{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb\n",
    "model_path = \"model.None\"\n",
    "model = xgb.Booster()\n",
    "model.load_model(model_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dv_path = \"dv.bin\"\n",
    "with open(dv_path, 'rb') as f_out:\n",
    "    dv = pickle.load(f_out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler_path = \"scaler.bin\"\n",
    "with open(scaler_path, 'rb') as f_out:\n",
    "    scaler = pickle.load(f_out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess(data):\n",
    "    \"\"\"Preprocessing of the data\"\"\"\n",
    "    # turn json input to dataframe\n",
    "    data = pd.DataFrame([data])\n",
    "\n",
    "    # define numerical and categorical features\n",
    "    numerical = [\"X1\", \"X2\", \"X3\", \"X4\", \"X5\", \"X7\"]\n",
    "    categorical = [\"X6\", \"X8\"]\n",
    "\n",
    "    # preprocess numerical features\n",
    "    X_num = scaler.transform(data[numerical])\n",
    "    # preprocess categorical features\n",
    "    data[categorical] = data[categorical].astype(\"string\")\n",
    "    X_dicts = data[categorical].to_dict(orient=\"records\")\n",
    "    X_cat = dv.transform(X_dicts)\n",
    "    # concatenate both\n",
    "    X = np.concatenate((X_num, X_cat), axis=1)\n",
    "\n",
    "    return X\n",
    "\n",
    "\n",
    "def predict(X):\n",
    "    \"\"\"make predictions\"\"\"\n",
    "    pred = model.predict(X)\n",
    "    print('prediction', pred[0])\n",
    "    return float(pred[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def main(input_data):\n",
    "    \"\"\"request input, preprocess it and make prediction\"\"\"\n",
    "    features = preprocess(input_data)\n",
    "    features_2 = xgb.DMatrix(features)\n",
    "    pred = predict(features_2)\n",
    "\n",
    "    result = {'heat load': pred}\n",
    "\n",
    "    return result\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prediction 15.648413\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'heat load': 15.648412704467773}"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "input_example = {\n",
    "    \"X1\": 0.98,\n",
    "    \"X2\": 514.50,\n",
    "    \"X3\": 294.00,\n",
    "    \"X4\": 110.25,\n",
    "    \"X5\": 7.00,\n",
    "    \"X6\": 2,\n",
    "    \"X7\": 0.00,\n",
    "    \"X8\": 0,\n",
    "}\n",
    "\n",
    "main(input_example)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mlops",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}