 4 & 8 & \\\\\r\n\\hline\r\n 5 & 1 & 6\r\n\\end{tabular}",
"Solution_3": "That is good, but \r\n[list=1][*]it will give 4 errors in a document! The problem is that \\times must be in maths mode so should be \\$\\times\\$ (AoPS treats maths and text modes differently so not important on the forum). Alternatively, use array instead of tabular and stick the whole lot in \\$...\\$. That's useful if otherwise you would have a lot of dollar signs to put in.\n[*]You can make the separation between the columns more or less \\setlength{\\tabcolsep}{[i]width[/i]}\n[*]the \\\\ on the first line after {ccc} isn't required.[/list]\r\nHere's an example with the tabcolsep changed to 0.5mm\r\n$ \\setlength{\\tabcolsep}{0.5mm}\r\n\\begin{array}{ccc}\r\n & 1 & 2\\\\ \r\n\\times & 4 & 3\\\\ \r\n\\hline \r\n & 3 & 6\\\\ \r\n 4 & 8 &\\\\ \r\n\\hline \r\n 5 & 1 & 6 \r\n\\end{array}$"
}
{
"Tag": [
"integration",
"LaTeX"
],
"Problem": "So, by default, when we use the two-dollar signs to wrap up our math stuff, they will be centered automatically. What should I do if I just want them to be, says, left hand align?",
"Solution_1": "Use one dollar sign on each side- this puts the math stuff inline, and you can simply add line breaks in if you want it to be at the beginning of a line.",
"Solution_2": "No, that's not what I want, because, for example, the $\\sum$, $\\int$ etc,.. will get smaller also. I want them to look as we use 2 dollars sign.",
"Solution_3": "Use \\displaystyle.\r\n\r\nYou don't need to do that on this forum- displaystyle is automatic here, as you can see by what you posted."
}
{
"Tag": [
"inequalities proposed",
"inequalities"
],
"Problem": "Given a triangle, and an inside point $ P$ of the triangle. Let $ a,b,c$ be the distances from $ P$ to the vortices and $ x,y,z$ be the distances from $ P$ to the sides of the triangle respectively.\r\nFind the minimum of :\r\n\r\n$ \\mbox{min}\\left ( \\frac{\\sqrt{a^2\\minus{}x^2}}{ax}\\plus{}\\frac{\\sqrt{b^2\\minus{}x^2}}{bx}\\plus{}\\frac{\\sqrt{a^2\\minus{}y^2}}{ay}\\plus{}\\frac{\\sqrt{c^2\\minus{}y^2}}{cy}\\plus{}\\frac{\\sqrt{b^2\\minus{}z^2}}{bz}\\plus{}\\frac{\\sqrt{c^2\\minus{}z^2}}{cz} \\right ) \\equal{}?$\r\nFor which $ x,y,z,a,b,c$ will it be minimal.\r\n \r\nI bumped into this in a physics problem, and I am looking for an elegant solution, not Lagrange method.\r\n\r\nIf someone has any idea for an elegant solution please share it :)",
"Solution_1": "your expression is homogenous of degree -1 so we can make it as small as we want. can you check it again.",
"Solution_2": "no thats it. I dont understand, why can it be as small as we want, since $ a,b,c,x,y,z$ arent independant variables, but are dependent, and are \"parts\" of a triangle, so they cant just take up any value. :huh:",
"Solution_3": "I mean that if you take bigger and bigger triangles similar to the initial one this expression goes to 0. and it is clearly non-negative so... :wink:",
"Solution_4": "Yes clearly if you could just take bigger and bigger triangles then it goes to 0. but then there would be no point of the problem. The lengths of the sides of the triangle are clearly fixed values, you cant just change them."
}
{
"Tag": [
"geometry",
"geometric transformation",
"reflection",
"probability",
"ratio",
"absolute value"
],
"Problem": "The numbers 1 to 9 are arranged randomly around the circumference of a circle. There is rotational indistinguishability and reflections are considered the same. The absolute value of the differences between adjacent terms around the circle are added together, for a total of 9 sums. What is the probability that this sum is a minimum?",
"Solution_1": "is it 2/8! ?",
"Solution_2": "[quote=\"aidan\"]The numbers 1 to 9 are arranged randomly around the circumference of a circle. There is rotational indistinguishability and reflections are considered the same. The absolute value of the differences between adjacent terms around the circle are added together, for a total of 9 sums. What is the probability that this sum is a minimum?[/quote]\r\n\r\nobviously in the sums, each number is going to be used twice\r\nso the sum will always be 90",
"Solution_3": "We're not talking about the sums of adjacent numbers.\r\n\r\n[hide=\"Solution\"]The minimum of the sums is $ 16$. Suppose a number $ n$ is next to numbers $ k$ and $ m$. WLOG $ k < m$.\n1. If $ k < n < m$, then in adding the absolute values of the differences $ n$ cancels out. $ |k \\minus{} n| \\plus{} |m \\minus{} n| \\equal{} m \\minus{} k$\n2. If $ n < k$ and $ n < m$, then $ n$ is subtracted twice in the resulting sum. $ |k \\minus{} n| \\plus{} |m \\minus{} n| \\equal{} m \\plus{} k \\minus{} 2n$\n3. If $ n > k$ and $ n > m$, then $ n$ is added twice in the resulting sum. $ |k \\minus{} n| \\plus{} |m \\minus{} n| \\equal{} 2n \\minus{} m \\minus{} k$\n$ 9$ will always be added twice and $ 1$ will always be subtracted twice. That's $ 16$ there. To get lower we have to have a number satisfying case 3 less than a number satisfying case 2. But if you partition the 9 number arrangement into monotonic subsequences, it's pretty clear that a number satisfying case 3 is always greater than another number satisfying case 2. (sorry for lack of rigor)\n\nThis also tells us the way to construct all arrangements achieving this is to place the $ 9$ and $ 1$ anywhere, then make sure that going from the $ 9$ to $ 1$ in either direction forms a monotonically decreasing sequence of numbers. Fix the position of the $ 9$ and ignore that we can do reflections (probability, being a ratio, will not change). There are $ 8!$ ways to arrange the other numbers now.\n\nSuppose, going clockwise, that there are $ m$ numbers between the $ 1$ and the $ 9$. Then $ 0 \\leq m \\leq 7$. The number of arrangements given $ m$ is $ {7 \\choose m}$, since there are $ 7$ numbers, and once we have decided which $ m$ are between the $ 9$ and $ 1$ going clockwise, there is only one way to arrange them. So the probability is\n\n$ \\frac{\\sum_{k\\equal{}0}^{7} {7 \\choose k}}{8!} \\equal{} \\frac{2^7}{8!} \\equal{} \\frac{1}{315}$\n\nunless I screwed up.[/hide]",
"Solution_4": "It is $ \\frac {1}{315}$ alright :)"
}
{
"Tag": [
"algebra",
"function",
"domain",
"system of equations",
"algebra unsolved"
],
"Problem": "Solve the system of equations \r\n$\\frac{1}{xy}=\\frac{x}{z}+1$ , $\\frac{1}{yz}=\\frac{y}{x}+1$ , $\\frac{1}{zx}=\\frac{z}{y}+1$\r\nin the domain of real numbers.",
"Solution_1": "http://www.mathlinks.ro/Forum/viewtopic.php?t=27577",
"Solution_2": "hello, eliminating $y,z$ we get for $x$:\n$\\left( 2\\,{x}^{2}-1 \\right) \\left( {x}^{12}+4\\,{x}^{10}+10\\,{x}^{8}+\n12\\,{x}^{6}+9\\,{x}^{4}+4\\,{x}^{2}+1 \\right)\n=0$\nSonnhard."
}
{
"Tag": [],
"Problem": "Two computers and a number of humans participated in a large round-robin chess tournament (i.e., every participant played every other participant exactly once). In every game, the winner of the game received one point, the loser zero. If a game ended in a draw, each player received half a point. At the end of the tournament, the sum of the two computers' scores was 38 points, and all of the human participants finished with the same total score. Describe (with proof) ALL POSSIBLE numbers of humans that could have participated in such a tournament.",
"Solution_1": "Let $ h$ be the numbers of humans with an equal final score of $ s$.\r\nThe amount of games played is $ (h \\plus{} 2)(h \\plus{} 1)/2$ so computing the sum of all the points given:\r\n$ 38 \\plus{} h \\times s \\equal{} (h \\plus{} 2)(h \\plus{} 1)/2$\r\n$ 76 \\equal{} h^2 \\plus{} 3h \\plus{} 2 \\minus{} 2sh$\r\n$ 74 \\equal{} h(h \\plus{} 3 \\minus{} 2s)$\r\nso $ h|74$ then $ h \\equal{} 1,2,37,74$ so the only possible solution is: $ (h,s) \\equal{} (37,19)$"
}
{
"Tag": [
"inequalities",
"inequalities open"
],
"Problem": "we know schur can be write as (3,0,0)+(1,1,1)>=2(2,1,0) and it seems that it can't be proved directlly by muirhead inequality.\r\nof course i kown schur is equivalent to (a+b-c)(b+c-a)(c+a-b)<=abc,and it can be proved using AM-GM,but i want to know whether it can be proved by muirhead directlly.thanks :D",
"Solution_1": "You'll never get it, i think..\r\n\r\nSchur :\r\n\r\n$ \\sum_{cyc} a^t(a\\minus{}b)(a\\minus{}c) \\ge0$\r\n\r\nSince there is $ [t,1,1]$ in LHS but $ [t\\plus{}1,1,0]$ in RHS, you will never get a proof with Muirhead only..\r\n\r\nYou'll need some manipulations first..",
"Solution_2": "Schur can be written (I prefer that form) as\r\n$ (a\\plus{}2b,0,0)\\plus{}(a,b,b)\\ge2(a\\plus{}b,b,0)$\r\nFrom Muirhead you can get $ (a\\plus{}2b,0,0)\\ge(a\\plus{}b,b,0)$, but also $ (a,b,b)\\le(a\\plus{}b,b,0)$, so it can't be proven just with Muirhead.",
"Solution_3": "@Bugi: Your notation is wrong. Schur is $ [a\\plus{}2b,0,0] \\plus{} [a,b,b] \\ge 2[a\\plus{}b,b,0]$! It should not be with ordinary parantheses :)\r\n\r\nYou probably confused it with $ (a\\plus{}2b,0,0) \\succ (a\\plus{}b,b,b)$"
}
{
"Tag": [
"geometry",
"geometric transformation",
"reflection",
"circumcircle",
"trigonometry",
"projective geometry",
"cyclic quadrilateral"
],
"Problem": "(A.Zaslavsky, 9--10) The reflections of diagonal $ BD$ of a quadrilateral $ ABCD$ in the bisectors of angles $ B$ and $ D$ pass\r\nthrough the midpoint of diagonal $ AC$. Prove that the reflections of diagonal $ AC$ in the bisectors of angles $ A$ and $ C$ pass\r\nthrough the midpoint of diagonal $ BD$ (There was an error in published condition of this problem).",
"Solution_1": "Notice that the quadrilateral also must be cyclic. In this case, denote by $ M$, $ N$ the midpoints of the diagonals $ BD$, $ AC$, respectively, and by $ P$ the intersection point of the diagonals of $ ABCD$. Since the lines $ BD$, and $ BM$ are isogonal with respect to angle $ ABC$, by Steiner's theorem, $ \\frac{AB^{2}}{BC^{2}}\\equal{}\\frac{AP}{PC}.$ Similarly, the lines $ DB$ and $ DM$ are isogonal with respect to angle $ CDA$, and therefore $ \\frac{AD^{2}}{DC^{2}}\\equal{}\\frac{AP}{PC}.$ Hence $ AB/BC\\equal{}AD/DC$; in other words, the quadrilateral $ ABCD$ is harmonic. Now, since the pencil $ B(B, A, D, C)$ is harmonic, by intersecting it with the line $ AC$, we deduce that $ T$ (the intersection point of the tangent at $ B$ to the circumcircle of $ ABCD$ and line $ AC$) is the harmonic conjugate of $ P$ with respect to segment $ (AC)$. On other hand, the pencil $ D(D, A, B, C)$ is harmonic, and thus the tangent at $ D$ to the circumcircle of $ ABCD$ passes through $ T$. \r\n\r\nIt is now well-known that if $ XYZ$ is the tangential triangle of a given triangle $ ABC$, the lines $ AX$, $ BY$, $ CZ$ are the symmedians of $ ABC$. According to this fact, and to the collinearity of the points $ A$, $ C$, and $ T$, the diagonal $ AC$ is the $ A$-symmedian of triangle $ ABD$, and simultanously the $ C$-symmedian of triangle $ BCD$. Therefore, the reflections of $ AC$ in the bisectors of angles $ A$, and $ C$ of the cyclic quadrilateral $ ABC$ pass through the midpoint of $ BD$. This completes our proof.",
"Solution_2": "I have another solution, which uses the following known fact: considering the projective plane with the circumcircle of $ ABC$ as reference, the symmedians of $ ABC$ are obtained connecting the corresponding vertex to the dual of the opposite side. Then the problem asks to prove that if $ AD$ is the $ A$-symmedian of $ ABC$ then $ BC$ is the $ B$ symmedian of $ ABD$.\r\n\r\nIn order to prove that, let $ P$ be the dual point of $ BC$ and $ Q$ be the dual point of $ AD$. Then $ P$ belongs to $ AD$ and the result follows from the duality theorem: the dual of $ P$, $ BC$ passes through the dual of $ AD$, $ Q$.",
"Solution_3": "[hide=\"Solution\"]Since $BD$ is a symmedian of triangles $ABC$, $ADC$, we obtain $\\frac{[ABD]}{[CBD]}=\\frac{\\frac{1}{2}BA\\cdot BD\\sin\\angle DBA}{\\frac{1}{2}BC\\cdot BD\\sin\\angle DBC}=\\frac{BA\\sin\\angle DBA}{BC\\sin\\angle DBC}=\\frac{BA^2}{BC^2}$. Similarly $\\frac{[ABD]}{[CBD]}=\\frac{DA^2}{DC^2}$; hence $\\frac{BA}{BC}=\\frac{DA}{DC}$. Let $D'=BD\\cap\\odot(ABC)$. Since $\\frac{D'A}{D'C}=\\frac{DA}{DC}=\\frac{BA}{BC}$, it follows that $A$, $D$, $D'$ lie on the $B$-Apollonius circle WRT triangle $ABC$. They are also, however, collinear, so $D'\\equiv D$. It follows that quadrilateral $ABCD$ is harmonic, and hence we may conclude. $\\blacksquare$[/hide]"
}
{
"Tag": [
"geometry",
"rectangle"
],
"Problem": "The side of square $ S$ is $ x$ units in length. The lengths of two\nnon-adjacent sides of the square are each increased by 1 unit and the\nother two sides are each decreased by 1 unit. Square $ S$ has thus been\ntransformed into rectangle $ R$. What is the number of square units in\nthe positive difference between the areas of $ S$ and $ R$?",
"Solution_1": "This is a geometric representation and demonstration of the sum and difference algebraic property. $ (x\\plus{}1)(x\\minus{}1)\\equal{}x^2\\minus{}1$, so the difference in area between the two rectangles is $ \\boxed{1}$."
}