Update app.py

app.py

@@ -16,7 +16,6 @@ logger = logging.getLogger(__name__)
 CONVERSATION_PROMPT = """
 You are an engaging and insightful career advisor. Have natural conversations to learn about their career.
 Use an enthusiastic, supportive tone and show genuine interest in their journey.
-
 CONVERSATION STYLE:
 - Be warm and engaging
 - Show genuine interest in their experiences
@@ -25,7 +24,6 @@ CONVERSATION STYLE:
 - Use conversational language, not formal queries
 - Express enthusiasm about their achievements
 - Dig deeper into interesting points they make
-
 INFORMATION TO GATHER (through natural conversation):
 1. Current Role Details:
    - Job title and responsibilities
@@ -33,47 +31,38 @@ INFORMATION TO GATHER (through natural conversation):
    - Team size and structure
    - Project scope and impact
    - Current compensation (base, bonus, equity)
-
 2. Experience Deep-Dive:
    - Career progression story
    - Leadership experience
    - Major projects and achievements
    - Technical skills and expertise
    - Industry knowledge
-
 3. Educational Background:
    - Degrees and certifications
    - Specialized training
    - Continuous learning
-
 4. Work Environment:
    - Location and market
    - Remote/hybrid setup
    - Growth opportunities
    - Company culture
-
 CONVERSATION FLOW:
 1. Start with: "Hi! I'd love to hear about your career journey. What kind of work are you doing currently?"
-
 2. After each response:
    - Pick up on specific details they mentioned
    - Ask engaging follow-up questions
    - Show genuine interest in their experiences
    - Build on previous information shared
-
 3. If they mention something interesting, probe deeper:
    - "That project sounds fascinating! What were some unique challenges you faced?"
    - "Leading a team must be exciting! How did you approach building and motivating your team?"
    - "Interesting technology stack! What made you choose those specific tools?"
-
 4. When compensation is mentioned:
    - Be tactful and professional
    - Acknowledge their goals
    - Ask about their desired growth
-
 5. Once you have enough information, say:
    "I've got a good understanding of your career profile now! Would you like to see your personalized salary growth projection? Just click 'Generate Analysis' and I'll create a detailed forecast based on our discussion."
-
 IMPORTANT:
 - Keep conversation flowing naturally
 - Don't rush to collect information
@@ -84,9 +73,7 @@ IMPORTANT:
 
 EXTRACTION_PROMPT = """
 Analyze the conversation and extract numerical scores from 0 to 1 based on salary growth potential.
-
 SCORING GUIDELINES:
-
 1. Industry Score (0-1):
    Industry Type & Growth:
    - 1.0: Cutting-edge AI/ML companies
@@ -99,7 +86,6 @@ SCORING GUIDELINES:
    +0.1: Market leader
    +0.1: High growth trajectory
    -0.1: Declining market position
-
 2. Experience Score (0-1):
    Years and Level:
    - 1.0: 15+ years with executive experience
@@ -112,7 +98,6 @@ SCORING GUIDELINES:
    +0.1: Rapid promotions
    +0.1: Significant achievements
    +0.1: High-impact projects
-
 3. Education Score (0-1):
    Formal Education:
    - 1.0: PhD from top institution
@@ -125,7 +110,6 @@ SCORING GUIDELINES:
    +0.1: Relevant certifications
    +0.1: Continuous learning
    +0.1: Field-specific expertise
-
 4. Skills Score (0-1):
    Technical Depth:
    - 1.0: Industry-leading expertise
@@ -138,7 +122,6 @@ SCORING GUIDELINES:
    +0.1: Multiple in-demand skills
    +0.1: Proven implementation
    +0.1: Cross-functional expertise
-
 5. Location Score (0-1):
    Market Strength:
    - 1.0: Major tech hubs (SF, NYC)
@@ -151,7 +134,6 @@ SCORING GUIDELINES:
    +0.1: Remote work option
    +0.1: High-growth market
    +0.1: Strategic location
-
 Return a JSON object with exactly these fields:
 {
     "industry_score": float,
@@ -161,7 +143,6 @@ Return a JSON object with exactly these fields:
     "location_score": float,
     "current_salary": float
 }
-
 Base scores on available information. Make reasonable assumptions for missing data based on context clues.
 """
 
@@ -169,10 +150,7 @@ class CodeEnvironment:
     """Environment for executing visualization code"""
     
     def __init__(self):
-        self.globals = {
-            'np': np,
-            'plt': plt
-        }
+        self.globals = {'np': np, 'plt': plt}
         self.locals = {}
     
     def execute(self, code: str, paths: np.ndarray = None) -> Dict[str, Any]:
@@ -181,81 +159,59 @@ class CodeEnvironment:
             self.globals['paths'] = paths
             
         result = {'figures': [], 'error': None}
-        
         try:
             exec(code, self.globals, self.locals)
-            
-            # Save current figure
             buf = io.BytesIO()
             plt.gcf().savefig(buf, format='png', dpi=300, bbox_inches='tight')
             buf.seek(0)
             result['figures'].append(buf.getvalue())
             plt.close('all')
-            
         except Exception as e:
-            result['error'] = str(e)
+            result['error'] = f"Visualization failed: {str(e)}"
             plt.close('all')
-            
         return result
 
 class SalarySimulator:
     """Monte Carlo simulation for salary projections"""
     
-    def __init__(self):
-        self.years = 5
-        self.num_paths = 1000
+    def __init__(self, years: int = 5, num_paths: int = 1000):
+        self.years = years
+        self.num_paths = num_paths
         
     def run_simulation(self, profile: Dict[str, float]) -> np.ndarray:
-        """Generate salary growth paths"""
+        """Generate salary growth paths using Monte Carlo simulation"""
         paths = np.zeros((self.num_paths, self.years + 1))
         paths[:, 0] = profile['current_salary']
         
-        # Base growth parameters
         base_growth = 0.02 + (profile['industry_score'] * 0.04)
         skill_premium = 0.01 + (profile['skills_score'] * 0.02)
         exp_premium = 0.01 + (profile['experience_score'] * 0.02)
         edu_premium = 0.005 + (profile['education_score'] * 0.015)
         location_premium = 0.01 + (profile['location_score'] * 0.02)
         
-        # Risk parameters
         volatility = 0.05 + (profile['industry_score'] * 0.05)
         disruption_chance = 0.1
         disruption_impact = 0.2
         
-        # Generate paths
         for path in range(self.num_paths):
             salary = paths[path, 0]
             for year in range(1, self.years + 1):
-                # Calculate base growth
-                growth = (base_growth + skill_premium + exp_premium + 
-                         edu_premium + location_premium)
-                
-                # Add volatility
+                growth = base_growth + skill_premium + exp_premium + edu_premium + location_premium
                 growth += np.random.normal(0, volatility)
-                
-                # Possible disruption
                 if np.random.random() < disruption_chance:
                     impact = disruption_impact * np.random.random()
-                    if np.random.random() < 0.7:  # 70% positive
-                        growth += impact
-                    else:
-                        growth -= impact
-                
-                # Apply bounds
-                growth = max(min(growth, 0.25), -0.1)  # -10% to +25%
-                
-                # Update salary
+                    growth += impact if np.random.random() < 0.7 else -impact
+                growth = max(min(growth, 0.25), -0.1)
                 salary *= (1 + growth)
                 paths[path, year] = salary
-        
         return paths
 
 class CareerAdvisor:
     """Main career advisor system"""
     
-    def __init__(self):
+    def __init__(self, years: int = 5, num_paths: int = 1000):
         self.chat_history = []
-        self.simulator = SalarySimulator()
+        self.simulator = SalarySimulator(years, num_paths)
         self.code_env = CodeEnvironment()
     
     def reset(self):
@@ -265,146 +221,94 @@ class CareerAdvisor:
     def chat(self, message: str, api_key: str) -> str:
         """Process user message and generate response"""
         if not api_key.strip().startswith("sk-"):
-            return "Please enter a valid OpenAI API key."
-        
+            return "Please enter a valid OpenAI API key starting with 'sk-'."
         try:
-            messages = [
-                {"role": "system", "content": CONVERSATION_PROMPT}
-            ] + self.chat_history + [
-                {"role": "user", "content": message}
-            ]
-            
-            response = completion(
-                model="gpt-4o-mini",
-                messages=messages,
-                api_key=api_key
-            )
-            
-            # Update chat history
+            messages = [{"role": "system", "content": CONVERSATION_PROMPT}] + \
+                       self.chat_history + [{"role": "user", "content": message}]
+            response = completion(model="gpt-4o-mini", messages=messages, api_key=api_key)
             self.chat_history.extend([
                 {"role": "user", "content": message},
                 {"role": "assistant", "content": response.choices[0].message.content}
             ])
-            
             return response.choices[0].message.content
-            
         except Exception as e:
-            return f"Error: {str(e)}"
+            return f"Chat error: {str(e)}. Please check your API key or try again."
     
     def generate_analysis(self, api_key: str) -> Tuple[str, bytes]:
         """Generate complete analysis with visualization"""
+        if not self.chat_history:
+            return "Please chat about your career first to generate an analysis.", None
         try:
-            # Extract profile
             profile = self._extract_profile(api_key)
-            
-            # Run simulation
             paths = self.simulator.run_simulation(profile)
             
-            # Generate visualization
             viz_code = """
 import matplotlib.pyplot as plt
 import numpy as np
-
-# Setup
 plt.style.use('dark_background')
 fig = plt.figure(figsize=(12, 16))
-
-# Top plot - Salary paths
 ax1 = plt.subplot2grid((2, 1), (0, 0))
-
-# Plot paths with transparency
-for path in paths[::20]:  # Plot every 20th path to reduce density
+for path in paths[::20]:
     ax1.plot(range(paths.shape[1]), path, color='#4a90e2', alpha=0.1, linewidth=1)
-
-# Calculate and plot percentiles
 percentiles = [10, 25, 50, 75, 90]
 colors = ['#ff9999', '#ffcc99', '#ffffff', '#ffcc99', '#ff9999']
 labels = ['10th', '25th', 'Median', '75th', '90th']
-
 for p, color, label in zip(percentiles, colors, labels):
     line = np.percentile(paths, p, axis=0)
     ax1.plot(range(paths.shape[1]), line, color=color, linewidth=2, label=f'{label} percentile')
-
 ax1.set_title('Salary Growth Projections\n', fontsize=16, pad=20)
 ax1.set_xlabel('Years', fontsize=12)
 ax1.set_ylabel('Salary ($)', fontsize=12)
 ax1.grid(True, alpha=0.2)
 ax1.legend(fontsize=10)
-
-# Format y-axis as currency
 ax1.yaxis.set_major_formatter(plt.FuncFormatter(lambda x, p: f'${x:,.0f}'))
-
-# Customize x-axis
 ax1.set_xticks(range(paths.shape[1]))
 ax1.set_xticklabels(['Current'] + [f'Year {i+1}' for i in range(paths.shape[1]-1)])
-
-# Bottom plot - Final distribution
 ax2 = plt.subplot2grid((2, 1), (1, 0))
 final_salaries = paths[:, -1]
-
-# Create histogram
 ax2.hist(final_salaries, bins=50, color='#4a90e2', alpha=0.7)
 ax2.set_title('Final Salary Distribution\n', fontsize=16, pad=20)
 ax2.set_xlabel('Salary ($)', fontsize=12)
 ax2.set_ylabel('Frequency', fontsize=12)
 ax2.grid(True, alpha=0.2)
-
-# Format x-axis as currency
 ax2.xaxis.set_major_formatter(plt.FuncFormatter(lambda x, p: f'${x:,.0f}'))
-
-# Add percentile lines
 for p, color in zip(percentiles, colors):
     value = np.percentile(final_salaries, p)
     ax2.axvline(x=value, color=color, linestyle='--', alpha=0.5)
-
 plt.tight_layout(pad=4)
 """
-            
-            # Execute visualization code
             viz_result = self.code_env.execute(viz_code, paths)
-            
             if viz_result['error']:
-                raise Exception(f"Visualization error: {viz_result['error']}")
-            
-            # Generate summary
+                return f"Analysis generated, but {viz_result['error']}", None
             summary = self._generate_summary(profile, paths)
-            
             return summary, viz_result['figures'][0]
-            
         except Exception as e:
-            return f"Error generating analysis: {str(e)}", None
+            return f"Analysis error: {str(e)}. Please ensure sufficient chat history.", None
     
     def _extract_profile(self, api_key: str) -> Dict[str, float]:
         """Extract profile scores from conversation"""
-        conversation = "\n".join([
-            f"{msg['role']}: {msg['content']}"
-            for msg in self.chat_history
-        ])
-        
+        conversation = "\n".join([f"{msg['role']}: {msg['content']}" for msg in self.chat_history])
         messages = [
             {"role": "system", "content": EXTRACTION_PROMPT},
             {"role": "user", "content": f"Extract profile from:\n{conversation}"}
         ]
-        
         response = completion(
             model="gpt-4o-mini",
             messages=messages,
             api_key=api_key,
             response_format={"type": "json_object"}
         )
-        
         return json.loads(response.choices[0].message.content)
     
     def _generate_summary(self, profile: Dict[str, float], paths: np.ndarray) -> str:
         """Generate analysis summary"""
         final_salaries = paths[:, -1]
         initial_salary = paths[0, 0]
-        cagr = (np.median(final_salaries) / initial_salary) ** (1/5) - 1
+        cagr = (np.median(final_salaries) / initial_salary) ** (1/self.simulator.years) - 1
         
         return f"""
         Career Profile Analysis
         ======================
-
         Current Situation:
         • Salary: ${profile['current_salary']:,.2f}
         • Industry Position: {profile['industry_score']:.2f}/1.0
@@ -412,110 +316,66 @@ plt.tight_layout(pad=4)
         • Education Rating: {profile['education_score']:.2f}/1.0
         • Skills Assessment: {profile['skills_score']:.2f}/1.0
         • Location Impact: {profile['location_score']:.2f}/1.0
-
-        5-Year Projection:
+        {self.simulator.years}-Year Projection:
         • Conservative (25th percentile): ${np.percentile(final_salaries, 25):,.2f}
         • Most Likely (Median): ${np.percentile(final_salaries, 50):,.2f}
         • Optimistic (75th percentile): ${np.percentile(final_salaries, 75):,.2f}
         • Expected Annual Growth: {cagr*100:.1f}%
-
         Key Insights:
         • Your profile suggests {cagr*100:.1f}% annual growth potential
         • {profile['industry_score']:.2f} industry score indicates {'strong' if profile['industry_score'] > 0.7 else 'moderate' if profile['industry_score'] > 0.5 else 'challenging'} growth environment
         • Skills rating of {profile['skills_score']:.2f} suggests {'excellent' if profile['skills_score'] > 0.7 else 'good' if profile['skills_score'] > 0.5 else 'potential for'} career advancement
         • Location score {profile['location_score']:.2f} {'enhances' if profile['location_score'] > 0.7 else 'supports' if profile['location_score'] > 0.5 else 'may limit'} opportunities
-
-        Based on {paths.shape[0]:,} simulated career paths
+        Based on {self.simulator.num_paths:,} simulated career paths
         """
 
 def create_interface():
-    """Create Gradio interface"""
-    advisor = CareerAdvisor()
+    """Create Gradio interface with configurable simulation parameters"""
+    advisor = None
+    
+    def init_advisor(years: int, num_paths: int):
+        nonlocal advisor
+        advisor = CareerAdvisor(years=max(1, years), num_paths=max(100, num_paths))
+        advisor.reset()
     
     def user_message(message: str, history: List, api_key: str) -> Tuple[str, List]:
-        """Handle user message"""
         if not message.strip():
             return "", history
-            
+        if not advisor:
+            return "Please set simulation parameters first.", history
         response = advisor.chat(message, api_key)
-        history = history + [(message, response)]
-        return "", history
+        return "", history + [(message, response)]
     
     def generate_analysis(api_key: str, history: List) -> Tuple[str, gr.Image]:
-        """Generate analysis"""
-        if not history:
-            return "Please chat about your career first.", None
-            
+        if not advisor or not history:
+            return "Please chat about your career and set parameters first.", None
         summary, figure_data = advisor.generate_analysis(api_key)
-        if figure_data:
-            return summary, figure_data
-        return summary, None
+        return summary, figure_data if figure_data else None
     
-    # Create interface
     with gr.Blocks(title="Monte Carlo Salary Prediction", theme=gr.themes.Soft()) as demo:
-        gr.Markdown("""
-        # 💰 Monte Carlo Simulation of Salary Prediction
+        gr.Markdown("# 💰 Monte Carlo Simulation of Salary Prediction\nChat about your career to see your growth potential!")
         
-        Chat about your career, and I'll analyze your growth potential using Monte Carlo simulation!
-        """)
-        
-        api_key = gr.Textbox(
-            label="OpenAI API Key",
-            placeholder="Enter your OpenAI API key",
-            type="password"
-        )
+        with gr.Row():
+            api_key = gr.Textbox(label="OpenAI API Key", placeholder="Enter your OpenAI API key", type="password")
+            years = gr.Number(label="Simulation Years", value=5, minimum=1, step=1)
+            num_paths = gr.Number(label="Number of Paths", value=1000, minimum=100, step=100)
         
-        chatbot = gr.Chatbot(
-            value=[],
-            height=400,
-            show_copy_button=True,
-            avatar_images=["/gradio/user.png", "/gradio/assistant.png"],
-            bubble_full_width=False
-        )
+        chatbot = gr.Chatbot(value=[], height=400, show_copy_button=True)
         
         with gr.Row():
-            msg = gr.Textbox(
-                label="Your message",
-                placeholder="Tell me about your career...",
-                lines=2,
-                show_copy_button=False
-            )
+            msg = gr.Textbox(label="Your message", placeholder="Tell me about your career...", lines=2)
             send = gr.Button("Send", variant="primary", scale=0)
         
         analyze = gr.Button("Generate Analysis", variant="secondary", size="lg")
         
         with gr.Row():
-            analysis = gr.Textbox(
-                label="Analysis",
-                lines=20,
-                show_copy_button=True
-            )
-            plot = gr.Image(
-                label="Projections",
-                show_download_button=True,
-                height=600
-            )
-        
-        # Wire up the interface
-        msg.submit(
-            user_message,
-            inputs=[msg, chatbot, api_key],
-            outputs=[msg, chatbot]
-        )
-        
-        send.click(
-            user_message,
-            inputs=[msg, chatbot, api_key],
-            outputs=[msg, chatbot]
-        )
-        
-        analyze.click(
-            generate_analysis,
-            inputs=[api_key, chatbot],
-            outputs=[analysis, plot]
-        )
+            analysis = gr.Textbox(label="Analysis", lines=10, show_copy_button=True)
+            plot = gr.Image(label="Projections", show_download_button=True, height=600)
         
-        demo.load(lambda: advisor.reset())
+        demo.load(lambda y, n: init_advisor(y, n), inputs=[years, num_paths], outputs=None)
+        msg.submit(user_message, inputs=[msg, chatbot, api_key], outputs=[msg, chatbot])
+        send.click(user_message, inputs=[msg, chatbot, api_key], outputs=[msg, chatbot])
+        analyze.click(generate_analysis, inputs=[api_key, chatbot], outputs=[analysis, plot])
     
     return demo