Chapter 8 - Hotel Cancelation .ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from lightgbm import LGBMRegressor,LGBMClassifier\n",
    "from sklearn.linear_model import LinearRegression,LogisticRegression\n",
    "from xgboost import XGBRegressor,XGBClassifier\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.metrics import mean_squared_error,f1_score,precision_score,recall_score\n",
    "\n",
    "from sklearnex import patch_sklearn,unpatch_sklearn\n",
    "\n",
    "from scipy.stats import f_oneway\n",
    "\n",
    "import gc\n",
    "import pickle\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from collections import Counter\n",
    "\n",
    "from Interaction_effect_features import get_data\n",
    "\n",
    "from MetaHeuristicsFS import FeatureSelection\n",
    "\n",
    "# from EvolutionaryFS import GeneticAlgorithmFS\n",
    "# from SimmulatedAnnealingFS import SimulatedAnnealingFS\n",
    "# from ACO import AntColonyOptimizationFS\n",
    "# from PSOFinal import ParticleSwarmOptimizationFS\n",
    "\n",
    "import shap\n",
    "\n",
    "np.seterr(divide = 'ignore')\n",
    "pd.options.mode.chained_assignment = None  # default='warn'\n",
    "import warnings\n",
    "# suppress warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_metrics(model,data_dict,problem_type,dependent_variable,features,cost_funct):\n",
    "    '''\n",
    "    problem_type = 'regression' or classification\n",
    "    '''\n",
    "    \n",
    "    test_tree_cost = {}\n",
    "    \n",
    "    external_tree_cost = []\n",
    "    external_tree_precision = []\n",
    "    external_tree_recall = []\n",
    "    \n",
    "    data_external_test = data_dict['data_external_test']\n",
    "    data_validation_test = data_dict['data_validation_test']\n",
    "    \n",
    "    #for each training data\n",
    "    for data_fold_index in data_dict['fold_dict'].keys():\n",
    "        fold_training = data_dict['fold_dict'][data_fold_index]['fold_training']\n",
    "        fold_test = data_dict['fold_dict'][data_fold_index]['fold_test']\n",
    "        \n",
    "        #based on regression/classification\n",
    "        if problem_type == 'regression':\n",
    "            cost_function = cost_funct\n",
    "        else:\n",
    "            cost_function = f1_score\n",
    "        \n",
    "        Y_train = fold_training[dependent_variable]\n",
    "        Y_test = fold_test[dependent_variable]\n",
    "        Y_validation = data_validation_test[dependent_variable]\n",
    "        Y_external = data_external_test[dependent_variable]\n",
    "        \n",
    "        model.fit(fold_training[features],Y_train)\n",
    "                \n",
    "        Y_test_tree = model.predict(fold_test[features])\n",
    "        Y_validation_tree = model.predict(data_validation_test[features])\n",
    "        Y_external_tree = model.predict(data_external_test[features])\n",
    "        \n",
    "        place_holder = np.mean([cost_funct(Y_test,Y_test_tree),cost_funct(Y_validation,Y_validation_tree)])\n",
    "        \n",
    "        #predict for test and calculate cost function\n",
    "        test_tree_cost['CV Sample-'+str(data_fold_index)] = place_holder\n",
    "\n",
    "        #predict for external test and calculate cost function\n",
    "        external_tree_cost.append(cost_funct(Y_external,Y_external_tree))\n",
    "        \n",
    "        if problem_type != 'regression':\n",
    "            external_tree_precision.append(precision_score(Y_external,Y_external_tree))\n",
    "            external_tree_recall.append(recall_score(Y_external,Y_external_tree))\n",
    "            \n",
    "    #return cost\n",
    "    if problem_type != 'regression':\n",
    "        #print('External Test precision: ' + str(round(np.mean(external_tree_precision),2)) + ', recall: ' + str(round(np.mean(external_tree_recall),2)))\n",
    "        print('External Test recall: ' + str(round(np.mean(external_tree_recall),2)))\n",
    "    return test_tree_cost,external_tree_cost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_plot(metric,title,test_tree_cost,external_tree_cost):\n",
    "    sns.set(rc={'figure.figsize':(16,8)})\n",
    "    sns.set_context('paper', font_scale = 1.3)\n",
    "    if metric=='RMSE':\n",
    "        fig, ax = plt.subplots(1, 2)\n",
    "    else:\n",
    "        fig, ax = plt.subplots(1, 2, sharey=True)\n",
    "        \n",
    "    validation_name = [' External_Test','External_Test ',' External_Test ',' External_Test  ','  External_Test']\n",
    "\n",
    "    ax[0].bar(list(test_tree_cost.keys()),list(test_tree_cost.values()))\n",
    "    ax[0].set_title(metric + ' for test data and validation data')\n",
    "\n",
    "    ax[1].bar(validation_name,external_tree_cost)\n",
    "    ax[1].set_title(metric + ' for external test data')\n",
    "\n",
    "    #plt.title('')\n",
    "    \n",
    "def get_data_for_FS(data_dict,dependent_variable,features_all):\n",
    "    \n",
    "    data_dict_for_SA = {}\n",
    "\n",
    "#     x_validation = data_dict['data_external_test'][features_all]\n",
    "#     y_validation = data_dict['data_external_test'][dependent_variable]\n",
    "    \n",
    "    x_validation = data_dict['data_validation_test'][features_all]\n",
    "    y_validation = data_dict['data_validation_test'][dependent_variable]\n",
    "\n",
    "    #for each training data\n",
    "    for data_fold_index in data_dict['fold_dict'].keys():\n",
    "        \n",
    "        for feature in features_all:\n",
    "            if feature not in data_dict['fold_dict'][data_fold_index]['fold_training'].columns:\n",
    "                print(feature,'not present')\n",
    "\n",
    "        data_dict_for_SA[data_fold_index]={'x_train':data_dict['fold_dict'][data_fold_index]['fold_training'][features_all],\n",
    "                                           'y_train':data_dict['fold_dict'][data_fold_index]['fold_training'][dependent_variable],\n",
    "                                           'x_test':data_dict['fold_dict'][data_fold_index]['fold_test'][features_all],\n",
    "                                           'y_test':data_dict['fold_dict'][data_fold_index]['fold_test'][dependent_variable]}\n",
    "    \n",
    "#     for feature in features_all:\n",
    "#         if '_Encoded' not in feature:\n",
    "            \n",
    "    \n",
    "    return data_dict_for_SA,x_validation,y_validation,features_all\n",
    "\n",
    "def last_preprocessing(data_dict):\n",
    "    \n",
    "    columns = ['LeadTime_Quartile_countEncoded','LeadTime_Quartile_percentEncoded','LeadTime_Quartile_countRankEncoded','LeadTime_Quartile_MeanEncoded','ADR_Quartile_countEncoded','ADR_Quartile_percentEncoded','ADR_Quartile_countRankEncoded','ADR_Quartile_MeanEncoded']\n",
    "\n",
    "    for feature in columns:\n",
    "        data_dict['data_external_test'][feature] = data_dict['data_external_test'][feature].astype(int)\n",
    "        data_dict['data_validation_test'][feature] = data_dict['data_validation_test'][feature].astype(int)\n",
    "        \n",
    "        for index in data_dict['fold_dict'].keys():\n",
    "            data_dict['fold_dict'][index]['fold_training'][feature] = data_dict['fold_dict'][index]['fold_training'][feature].astype(int)\n",
    "            data_dict['fold_dict'][index]['fold_test'][feature] = data_dict['fold_dict'][index]['fold_test'][feature].astype(int)\n",
    "\n",
    "\n",
    "    return data_dict\n",
    "\n",
    "def last_preprocessing_linear(data_dict):\n",
    "    \n",
    "    columns = ['LeadTime_Quartile_countEncoded','LeadTime_Quartile_percentEncoded','LeadTime_Quartile_countRankEncoded','LeadTime_Quartile_MeanEncoded','ADR_Quartile_countEncoded','ADR_Quartile_percentEncoded','ADR_Quartile_countRankEncoded','ADR_Quartile_MeanEncoded']\n",
    "    for feature in columns:\n",
    "        data_dict['data_external_test'][feature] = data_dict['data_external_test'][feature].astype(int)\n",
    "        data_dict['data_validation_test'][feature] = data_dict['data_validation_test'][feature].astype(int)\n",
    "        \n",
    "        for index in data_dict['fold_dict'].keys():\n",
    "            data_dict['fold_dict'][index]['fold_training'][feature] = data_dict['fold_dict'][index]['fold_training'][feature].astype(int)\n",
    "            data_dict['fold_dict'][index]['fold_test'][feature] = data_dict['fold_dict'][index]['fold_test'][feature].astype(int)\n",
    "\n",
    "\n",
    "    return data_dict\n",
    "\n",
    "\n",
    "def cost_funct(y_true,y_pred):\n",
    "    return precision_score(y_true,y_pred)\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def anova_fs(linear_features_all,data_dict,dependent_variable):\n",
    "    \n",
    "    common_list = []\n",
    "    \n",
    "    for feature in linear_features_all:\n",
    "        \n",
    "        p_counter = 0\n",
    "        \n",
    "        for index in data_dict['fold_dict'].keys():\n",
    "            \n",
    "            indx_df = data_dict['fold_dict'][index]['fold_training']\n",
    "            \n",
    "            class1 = indx_df[indx_df[dependent_variable[0]]==1][feature]\n",
    "            class2 = indx_df[indx_df[dependent_variable[0]]==0][feature]\n",
    "            \n",
    "            stat,p = f_oneway(class1, class2)\n",
    "            \n",
    "            if p <= 0.05:\n",
    "                p_counter += 1\n",
    "        \n",
    "        if p_counter == 5:\n",
    "            common_list.append(feature)\n",
    "        \n",
    "    return common_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hotel Cancellations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "problem = 'HotelCancellations'\n",
    "data_dict,dependent_variable,categorical_linear,categorical_tree,categorical_features,numerical_features_linear,numerical_features_tree,ordinal_features,ordinal_features_engineered = get_data(problem)\n",
    "\n",
    "data_dict = last_preprocessing(data_dict)\n",
    "data_dict_linear = last_preprocessing_linear(data_dict)\n",
    "\n",
    "linear_features_all = categorical_linear+numerical_features_linear+ordinal_features_engineered\n",
    "tree_features_all = categorical_tree+numerical_features_tree+ordinal_features_engineered\n",
    "\n",
    "#data_dict_for_linear,x_validation_linear,y_validation_linear = get_data_for_SA(data_dict,dependent_variable,features_all=linear_features_all)\n",
    "\n",
    "data_dict_for_tree,x_validation_tree,y_validation_tree,features_all = get_data_for_FS(data_dict,dependent_variable,features_all=tree_features_all)\n",
    "\n",
    "data_dict_for_linear,x_validation_linear,y_validation_linear,features_all_linear = get_data_for_FS(data_dict_linear,dependent_variable,features_all=linear_features_all)\n",
    "\n",
    "#common_list = anova_fs(linear_features_all,data_dict,dependent_variable)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'ReservedRoomType_countEncoded' in tree_features_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LightGBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_use = LGBMClassifier(random_state=1,n_jobs=-1,n_estimators=200)\n",
    "fsObj = FeatureSelection(columns_list = tree_features_all,\n",
    "                         data_dict = data_dict_for_tree,\n",
    "                         x_validation_dataframe = x_validation_tree,\n",
    "                         y_validation_dataframe = y_validation_tree,\n",
    "                         model = model_use,\n",
    "                         cost_function = cost_funct,\n",
    "                         cost_function_improvement = 'increase',\n",
    "                         average = None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.7835892709352555\n",
      "Best combined performance on test and validation data for iteration 1: 0.7969846870718833\n",
      "Best combined performance on test and validation data for iteration 2: 0.7970487272810414\n",
      "Best combined performance on test and validation data for iteration 3: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 4: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 5: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 6: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 7: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 8: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 9: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 10: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 11: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 12: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 13: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 14: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 15: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 16: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 17: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 18: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 19: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 20: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 21: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 22: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 23: 0.7971593925203762\n",
      "Best combined performance on test and validation data for iteration 24: 0.7971593925203762\n",
      "================= Best result: 0.7971593925203762 =================\n",
      "================= Execution time in minutes: 93.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Particle Swarm Optimization\n",
    "best_columns_PSO_HotelCancel_lightgbm = fsObj.ParticleSwarmOptimization(iterations = 25,                                                                   \n",
    "                                                                       swarmSize = 75,\n",
    "                                                                       run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_PSO_HotelCancel_lightgbm.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_PSO_HotelCancel_lightgbm, output_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.7844175434904177\n",
      "Best combined performance on test and validation data for iteration 1: 0.8166748232130565\n",
      "Best combined performance on test and validation data for iteration 2: 0.8209488291141775\n",
      "Best combined performance on test and validation data for iteration 3: 0.8211862374198946\n",
      "Best combined performance on test and validation data for iteration 4: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 5: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 6: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 7: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 8: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 9: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 10: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 11: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 12: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 13: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 14: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 15: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 16: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 17: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 18: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 19: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 20: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 21: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 22: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 23: 0.8212852202632543\n",
      "Best combined performance on test and validation data for iteration 24: 0.8212852202632543\n",
      "================= Best result: 0.8212852202632543 =================\n",
      "================= Execution time in minutes: 93.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Ant Colony Optimization\n",
    "best_columns_ACO_HotelCancel_lightgbm = fsObj.AntColonyOptimization(iterations = 25,\n",
    "                                                                N_ants = 75,\n",
    "                                                                run_time=1200,\n",
    "                                                                evaporation_rate=0.8,\n",
    "                                                                Q=0.2)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_ACO_HotelCancel_lightgbm.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_ACO_HotelCancel_lightgbm, output_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.7082985791015886\n",
      "Best combined performance on test and validation data for iteration 1: 0.7545843775852539\n",
      "Best combined performance on test and validation data for iteration 2: 0.7631305432256494\n",
      "Best combined performance on test and validation data for iteration 3: 0.7641235736850582\n",
      "Best combined performance on test and validation data for iteration 4: 0.7641235736850582\n",
      "Best combined performance on test and validation data for iteration 5: 0.7760484036901578\n",
      "Best combined performance on test and validation data for iteration 6: 0.7824246400112626\n",
      "Best combined performance on test and validation data for iteration 7: 0.7824246400112626\n",
      "Best combined performance on test and validation data for iteration 8: 0.783123110285967\n",
      "Best combined performance on test and validation data for iteration 9: 0.7866628575094309\n",
      "Best combined performance on test and validation data for iteration 10: 0.78863118712141\n",
      "Best combined performance on test and validation data for iteration 11: 0.78863118712141\n",
      "Best combined performance on test and validation data for iteration 12: 0.8103672436891977\n",
      "Best combined performance on test and validation data for iteration 13: 0.8103672436891977\n",
      "Best combined performance on test and validation data for iteration 14: 0.8141042706314272\n",
      "Best combined performance on test and validation data for iteration 15: 0.8148701587502604\n",
      "Best combined performance on test and validation data for iteration 16: 0.8148701587502604\n",
      "Best combined performance on test and validation data for iteration 17: 0.819550797300739\n",
      "Best combined performance on test and validation data for iteration 18: 0.8208566390975927\n",
      "Best combined performance on test and validation data for iteration 19: 0.826222329303625\n",
      "Best combined performance on test and validation data for iteration 20: 0.826222329303625\n",
      "Best combined performance on test and validation data for iteration 21: 0.8266098906330646\n",
      "Best combined performance on test and validation data for iteration 22: 0.8281076876961816\n",
      "Best combined performance on test and validation data for iteration 23: 0.8281076876961816\n",
      "Best combined performance on test and validation data for iteration 24: 0.8304294520460024\n",
      "Best combined performance on test and validation data for iteration 25: 0.830847012536372\n",
      "Best combined performance on test and validation data for iteration 26: 0.8314125178216598\n",
      "Best combined performance on test and validation data for iteration 27: 0.8314125178216598\n",
      "Best combined performance on test and validation data for iteration 28: 0.8314414468782481\n",
      "Best combined performance on test and validation data for iteration 29: 0.8317893110053195\n",
      "Best combined performance on test and validation data for iteration 30: 0.8325651589460866\n",
      "Best combined performance on test and validation data for iteration 31: 0.8325651589460866\n",
      "Best combined performance on test and validation data for iteration 32: 0.8343924084196814\n",
      "Best combined performance on test and validation data for iteration 33: 0.8343924084196814\n",
      "Best combined performance on test and validation data for iteration 34: 0.8343924084196814\n",
      "================= Best result: 0.8343924084196814 =================\n",
      "================= Execution time in minutes: 95.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Simulated Annealing\n",
    "best_columns_SA_HotelCancel_lightgbm = fsObj.SimulatedAnnealing(temperature = 1500,\n",
    "                                                               iterations = 35,\n",
    "                                                               n_perturb = 75,\n",
    "                                                               run_time=1200,\n",
    "                                                               n_features_percent_perturb=1,\n",
    "                                                               alpha=0.9)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_SA_HotelCancel_lightgbm.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_SA_HotelCancel_lightgbm, output_file)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for generation 0 : 0.7988091286792203\n",
      "Best combined performance on test and validation data for generation 1 : 0.8077927345317603\n",
      "Best combined performance on test and validation data for generation 2 : 0.8072789207708537\n",
      "Best combined performance on test and validation data for generation 3 : 0.8192945884103073\n",
      "Best combined performance on test and validation data for generation 4 : 0.8253564542084055\n",
      "Best combined performance on test and validation data for generation 5 : 0.8226497115659622\n",
      "Best combined performance on test and validation data for generation 6 : 0.8148813031096859\n",
      "Best combined performance on test and validation data for generation 7 : 0.8190026806828788\n",
      "Best combined performance on test and validation data for generation 8 : 0.8214159615296316\n",
      "Best combined performance on test and validation data for generation 9 : 0.820422598902326\n",
      "Best combined performance on test and validation data for generation 10 : 0.8215958849627979\n",
      "Best combined performance on test and validation data for generation 11 : 0.8255087860684679\n",
      "Best combined performance on test and validation data for generation 12 : 0.8187240513375544\n",
      "Best combined performance on test and validation data for generation 13 : 0.8193134454430909\n",
      "Best combined performance on test and validation data for generation 14 : 0.8197160858044628\n",
      "Best combined performance on test and validation data for generation 15 : 0.8229444170300143\n",
      "Best combined performance on test and validation data for generation 16 : 0.8215547944162853\n",
      "Best combined performance on test and validation data for generation 17 : 0.8227853897575441\n",
      "Best combined performance on test and validation data for generation 18 : 0.8226017153301255\n",
      "Best combined performance on test and validation data for generation 19 : 0.82386290167348\n",
      "================= Best result: 0.8255087860684679 =================\n",
      "================= Execution time in minutes: 203.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Genetic Algorithm\n",
    "best_columns_GA_HotelCancel_lightgbm = fsObj.GeneticAlgorithm(generations=20,\n",
    "                                                             population=75,\n",
    "                                                             run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_GA_HotelCancel_lightgbm.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_GA_HotelCancel_lightgbm, output_file)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.83\n",
      "Average Test and Validation Cost: 0.8 Average External Test Cost: 0.83\n",
      "Hotel Cancellations dataset - Particle Swarm Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_PSO_HotelCancel_lightgbm,\n",
    "                                                  cost_funct=cost_funct)\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Particle Swarm Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.83\n",
      "Average Test and Validation Cost: 0.82 Average External Test Cost: 0.83\n",
      "Hotel Cancellations dataset - Ant Colony Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1)#,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_ACO_HotelCancel_lightgbm,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Ant Colony Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.83\n",
      "Average Test and Validation Cost: 0.83 Average External Test Cost: 0.83\n",
      "Hotel Cancellations dataset - Simulated Annealing\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=200,random_state=1,n_jobs=-1)\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_SA_HotelCancel_lightgbm,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Simulated Annealing'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.83\n",
      "Average Test and Validation Cost: 0.83 Average External Test Cost: 0.83\n",
      "Hotel Cancellations dataset - Genetic Algorithm\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_GA_HotelCancel_lightgbm,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Genetic Algorithm'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Xgboost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model_use = LGBMRegressor(random_state=1,n_jobs=-1,n_estimators=100)\n",
    "model_use = XGBClassifier(n_estimators=200,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "# model_use = RandomForestRegressor(random_state=1,n_jobs=-1,n_estimators=200)\n",
    "\n",
    "fsObj = FeatureSelection(columns_list = tree_features_all,\n",
    "                         data_dict = data_dict_for_tree,\n",
    "                         x_validation_dataframe = x_validation_tree,\n",
    "                         y_validation_dataframe = y_validation_tree,\n",
    "                         model = model_use,\n",
    "                         cost_function = cost_funct,\n",
    "                         cost_function_improvement = 'increase',\n",
    "                         average = None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.8244271829323939\n",
      "Best combined performance on test and validation data for iteration 1: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 2: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 3: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 4: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 5: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 6: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 7: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 8: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 9: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 10: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 11: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 12: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 13: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 14: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 15: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 16: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 17: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 18: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 19: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 20: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 21: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 22: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 23: 0.8421729271692419\n",
      "Best combined performance on test and validation data for iteration 24: 0.8421729271692419\n",
      "================= Best result: 0.8421729271692419 =================\n",
      "================= Execution time in minutes: 195.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Particle Swarm Optimization\n",
    "best_columns_PSO_HotelCancel_xgboost = fsObj.ParticleSwarmOptimization(iterations = 25,                                                                   \n",
    "                                                                       swarmSize = 75,\n",
    "                                                                       run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_PSO_HotelCancel_xgboost.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_PSO_HotelCancel_xgboost, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.8367793318458914\n",
      "Best combined performance on test and validation data for iteration 1: 0.8521934941187854\n",
      "Best combined performance on test and validation data for iteration 2: 0.8552180885138283\n",
      "Best combined performance on test and validation data for iteration 3: 0.8570069898519017\n",
      "Best combined performance on test and validation data for iteration 4: 0.8570069898519017\n",
      "Best combined performance on test and validation data for iteration 5: 0.8570069898519017\n",
      "Best combined performance on test and validation data for iteration 6: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 7: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 8: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 9: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 10: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 11: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 12: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 13: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 14: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 15: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 16: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 17: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 18: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 19: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 20: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 21: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 22: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 23: 0.8573772222260324\n",
      "Best combined performance on test and validation data for iteration 24: 0.8573772222260324\n",
      "================= Best result: 0.8573772222260324 =================\n",
      "================= Execution time in minutes: 197.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Ant Colony Optimization\n",
    "best_columns_ACO_HotelCancel_xgboost = fsObj.AntColonyOptimization(iterations = 25,\n",
    "                                                                N_ants = 75,\n",
    "                                                                run_time=1200,\n",
    "                                                                evaporation_rate=0.8,\n",
    "                                                                Q=0.2)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_ACO_HotelCancel_xgboost.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_ACO_HotelCancel_xgboost, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.7167448383370251\n",
      "Best combined performance on test and validation data for iteration 1: 0.7235161125829991\n",
      "Best combined performance on test and validation data for iteration 2: 0.7325854822984283\n",
      "Best combined performance on test and validation data for iteration 3: 0.7679653693544285\n",
      "Best combined performance on test and validation data for iteration 4: 0.7982988050529538\n",
      "Best combined performance on test and validation data for iteration 5: 0.8298992154982374\n",
      "Best combined performance on test and validation data for iteration 6: 0.8421314966237313\n",
      "Best combined performance on test and validation data for iteration 7: 0.8495068538354976\n",
      "Best combined performance on test and validation data for iteration 8: 0.8629978584506649\n",
      "Best combined performance on test and validation data for iteration 9: 0.8637098688764088\n",
      "Best combined performance on test and validation data for iteration 10: 0.8637098688764088\n",
      "Best combined performance on test and validation data for iteration 11: 0.8648029523819165\n",
      "Best combined performance on test and validation data for iteration 12: 0.8729971673378925\n",
      "Best combined performance on test and validation data for iteration 13: 0.8729971673378925\n",
      "Best combined performance on test and validation data for iteration 14: 0.8731238970027061\n",
      "Best combined performance on test and validation data for iteration 15: 0.8731238970027061\n",
      "Best combined performance on test and validation data for iteration 16: 0.8737583653817141\n",
      "Best combined performance on test and validation data for iteration 17: 0.8737583653817141\n",
      "Best combined performance on test and validation data for iteration 18: 0.8737744582780842\n",
      "Best combined performance on test and validation data for iteration 19: 0.8748052185607916\n",
      "Best combined performance on test and validation data for iteration 20: 0.8748052185607916\n",
      "Best combined performance on test and validation data for iteration 21: 0.8748052185607916\n",
      "Best combined performance on test and validation data for iteration 22: 0.8748052185607916\n",
      "Best combined performance on test and validation data for iteration 23: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 24: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 25: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 26: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 27: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 28: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 29: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 30: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 31: 0.8754914817670031\n",
      "Best combined performance on test and validation data for iteration 32: 0.876341811379094\n",
      "Best combined performance on test and validation data for iteration 33: 0.876720141909787\n",
      "Best combined performance on test and validation data for iteration 34: 0.876720141909787\n",
      "================= Best result: 0.876720141909787 =================\n",
      "================= Execution time in minutes: 225.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Simulated Annealing\n",
    "best_columns_SA_HotelCancel_xgboost = fsObj.SimulatedAnnealing(temperature = 1500,\n",
    "                                                               iterations = 35,\n",
    "                                                               n_perturb = 75,\n",
    "                                                               run_time=1200,\n",
    "                                                               n_features_percent_perturb=1,\n",
    "                                                               alpha=0.9)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_SA_HotelCancel_xgboost.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_SA_HotelCancel_xgboost, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for generation 0 : 0.8335388191254951\n",
      "Best combined performance on test and validation data for generation 1 : 0.8426055067319773\n",
      "Best combined performance on test and validation data for generation 2 : 0.8390713701643799\n",
      "Best combined performance on test and validation data for generation 3 : 0.8386006049279884\n",
      "Best combined performance on test and validation data for generation 4 : 0.8468804107743809\n",
      "Best combined performance on test and validation data for generation 5 : 0.8518674037953382\n",
      "Best combined performance on test and validation data for generation 6 : 0.8482528559153735\n",
      "Best combined performance on test and validation data for generation 7 : 0.8500596380719708\n",
      "Best combined performance on test and validation data for generation 8 : 0.8535473531008269\n",
      "Best combined performance on test and validation data for generation 9 : 0.856882402507962\n",
      "Best combined performance on test and validation data for generation 10 : 0.8545666423636407\n",
      "Best combined performance on test and validation data for generation 11 : 0.8584787270521124\n",
      "Best combined performance on test and validation data for generation 12 : 0.858628480645379\n",
      "Best combined performance on test and validation data for generation 13 : 0.8589812233378129\n",
      "Best combined performance on test and validation data for generation 14 : 0.8639789214681569\n",
      "Best combined performance on test and validation data for generation 15 : 0.8602024619873484\n",
      "Best combined performance on test and validation data for generation 16 : 0.8584137203615324\n",
      "Best combined performance on test and validation data for generation 17 : 0.8597112151270536\n",
      "Best combined performance on test and validation data for generation 18 : 0.8578861191070599\n",
      "Best combined performance on test and validation data for generation 19 : 0.858616572227821\n",
      "================= Best result: 0.8639789214681569 =================\n",
      "================= Execution time in minutes: 507.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Genetic Algorithm\n",
    "best_columns_GA_HotelCancel_xgboost = fsObj.GeneticAlgorithm(generations=20,\n",
    "                                                             population=75,\n",
    "                                                             run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_GA_HotelCancel_xgboost.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_GA_HotelCancel_xgboost, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.64\n",
      "Average Test and Validation Cost: 0.84 Average External Test Cost: 0.88\n",
      "Hotel Cancellations dataset - Particle Swarm Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABQ0AAAKrCAYAAABbZqvkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/av/WaAAAACXBIWXMAAA9hAAAPYQGoP6dpAABovUlEQVR4nO3dd5hU5fk/4GepIqAiGjU2ULNLd1EEARtIMfbewAaiErFixRYxKnYFpFlAsaOoUaOJjYi9d8VKETsoUkTa+/uD3044ssAu7ALr976vi0v3zCnvvM+cmXc+c0peSikFAAAAAMD/V2lVNwAAAAAAWL0IDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCJRaSqlCrnt12ibF+6PV4o/2fACAVauijy18j4CKRWjIauucc86JgoKCzL+GDRvGdtttF0ceeWQ8//zzK60tX331VRQUFMQtt9xS4mXOOeecaN68eTm2anHDhg2LHXbYIZo2bRp9+vQpl2088MADcf7556926z7iiCNizz33XKnbXBXat28fxx9//ErZ1u9fwyXp4+XZVyIifvnllzjvvPPiqaeeKtX2Vobl7fNhw4bFddddVw4tAoDFGTuX3soYO5eVzz77LI444oj46aefVnVTFvPKK69EQUFBPPHEE0ucp7ixXllZkXWX5dgV/oiqrOoGwNKsscYacdttt+X+nj9/fnz//fcxaNCg6NGjR9x6663RunXrcm/Hn/70p7j33nvjz3/+c4mX+dvf/haHH354ObYq69tvv41rrrkmdtlllzj22GNjvfXWK5ft3HjjjfGXv/ylwq17ddpmRXXRRRfFvHnzymXdH330Udx///2x4447rpTtrQzXXHNNdOnSZVU3A4D/Q4ydS25ljZ3LyuOPPx6vvvrqqm7GciturFcR1r06bRNWBaEhq7VKlSpFYWHhYtObNGkSnTp1WmkDn2rVqhXbjqXZbLPNYrPNNiufBhXj559/joiITp06xXbbbbfStsv/HVtttdUfensAUNEZO5ecsTPAsjk9mQpp0003jTp16sTEiRMj4n+HxN95553RuXPnaNasWe4Q8ylTpsT5558fbdu2jSZNmsRee+0VDz300GLrfOqpp+LQQw+N5s2bR5s2beL000+PyZMnR0Txh63fddddsddee8XWW28dLVu2jBNOOCE+/vjj3OPFnWLxxhtvRLdu3aJVq1bRvHnzOOqoo+L111/PzFNQUBC33npr9OvXL3e6xKGHHhpvvvnmEvtjwIABsc8++0RERJ8+faKgoCC++uqriIj45JNPolevXrHDDjtEs2bN4uCDD17sMPr27dvHhRdeGD179ozCwsLYf//9i70+R0FBQUyePDnGjBmT2UZJ+vjrr7+OE088Mdq2bRtNmzaN3XbbLQYPHhzz589f6rqL884778SRRx4ZzZs3jx133DGGDRu22Dwppbj99ttjv/32i8LCwmjSpEn89a9/jREjRizz+bz99ttxwgknxPbbbx+NGzeONm3axJlnnhlTpkxZYpsiIubMmRMDBw6MPfbYI5o1axbNmjWLfffdNx555JHcPEWv1eeeey5OPvnk2GabbWKbbbaJk08+Ob777rvM+v773//GQQcdFFtvvXV06NAh7r///qVuPyLi2GOPjfbt2y9Wv1GjRkVBQUF8+umnERHx+eefx+mnnx477LBDNG7cOFq1ahUnnnhijB8/fonr/v3pwimlGDFiRHTq1CmaNm0aBx54YLzzzjuLLffdd9/FBRdcEO3atYsmTZrEtttuG8ccc0y8++67ERExevToOPLIIyMi4pRTTokjjjii2O1FRDzzzDNx2GGHRYsWLaJFixbRs2fPGDduXO7xon31oYceij59+kSrVq1i6623ju7du8fnn3++zP4rSZ8vq85FbYiIuPPOO3P/X7T+o48+Orbbbrto3Lhx7LzzztG3b9+YOXPmMtsGAMvL2DlrZY2dIyLmzZsXgwYNik6dOkWTJk2iXbt2cd1118WcOXMiImLWrFnRuXPnaNu2bS7IjIgYPHhwFBQUxOOPPx7nnHNODBw4MCIiWrduHQMGDMj065577hlNmzaNHXbYIfr27RszZszIPT569OgoKCiIBx54IHbZZZdo3rx5PPbYYzFgwIBo1apVvP7663HIIYdEs2bNom3btnHFFVfE3Llzc8uXZHy7LEsa60VEvPTSS3HEEUdEYWFhbLPNNtGzZ8/44osvMsv/+9//jv333z+aN28e22yzTRx11FHxyiuvLHPdv1feY9eSfP+ACifBaurss89OhYWFxT42ZcqU1KBBg3T00UenlFJ6+eWXU35+fmrRokV64IEH0pNPPpm+/PLLNG3atNSxY8fUtm3bdPfdd6fnnnsuXXDBBSk/Pz8NHz48t77Ro0en/Pz8dOKJJ6ZnnnkmPfbYY6lDhw6pQ4cOadasWWnSpEkpPz8/3XzzzSmllB577LGUn5+frrvuuvTyyy+nJ554Iu2+++6pVatWaebMmcW2/7HHHksFBQWpW7du6cknn0xPPPFEOvTQQ1OjRo3SmDFjcvMVPY/jjz8+jRkzJj3++OOpXbt2qWXLlmn27NnF9sc333yTHnzwwVyb3nrrrfTbb7+l119/PTVp0iTtt99+6bHHHkvPPPNMOuGEE1J+fn669957c8u3a9cuNWrUKPXp0ye9+OKL6Zlnnil2O2+99VZq27ZtOvTQQ3PbKEkfz58/P+2+++5pzz33TP/+97/Tyy+/nK6++uqUn5+fhg4dusR1F2fcuHGpadOm6YADDkhPPvlkevTRR1OnTp1So0aN0h577JGb79prr02NGjVKgwYNSi+//HJ66qmnUrdu3VJ+fn7673//u8RtfvLJJ6lx48bp2GOPTc8++2x68cUX04ABA1LDhg3TqaeeWmybipx++umpsLAwjRw5Mr3yyivp8ccfT/vvv38qKChIn376aUrpf6/Vli1bpn79+qUXX3wxjRw5MjVt2jT16NEjt67nnnsuNWjQIHXv3j09++yzadSoUalt27apYcOG6bjjjltiG4pem6+++mpm+uGHH54OOOCAlFJKP/74Y9puu+3SQQcdlP7zn/+kl19+OQ0fPjwVFhamgw46KLfM71/DXbt2XayPCwoK0qWXXprGjh2bBgwYkJo1a5bZV2bPnp06dOiQOnfunB555JH0yiuvpHvuuSe1adMm7bzzzmnOnDlpypQp6Y477kj5+flpyJAhub76/fZuvvnmlJ+fn04//fT07LPPpn/+859p9913T4WFhenDDz9MKaXcvtqiRYt0zjnnpLFjx6bRo0enli1bZtZVnJL2+bLq/Ntvv6W33nor5efnp5NOOim99dZbKaWUxo4dmwoKCtKZZ56Zxo4dm55//vn0j3/8I+Xn56errrpqqW0DgGUxdl79xs4ppXTSSSelxo0bpxtuuCE9//zzaejQoalp06bpb3/7W26ed955JzVq1Cj17t078/cFF1yQUkppwoQJqU+fPik/Pz8999xz6ZtvvkkppdSvX79UUFCQLrnkkjR27Nh0xx13pJYtW6YDDzwwN5Z+4IEHUn5+ftppp53SE088kf75z3+mKVOmpP79+6fGjRunHXfcMY0YMSK99NJLqW/fvik/Pz/ddNNNubaVZnz7+OOPL/H1V9xY76mnnkoNGjRI3bp1S0899VR69NFH0957751atmyZJk6cmFJK6fXXX08NGzZMF1xwQXrxxRfTs88+mw499NDUuHHjNHny5CWuuzjlPXYtyfcPqGiEhqy2igYOc+fOzf2bPn16euedd9KRRx6Z8vPz01NPPZVS+t8H1SWXXJJZR//+/VODBg3SBx98kJl+wQUXpMLCwjRt2rS0YMGCtOOOO6aDDz44M88777yT2rVrl1555ZXFBj4XXnhhKiwszAxEPvzww3T11Ven7777LtP+lFJuG/vss0+aP39+bpnffvstdezYMXXq1Ck3LT8/P3Xu3DkzX9HA7Pnnn19if3344YcpPz8/PfDAA7lphxxySGrbtm1uMFakS5cuadttt02zZs1KKaXcwGru3LlLXH+Rdu3aZQKUkvTxDz/8kPLz89ONN96YmWfIkCHpkUceWeK6i3PaaaelbbfdNv3888+5aV999VVq3LhxJhTq3bt3uuGGGzLLTp06NeXn56d+/fotcZsPPfRQ6tq1a/r1118zyx5//PGpZcuWS2zXb7/9lo499th01113Zaa/9957KT8/P40cOTKl9L/X6nnnnZeZ75xzzkkFBQW5Ad7BBx+cdt1110xN3njjjZSfn7/UPvrtt99Sy5Yt0/nnn5+bNnHixFRQUJDuvPPOlFJKzz//fDr00EPTt99+m1m2aKA4Y8aMlNLSQ8Pp06enxo0bpzPPPDOzjsGDB2f2lY8++ih16dIlvfvuu5n5brnllpSfn5/GjRuX6ZdFB5uLbm/atGmpWbNm6YQTTsisZ+rUqWm77bZLxxxzTErpf6Fh0ZeiIgMGDEj5+fnpyy+/XGLflaTPS1rnlBbuyxdffHHmOffs2XOx7e65555pv/32W2K7AKAkjJ1Xv7FzUT+PGDEiM/2RRx5ZrH1FY5XHH388dezYMe25556Z/urfv3/Kz89PU6ZMSSktHN81aNAgXXrppZl1v/nmm5mQsyg0XDT0XXR9Dz/8cGZ6+/bt0yGHHJJSKv34dkmhYXHzLFiwILVr1y7tt99+mdpNnz49tWrVKp1++ukppZSGDh2a8vPzc0FpSil9/fXXqV+/fumzzz4r8fZXxti1pN8/oCJxTUNWa7NmzYrGjRsvNn2jjTaKf/zjH7Hrrrtmpjdq1Cjz9wsvvBCbbLJJ5OfnZ26osOuuu8a9994br732WmyxxRbx3XffxTHHHJNZtlmzZvHMM89ERCx2qmzr1q3jnnvuiT322CM6duwYbdq0iVatWkXv3r2LfR5ffPFFfPfdd3HkkUdGpUr/uypAtWrVYs8994wbb7wxvv7669zFops3b56Zb8MNN4yIKNUpjL/++mu8/fbbcdhhh8Waa66ZeWy//faLPn36xPvvv5+7hkt+fn5UqVL6t4SS9HH79u0jPz8/Bg4cGB988EG0adMmdtppp+W6I+2rr74aLVu2jLXXXjs3beONN47CwsLMKR1XX311RERMmzYtxo8fH5MmTYr3338/IiJ3Okhx9tlnn9hnn31izpw58fnnn8fEiRPjs88+i88//3ypy1WrVi1uuummiFh4Wk/RNotOnfj9sttuu23m7w033DBSSjFr1qxYsGBBvPPOO3HEEUdkarLNNtss84Li1apVi7322isefvjhuOCCC6JatWrx8MMP515rERFt27aNtm3bxvz582P8+PExceLE+OKLL3Kn8cyZMydq1qy51O289dZbMXfu3OjUqVNm+h577JG5Y3CDBg3ijjvuiJRSfPXVVzFx4sQYP358PPvss8X2y9K2N3v27Nh7770z0+vUqRPt2rWLxx57LHeqe0Tx/Rux5H1o9uzZJerz0tZ5Ud26dYtu3brF7Nmzc8t98sknMWXKlKhbt25JugEAlsrYeaHVZez8wgsvREREhw4dMv25yy67RKVKleK5556Ltm3bRkREz5494/nnn4/TTjstqlevHg888EBUr159iet+8cUXY8GCBYutu2nTprH++uvHc889FwcffHBuesOGDYtdT3Fjpl9++SUiVmzcsyzjx4+PyZMnx/777x8LFiyIBQsWRMTCm/m0bt06xo4dGxERrVq1isqVK8eBBx4YnTp1ijZt2kTr1q3j7LPPLtX2VsbYdXm/f8DqTGjIam2NNdaIO+64I/d31apVo06dOrHBBhsUO//vP+CnTp0aEydOLHbwFLHwrml16tSJiCjVl/bddtstBg4cGHfeeWeMHDkybr311qhZs2bsv//+cdZZZ0W1atUy8xeFWeuvv/5i6yqaNn369Ny0GjVqZOYpGgQVfZiWxC+//BIppRJv8/d9V1Il6eO8vLwYMWJEDB06NJ588sncdWEaN24cffr0iRYtWpR4ez///HOsu+66i01ff/31M6Hhxx9/HP/4xz/itddei6pVq0a9evVim222iYhY4jVnIhZ+oF9xxRXxwAMPxK+//hobbLBBNGrUqET988orr8Tll18eH330UVSvXj223HLLyM/PL3abS6vxtGnTIqW0xOe5LAcccECMHDkynn322ejcuXP885//jA4dOsRaa62Va8vAgQNj5MiRMW3atFh33XWjYcOGuTYtrX+KFPV10f5T5E9/+tNi8955550xZMiQ+P7772OttdaKgoKCUm0rYuHgK2LJ+9DcuXPj119/zU0r7T5Umj4vTZ0X9csvv8TFF18c//73v2PevHnx5z//OZo2bRprrLFGifsBAJbG2Hmh1WXsPHXq1IhYeA3E4nz77be5/69cuXLst99+8dZbb8WWW24Z9evXL9G6l3QNv0XXvbT2Ftd3i45LlnfcsyxF7R8wYEDmGo2L+vXXX2PrrbeOESNGxPDhw2P06NFx5513RvXq1WO33XaLCy64IGrXrl2i7a2Msevyfv+A1ZnQkNVapUqVomnTpsu9fNGb/KWXXlrs43/+85/jp59+ioj/fXAtasyYMbkPxd/r2LFjdOzYMWbPnh1vvvlmjB49OkaOHBkbbbRRdO/ePTPvOuusExERP/zww2LrKbr5xe8/wFZU7dq1o1KlSuW+zZL0ccTCgWWfPn2iT58+MXHixBg7dmwMHTo0TjzxxBg7duxig8UlqVOnTvz444+LTS+qY0TEjBkzolu3brHhhhvGAw88EAUFBVG1atWYNWtW3HvvvUtd/6WXXhoPPvhg9O3bN3bdddfcQOTkk0/OXTy8OJMmTYrjjz8+WrRoEY899lhsscUWUalSpfj000+LvXj4sp5jpUqVlvg8l1W3hg0bRsOGDePRRx+NDTbYICZMmBAXXnhh7vGbbropBg4cGH369Im99torF5RdccUV8cYbb5SojUXL/L6Ni9YhIuJf//pX9O3bN3r06BFHHHFE7kvLnXfeGc8991yJthURuSNLl/R6rlatWtSqVSsTHJdGSft8Rep8+umnx/vvv5+78HjR4P2AAw6I3377bbnaDQCLMnZefuUxdl5rrbUiLy8v7rrrrqhatepijy965sx3330X1157bTRq1Cg++OCDuPnmm+O4445b6rojIgYOHJg7snJRvw8Dl0dZjm9/r+i5n3DCCdGhQ4di5yn6ftCyZcto2bJlzJ07N95555144okn4o477ohatWplxrhLU95j1xX5/gGrM3dP5g+tVatWMWnSpPjTn/4UTZs2zf2bMGFC9O/fP6ZNmxZbbLFFrLfeevHkk09mlv3oo4/i+OOPzx1+v6g+ffrEQQcdFCmlWGONNaJNmzbRr1+/qFy5cnz99deLzV+/fv3YYIMN4tFHH8384jl37tx47LHHon79+sX+yrUi1lxzzdh6663jySefjFmzZmUee/jhh6N27dpL/BV5aRY99SOiZH388ccfx4477hj/+c9/IiJis802iy5dusSBBx4YP//8c+7Ukd+vuzht27aNV155JfOBP2XKlMydz7744ouYMmVKHHbYYdGkSZPcIK3otIJFa/D7bb722mvRpEmT2HfffXOB4fTp0+ONN95Y6q/V7733Xvz666/RvXv32GqrrXLrLW6by1KtWrVo0aJFPPXUU5kwady4cUu9q/SiDjjggBg7dmyMHj06Ntxww2jTpk3mOW644YZx1FFH5QZQc+fOjeeffz4iSvZLaPPmzaNGjRrx2GOPZaY//fTTmb9fffXVqFSpUpx88smZoxyK+qVoW5UrV17m9tZYY4345z//mZn+888/x5gxY6Jly5bLbPPSlLTPS1Pn4l5bO+64Y7Rr1y4XGE6ePDnGjRtXqtcHAJQXY+eyHTu3atUqUkrx888/Z/qzZs2acfXVV+fuHp1SinPOOSciFv64e+ihh0b//v3jgw8+yK2ruDF4xMIjChdd9yabbBLXXnttvPbaa6Xug98ry/Ht78d6W2yxRay//vrxxRdfZNrftGnTuPPOO+OBBx6IypUrx4ABA6J9+/YxZ86cqFq1arRo0SLOP//82GijjXKvnWWNIyPKf+xamu8fUJE40pA/tG7dusUjjzwSRx11VBx33HGx8cYbx7vvvhsDBw6MRo0aRb169aJSpUpx+umnR58+feL000+PffbZJ2bMmBE33HBD/OUvf4nOnTsv9kvqDjvsEA888ECceeaZsc8++0ReXl7cf//9ERHx17/+dbF2VKpUKc4+++zo3bt39OjRIw477LBIKcWIESPiq6++ikGDBpXL8z/jjDPi6KOPjiOOOCK6d+8ea665ZowaNSpee+21uPDCC0t8dN+i1lprrfj888/jlVdeiWbNmpWoj1NKsfbaa0ffvn3j559/jnr16sX48ePjrrvuih122CH3q+3v113cL6QnnnhiPPXUU3H00UdHr169Ii8vLwYPHpz5IN5iiy1irbXWihEjRsQ666wTtWrVitdeey1uueWWyMvLy5zG+vttFhYWxsMPPxzDhw+Pxo0bx+TJk+PWW2+NKVOmRETE/Pnzix2YFA0O+vfvH3PmzIkqVarEmDFj4q677oqIyGyzJHr37h1du3aNbt26Rffu3WPGjBlx/fXXl7hme+21V1x55ZVx//33R48ePTIDzcLCwnjuuefiuuuui7Zt28aPP/4YI0eOjE8//TQiFl4PaVmnHK255ppx6qmnxuWXXx7nn39+7LbbbjFu3LjcdW8W3dbdd98dffv2jT333DNmzJgR999/fy6gLBqUF/1a/uKLL8bmm2++2HV31lprrTjxxBPjmmuuid69e8fee+8d06dPjyFDhsRvv/0Wp512Won6ZWlK0uelqfNaa60VH3zwQbz22mux7bbbRmFhYTz99NMxatSo2HzzzePzzz+PYcOGxbx580r9+gCA8mDsXLZj55122il22GGHOPvss+O4446Lpk2bxjfffBMDBgyI2bNn544Kve222+LFF1+Mq666KtZbb70488wz47nnnoszzjgjHnzwwVhjjTVyR+Y9+eST0aZNm9hqq63i4IMPjquuuip+/PHHaNWqVfz8888xdOjQmDhxYpxxxhkr3B9lOb4tbqx31llnxZlnnhm9e/eOPffcMypXrhyjRo2K//znP3HxxRdHRESbNm1i8ODB0bNnz+jatWusscYa8eSTT8bXX3+duybmssaREeU/di3N9w+oSISG/KHVrVs37rvvvrjhhhvimmuuiWnTpsUGG2wQXbp0iRNPPDEXpBxwwAFRq1atGDZsWPTq1SvWXnvt2HHHHeP000+PNddcc7GBz+677x5z5syJ22+/PU455ZRYsGBBNGzYMIYMGbLE6/PtscceUatWrRg6dGj07t07qlSpEoWFhTFy5MhSXdOvNFq0aBF33nln9O/fP84///xIKUWDBg3ixhtvXOJpAMty7LHHxqWXXhrHHntsDB8+PFq0aFGiPr755pvj+uuvj4EDB8bUqVNj3XXXjT333DNOOeWUpa779zbddNO4++6748orr4xzzz031lhjjTj44INjyy23jHHjxkVERK1atWLIkCFx1VVX5a6TU69evbj00kvj0Ucfjddffz1SSpGXl7fYNs8555xIKcVNN90UM2fOjA033DDatWsXRx99dPTp0ydef/313C+7i9pss83ihhtuiP79+8fJJ58cNWvWjK222iqGDRsW/fr1K/WvvYWFhTFixIi47rrr4rTTTot11lknunfvHo8//niJll9nnXVi1113jccffzz233//zGM9evSI6dOnx4MPPhjDhw+P9ddfP1q3bh3HH398HH/88fHaa6/FpptuusxtHH300VGzZs0YPnx4PPzww1GvXr245pprolu3brl59t133/juu+/i3nvvjYcffjjq1q0bhYWFcdddd8Xhhx8er732Wmy//fax1VZbxZ577hkPPfRQvP766/Gvf/1rse0dd9xxscEGG8Rtt90WvXr1ijXXXDO22267uOaaa6KgoKBE/bI0Jenz0tS5Z8+eMWjQoOjRo0c8+uijccUVV8Sll14aV199dcyZMyf+/Oc/x0EHHRSVKlWK66+/PiZNmlSifgeA8mLsXLZj56IftwcPHpzr13XWWSdatWoVJ598cmy00UYxbty4uOaaa6J9+/a5G77VrFkzLrnkkujWrVtcccUVcdFFF0WnTp3ioYceiksuuST233//6Nu3b1x88cWx5ZZbxqhRo+KWW26JWrVqRWFhYVx66aXLdUbR75Xl+La4sd7ee+8da6+9dgwbNixOO+20qFSpUmy11VZx/fXX58LkbbfdNgYPHhxDhw6Ns846K3777bfYcsst48orr8zd5K8k48iI8h+7lvT7B1QkeckVOQEAAACARbimIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAICMKqu6Acvjl19+jfnzF6zqZqwW6tSpGT/9NHNVN4MSUKuKQ60qBnWqONTqfypXrhRrrVVjVTeDMmA8+j/28YpDrSoOtao41KpiUKesko5JK2RoOH/+gpg3zyAtL2/hf+fPXxAprdq2sHRqVXGoVcWgThWHWvFHZTy6kH284lCrikOtKg61qhjUafk5PRkAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZVVZ1AwAAAGBlyMtb1S1YtqI2VoS2RkSktKpbAJQXoSEAAAB/eGuvvWZUq1Z5VTejxOrWrb2qm1Aic+bMj2nTZpX5eitCaFqRAl7hLstDaAgAAMAfWl5eRLVqlePovv+OWbPnrerm/GGsuUaVGHFh58jLK9tQSsBb9sor3OWPTWgIAADA/wmzZs+LX38TGq7OBLxlr7zCXf74hIYAAADAakXAC6ueuycDAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJDh7skAAAAALJe8vFXdgqUrat/q3s4iKa3qFvyP0BAAAACAUlt77TWjWrXKq7oZJVK3bu1V3YQSmTNnfkybNmtVNyMihIYAAAArbHU/gsWRNkBZy8uLqFatchzd998xa/a8Vd2cP4Q116gSIy7sHHl5q8f7oNAQAABgBTjSpuytTkfaAEs3a/a8+PU3oeEfkdAQAABgOTnSpuytbkfaAPxfJTQEAABYQY60AeCPptKqbgAAAAAAsHoRGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMqqs6gYA/3fk5a3qFixbURsrQltTWtUtAAAA4I9KaAisFGuvvWZUq1Z5VTejxOrWrb2qm7BMc+bMj2nTZq3qZgAAAPAHJDQEyl1eXkS1apXj6L7/jlmz563q5vwhrLlGlRhxYefIy3PEIQAAAGVPaAisNLNmz4tffxMaAgAAwOpOaEiFVxGuPVeRrpMX4cg1Vv/Xqn0KAACgfAkNqdBcJ698uFbe/20Vab+yTwEAAJQPoSEVluvklQ/Xyvu/zX5V9uxTAABARSQ0pMJznTwoe/YrAACA/9sqreoGAAAAAACrF0caAkAFVRFuBOOmNQAAUDEJDQGgAqpIN6yJcNMaAACoaISGAFDBuGFN+XDTGgAA+B+hIQBUUG5YAwAAlBc3QgEAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZboQCAACrqby8Vd2CpStq3+reziLujA4AJSc0BACA1dDaa68Z1apVXtXNKJG6dWuv6iaUyJw582PatFmruhkAUCEIDQEAYDWTlxdRrVrlOLrvv2PW7Hmrujl/CGuuUSVGXNg58vIccQgAJSE0BACA1dSs2fPi19+EhgDAyudGKAAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgY7lCw7vvvjs6duwYzZo1i0MOOSTefvvtpc7/xhtvxKGHHhrNmzePjh07xrBhw2L+/PnLs2kAAAAAoJyVOjQcPXp09O3bN/bZZ58YMGBA1K5dO7p37x5ff/11sfNPnDgxjj322Khdu3YMGDAgDj/88Ojfv3/cdNNNK9z48pSXVzH+VbS2AgAAALD6q1KamVNKMXDgwDj88MOjV69eERHRpk2b2G233eK2226Lc889d7FlnnjiiViwYEH0798/atSoETvssENMnDgx7r333jjhhBPK5lmUsbXXXjOqVau8qptRYnXr1l7VTSiROXPmx7Rps1Z1MwAAAABYhlKFhhMmTIjJkydH+/btc9OqVq0au+yyS4wdO7bY0HDOnDlRuXLlqF69em7a2muvHdOmTVuBZpefvLyIatUqx9F9/x2zZs9b1c35w1hzjSox4sLOkZcXkdKqbg0AAAAAS1Oq0HD8+PEREbHZZptlpm+66aZxzz33REop8n53Huree+8dt912W1x77bXRo0eP+OKLL+K+++6LPffcc4UaXl6nuxatd9bsefHrb0LDslaWdXPKc/lSq4qhrPtWrcqPfari0L+UVHm+VrwOy4/PzorDZ2fFoVYVg/e/imN16dtShYYzZsyIiIiaNWtmptesWTPmzZsXs2fPjho1amQe22yzzeKss86KCy+8MHcdw2233TbOO++85W50nTo1lz0Tq6WKcio1alVRqFPFoVYVh1pREsajFZd9vOJQq4pDrSoGdao4VpdalfqahhGx2NGERSpVWvy+Kvfdd19ccMEFcdhhh0Xnzp3j22+/jeuvvz569eoVQ4cOLXaZZfnpp5kxf/6CUi9XEnl5q09x/oimTJleZqcnq1X5UquKoSzrFKFW5ck+VXGU9X61qMqVKwmb/iDKczwaYT8vTz47Kw6fnRWHWlUM3v8qjvIcj0aUfExaqtCwdu2FL4aZM2dGnTp1ctNnzpwZVatWzVy3sMhNN90U7du3j7///e+5aVtuuWUcdNBB8eyzz8auu+5amibkuC5exZSS2lUUalUxqFPFoVYVh1pRUl4nFZN9vOJQq4pDrSoGdao4Vpdaleowv8033zwiIiZNmpSZPmnSpKhfv36xy3zzzTdRWFiYmda0adOoWrVqfP7556XZPAAAAACwEpQqNKxXr15stNFG8fTTT+emzZ07N8aMGROtW7cudpn69evHm2++mZn2wQcfxNy5c2PjjTdejiYDAAAAAOWpVKcn5+XlRY8ePeKSSy6JWrVqRfPmzeOOO+6IadOmxVFHHRURERMnToypU6fmji7829/+FqeeempceOGF8de//jW+/fbb6N+/f/zlL3+Jjh07lvkTAgAAAABWTKlCw4iILl26xOzZs2PkyJExfPjwaNiwYdxyyy25owYHDRoUDz74YIwbNy4iIv76179GpUqVYvDgwTF69OhYf/31Y6eddorTTz89qlWrVrbPBgAAAABYYaUODSMiunfvHt27dy/2sX79+kW/fv0y0zp37hydO3denk0BAAAAACtZqa5pCAAAAAD88QkNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKWKzS8++67o2PHjtGsWbM45JBD4u23317q/D/++GOcccYZsd1220XLli3j5JNPjm+//XZ5Ng0AAAAAlLNSh4ajR4+Ovn37xj777BMDBgyI2rVrR/fu3ePrr78udv45c+bEMcccEx9++GFcdtll0bdv3/j444+jZ8+ekVJa4ScAAAAAAJStKqWZOaUUAwcOjMMPPzx69eoVERFt2rSJ3XbbLW677bY499xzF1vmoYceiq+++ioef/zx2HDDDSMiom7dutG7d+8YP3581K9fvwyeBgAAAABQVkp1pOGECRNi8uTJ0b59+9y0qlWrxi677BJjx44tdpmnnnoqdtppp1xgGBGx3XbbxXPPPScwBAAAAIDVUKlCw/Hjx0dExGabbZaZvummm8aECROKPd34k08+ifr168cNN9wQbdq0iaZNm8aJJ54YP/744/K3GgAAAAAoN6U6PXnGjBkREVGzZs3M9Jo1a8a8efNi9uzZUaNGjcxjU6dOjfvvvz823XTTuPzyy+OXX36JK6+8Mnr16hV333135OXlLVfDl3OxVbZeFirL/lWr8qVWFUNZ961alR/7VMWhfymp8nyteB2WH5+dFYfPzopDrSoG738Vx+rSt6W+pmFELDHoq1Rp8QMX582bF/PmzYubbropatWqFRELr2l4zDHHxMsvvxytW7cubZujTp2ay56J1VLdurVXdRMoIbWqGNSp4lCrikOtKAnj0YrLPl5xqFXFoVYVgzpVHKtLrUoVGtauvbDRM2fOjDp16uSmz5w5M6pWrRrVq1dfbJk111wzCgsLc4FhRETLli2jSpUq8cknnyxXaPjTTzNj/vwFpV6uJPLyVp/i/BFNmTI9yuqm2WpVvtSqYijLOkWoVXmyT1UcZb1fLapy5UrCpj+I8hyPRtjPy5PPzorDZ2fFoVYVg/e/iqM8x6MRJR+Tlio03HzzzSMiYtKkSbHJJpvkpk+aNGmJNzXZbLPNYu7cuZlpCxYsiAULFiz3qckRUa6dR/lJSe0qCrWqGNSp4lCrikOtKCmvk4rJPl5xqFXFoVYVgzpVHKtLrUp1I5R69erFRhttFE8//XRu2ty5c2PMmDFLPGKwbdu28cYbb2RufPLCCy/EggULonnz5svZbAAAAACgvJTqSMO8vLzo0aNHXHLJJVGrVq1o3rx53HHHHTFt2rQ46qijIiJi4sSJMXXq1CgsLIyIiKOOOioeeOCBOPbYY+Okk06K6dOnx5VXXhk777xzNG3atMyfEAAAAACwYkoVGkZEdOnSJWbPnh0jR46M4cOHR8OGDeOWW26JjTfeOCIiBg0aFA8++GCMGzcuIiLWW2+9uOuuu6Jfv35xxhlnRLVq1WK33XaLs88+u2yfCQAAAABQJkodGkZEdO/ePbp3717sY/369Yt+/fplptWrVy+GDBmyPJsCAAAAAFayUl3TEAAAAAD44xMaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQsV2h49913R8eOHaNZs2ZxyCGHxNtvv13iZQcOHBgFBQXLs1kAAAAAYCUodWg4evTo6Nu3b+yzzz4xYMCAqF27dnTv3j2+/vrrZS77ySefxJAhQ5aroQAAAADAylGq0DClFAMHDozDDz88evXqFTvvvHMMHjw41llnnbjtttuWuuz8+fOjT58+se66665QgwEAAACA8lWq0HDChAkxefLkaN++fW5a1apVY5dddomxY8cuddkRI0bEzJkzo2vXrsvXUgAAAABgpShVaDh+/PiIiNhss80y0zfddNOYMGFCpJSKXW7ChAkxYMCAuOSSS6JatWrL11IAAAAAYKWoUpqZZ8yYERERNWvWzEyvWbNmzJs3L2bPnh01atTIPJZSivPPPz/22WefaNGiRbz//vsr2OSF8vLKZDUrbb0sVJb9q1blS60qhrLuW7UqP/apikP/UlLl+VrxOiw/PjsrDp+dFYdaVQze/yqO1aVvSxUaFh1JmLeE1leqtPiBi/fcc09MmDAhBg8evBzNK16dOjWXPROrpbp1a6/qJlBCalUxqFPFoVYVh1pREsajFZd9vOJQq4pDrSoGdao4VpdalSo0rF17YaNnzpwZderUyU2fOXNmVK1aNapXr56Z/5tvvomrrroqLr/88lhjjTVi3rx5ueBx3rx5UalSpWKDxmX56aeZMX/+glIvVxJ5eatPcf6IpkyZHks4i73U1Kp8qVXFUJZ1ilCr8mSfqjjKer9aVOXKlYRNfxDlOR6NsJ+XJ5+dFYfPzopDrSoG738VR3mORyNKPiYtVWi4+eabR0TEpEmTYpNNNslNnzRpUtSvX3+x+V966aWYOXNmnHzyyYs91rhx4+jVq1ecdNJJpWlCTnl2HuUnJbWrKNSqYlCnikOtKg61oqS8Tiom+3jFoVYVh1pVDOpUcawutSpVaFivXr3YaKON4umnn47WrVtHRMTcuXNjzJgxseuuuy42f7t27eL+++/PTHvsscdi+PDhcf/998ef/vSnFWg6AAAAAFAeShUa5uXlRY8ePeKSSy6JWrVqRfPmzeOOO+6IadOmxVFHHRURERMnToypU6dGYWFh1KlTJ3Mac0TEG2+8ERERTZs2LaOnAAAAAACUpVKFhhERXbp0idmzZ8fIkSNj+PDh0bBhw7jlllti4403joiIQYMGxYMPPhjjxo0r88YCAAAAAOWv1KFhRET37t2je/fuxT7Wr1+/6Nev3xKXPfroo+Poo49ens0CAAAAACtB6W9dDAAAAAD8oQkNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAxnKFhnfffXd07NgxmjVrFocccki8/fbbS53/2WefjYMOOiiaN28e7du3j3/84x8xY8aM5dk0AAAAAFDOSh0ajh49Ovr27Rv77LNPDBgwIGrXrh3du3ePr7/+utj5X3nllejZs2dstdVWMWDAgOjZs2c89thj0bt37xVuPAAAAABQ9qqUZuaUUgwcODAOP/zw6NWrV0REtGnTJnbbbbe47bbb4txzz11smVtvvTW22WabuPzyy3PTatWqFaeeemp88cUXscUWW6zgUwAAAAAAylKpjjScMGFCTJ48Odq3b5+bVrVq1dhll11i7NixxS7TrFmz6NKlS2Za/fr1IyJi8uTJpW0vAAAAAFDOSnWk4fjx4yMiYrPNNstM33TTTeOee+6JlFLk5eVlHjvxxBMXW8+zzz4bEf8LDwEAAACA1UepQsOim5fUrFkzM71mzZoxb968mD17dtSoUWOp6/j4449j2LBh8de//jU22WSTUjb3f36XTZaZ8lovC5Vl/6pV+VKriqGs+1atyo99quLQv5RUeb5WvA7Lj8/OisNnZ8WhVhWD97+KY3Xp21Jf0zAiFjuasEilSks/2/nTTz+N7t27x4YbbhgXX3xxaTadUadOzWXPxGqpbt3aq7oJlJBaVQzqVHGoVcWhVpSE8WjFZR+vONSq4lCrikGdKo7VpValCg1r117Y6JkzZ0adOnVy02fOnBlVq1aN6tWrL3HZ119/Pf72t7/FuuuuG8OHD4+11157OZsc8dNPM2P+/AXLvfzS5OWtPsX5I5oyZXr8/+x5halV+VKriqEs6xShVuXJPlVxlPV+tajKlSsJm/4gynM8GmE/L08+OysOn50Vh1pVDN7/Ko7yHI9GlHxMWqrQcPPNN4+IiEmTJmVOLZ40adJSr0/4zDPPxKmnnhpbbLFF3HLLLVG3bt3SbLZY5dl5lJ+U1K6iUKuKQZ0qDrWqONSKkvI6qZjs4xWHWlUcalUxqFPFsbrUqlR3T65Xr15stNFG8fTTT+emzZ07N8aMGROtW7cudpl33303TjnllGjSpEnccccdZRIYAgAAAADlp1RHGubl5UWPHj3ikksuiVq1akXz5s3jjjvuiGnTpsVRRx0VERETJ06MqVOnRmFhYUREXHTRRVGlSpU4/vjj47PPPsusr379+it0mjIAAAAAUPZKFRpGRHTp0iVmz54dI0eOjOHDh0fDhg3jlltuiY033jgiIgYNGhQPPvhgjBs3Lr7++uv48MMPIyLiuOOOW2xdN954Y3To0GEFnwIAAAAAUJZKHRpGRHTv3j26d+9e7GP9+vWLfv36RUTEn//85xg3btzytw4AAAAAWOlKdU1DAAAAAOCPT2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAgQ2gIAAAAAGQIDQEAAACADKEhAAAAAJAhNAQAAAAAMoSGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAAAAIENoCAAAAABkCA0BAAAAgAyhIQAAAACQITQEAAAAADKEhgAAAABAhtAQAAAAAMgQGgIAAAAAGUJDAAAAACBDaAgAAAAAZAgNAQAAAIAMoSEAAAAAkLFcoeHdd98dHTt2jGbNmsUhhxwSb7/99lLn//jjj+PII4+M5s2bxy677BI33XTT8mwWAAAAAFgJSh0ajh49Ovr27Rv77LNPDBgwIGrXrh3du3ePr7/+utj5p0yZEt26dYsqVarE9ddfHwcffHBcd911MWLEiBVtOwAAAABQDqqUZuaUUgwcODAOP/zw6NWrV0REtGnTJnbbbbe47bbb4txzz11smTvuuCNSSjFo0KBYY401Yuedd445c+bE0KFD44gjjojKlSuXzTMBAAAAAMpEqY40nDBhQkyePDnat2+fm1a1atXYZZddYuzYscUu89JLL0WbNm1ijTXWyE3r0KFDTJ06NT744IPlbDYAAAAAUF5KFRqOHz8+IiI222yzzPRNN900JkyYECmlYpfZfPPNF5s/IuLLL78szeYBAAAAgJWgVKcnz5gxIyIiatasmZles2bNmDdvXsyePTtq1Kix2DLFzb/o+kqrcuXyu+lzXt7C/zasVydmz5lfbtv5v2aNagtPQ69SpVIUky0vF7UqH2pVMZRHnSLUqjzYpyqO8tqvFlWeYxhWrvKupf287PnsrDh8dlYcalUxeP+rOFbGeDSi5OOYvFTc4YFL8Mgjj8QZZ5wRL7/8ctSpUyc3fdSoUXH++efHu+++G9WrV88s06RJkzj99NOjW7duuWnz5s2Lxo0bx9///vc47LDDSrp5AAAAAGAlKNVPpLVr146IiJkzZ2amz5w5M6pWrbpYYBgRUatWrWLnX3R9AAAAAMDqo1ShYdG1CSdNmpSZPmnSpKhfv36xy9SrV6/Y+SNiicsAAAAAAKtOqULDevXqxUYbbRRPP/10btrcuXNjzJgx0bp162KX2X777eOFF16IX3/9NTftqaeeinXXXTcKCgqWs9kAAAAAQHkp1Y1Q8vLyokePHnHJJZdErVq1onnz5nHHHXfEtGnT4qijjoqIiIkTJ8bUqVOjsLAwIiIOP/zwGDlyZBx33HHRrVu3+Pjjj2PYsGFx1llnRZUqpdo8AAAAALASlOpGKEVuueWWGDlyZPz000/RsGHDOPfcc2PrrbeOiIhzzjknHnzwwRg3blxu/nfffTcuu+yy+OCDD2K99daLrl27Rvfu3cvuWQAAAAAAZWa5QkMAAAAA4I+rVNc0BAAAAAD++ISGAAAAAECG0BAAAAAAyBAaAgAAAAAZQkMAAPiDcs/D1ZfaLNvq1EerU1sonhpVfGpYMiuzn4SGv/Ovf/0rjjzyyGjVqlW0aNEiDjnkkHjsscdyjz/00ENRUFAQr7766hLX0bFjxzjttNOW+PgPP/wQF110Ueyyyy7RpEmTaNu2bZx22mnx2WeflelzWV5fffVVFBQUxFNPPbXC65o8eXL87W9/i2233TbatGkTV155ZcydO7cMWqlWEWVbqyLz5s2LAw88MEaMGFEm61Onsq3Tm2++GUcccUS0aNEidthhhzjrrLPixx9/LINWqlVE2dbqo48+iiOPPDKaN28eO++8cwwYMMD7Xxkqj/e/iIgHH3wwCgoK4quvvirT9UJJjR49OgoKCpb477DDDivxuu69994YMGBAOba2dNq3bx+XXnppieYt2seX9q99+/Yr3KZvvvkmunXrFr/88ssy5z3iiCOW2p5Ro0aVaJu//PJL9O7dO8aNG7eizS8Tr7zyShQUFMRHH31UovkHDBiwzNqUxevuqaeeir///e/Lvbx9aaHVcV8qokYLrc41KolltX3ChAklWs+nn34a3bp1K9O2rYgBAwZEixYtSjz/sj4jCgoK4pVXXlnhdg0cODDuueeeFV5PSVVZaVuqAPr06ROPPvpoHH744dGtW7fIy8uLp59+Ok4//fT45JNP4rTTTotOnTrFxRdfHI8//ni0bNlysXW8/fbbMXHixDj//POL3cbMmTPj8MMPj5o1a8Zpp50WG220UXz//fcxYsSIOOSQQ+K+++6LLbfcsryf6koxZ86c6NatW9SsWTOuvPLK+Oabb+Lqq6+OuXPnxnnnnbdC61ar8jF37tw455xz4r333os999xzhdenTmXr888/j6OPPjratGkT11xzTfzyyy9xww03xLHHHhv3339/VKmy/G/palW2vv322zj66KOjadOmMWDAgJg8eXJceeWVMWPGjDj33HNXaN1qVX5++OGHuPzyy1d1MyAiIoYPHx5rrrnmYtNr1apV4nUMHTo0dt1117Js1krzpz/9Ke69997c32PGjInBgwdn+qVatWorvJ2XXnopXnjhhRLPv91228UZZ5xR7GObbbZZidbx0UcfxaOPPhrHHntsibe7OjnooINixx13zP198cUXR82aNTP9suGGG67wdm677baoXbv2Cq/HvrR67kuLUqPVv0bLctRRR8Xuu+9e7GMbbbRRidbxxBNPxHvvvVeWzVqpLrroopgxY0bu70MOOWSxftlqq61WeDsDBgxY4e8TpSE0/P9GjRoVo0ePjltuuSXatm2bm77zzjtHnTp1YsiQIbH77rtHQUFBdOzYMf7973/H+eefH5UrV86s59FHH431118/dthhh2K385///Ce++uqreOGFF2LdddfNTd9ll12iY8eOMWLEiLjkkkvK50muZP/85z9j8uTJ8cwzz8Sf/vSniIioXr16XHzxxdGzZ8/M8y8NtSofH3/8cVx00UXx5Zdflsn61Kns3XnnnbH++uvHgAEDomrVqhERsfnmm8dBBx0UL730UmYAXxpqVfbuu+++SCnFgAEDokaNGhGxMJC66aab4uyzz45KlZbvQH+1Kl99+/aNNdZYI6ZNm7aqmwLRpEmTWGuttVZ1M1aZatWqRWFhYe7vL774IiJWfb+stdZamXb9X7ThhhtmQsFatWpF7dq1V9t+WdWvmVVtdd2XFrU6tWVVqAg1WpY///nPq+17wMpSXCD4R+gXpyf/fyNGjIh27dplvoQVOfbYY6NLly4xf/78iIjYd999Y8qUKYud9jV//vz417/+FXvttddiX9CKTJkyJSIiFixYkJleq1atOO+88zJf+qdPnx59+/bNnRrWpk2buOCCC2LWrFkR8b/DmJ9++uk48sgjo1mzZtG5c+d47rnn4v33348DDzwwtt566zjooIPik08+ya23ffv2MWTIkDj77LOjsLAwdt555xg6dOhS++fdd9+NI444Ipo1axZt2rSJyy+/PObMmbPUZV566aVo0qRJLjCMiOjQoUPMnTs3Xn755aUuuzRqVfa1iog4++yzo0aNGnH//fcvc96SUKeyr9NWW20V3bp1ywWGERFbbLFFRCy8FMDyUquyr1XXrl1jxIgRucAwIqJq1aoxf/78xZ5/aahV+bz/RSz8dfu1116Lk046qUTzw6p20003RUFBQbz00ku5adddd100adIkPvnkk2jfvn1Mnjw5br/99igoKMjNs6z95Igjjojzzz8/jjnmmGjWrFlcccUVMWDAgNh///3jn//8Z3Tq1CmaNm0ahxxySHzwwQe55VJKMXz48Nhjjz2iadOmsc0220SPHj1i/Pjx5doPX375ZRx//PHRvHnz2G677eLcc8/NBP+zZs2K8847L3bYYYdo1qxZ7LfffvGf//wnIhaeGll0tMZ2221XJqc2zps3L/bee+/o2LFjrl9nz54dnTt3jsMPPzxeeumlOPLIIyNi4fv0Oeeck1t2+PDh0aFDh2jSpEnstddemcsvFL2XFn0OtGjRIt59991o37593HTTTXHBBRfEdtttFy1atIjzzz8/Zs+enVv2u+++i7PPPjvatm0bjRs3jp122imuvvrqmDdv3go/36V56KGHYo899ogmTZpEp06dMkdSRUS888470aVLl2jevHm0bNkyTj755Nx45ogjjohXX301nn766XK/ZIR9aaHVbV9alBottDrXaFkmTZoUhYWFceaZZ+amffbZZ9GkSZO49tprY8CAATFw4MCYPn16FBQUxOjRoyMiYsaMGXHRRRdF69ato1mzZnH00UdnxoujR4+OVq1axbBhw6Jly5bRvn37+OKLL3Jjz27dusXWW28dO++8cwwfPjzTprfffju6d+8e2267bTRp0iR23333eOihh8q1H+bOnRvXXntt7LTTTtG0adM46KCD4vXXX8/M8+CDD+ZeWzvttFNcdtll8dtvv0VE5F7fl19+eZmcsl4iifTtt9+m/Pz8dPfdd5do/vnz56cdd9wxXXDBBZnpzz33XMrPz0/jxo1b4rIffPBBatCgQerQoUO65ZZb0scff5wWLFhQ7LzHHXdcat++fXr00UfTSy+9lAYOHJgaNGiQBg4cmFJKadKkSSk/Pz+1bNkyDR8+PD3//PNpv/32Sy1btkwdO3ZM9913X3r22WfTjjvumA488MDcetu1a5e23Xbb1L179zRmzJh03XXXpYKCgjRs2LDMep988smUUkrjxo1LzZo1S927d0/PPvtsuueee9J2222XTj311KX20/7775/OOuusxaZvs8026cYbb1zqskuiVuVTq6Jli+Tn56fhw4cvc5klUafyq9PvPfzwwyk/Pz+98sorpV42JbVaGbWaMWNGeuaZZ1LLli0X67fSUKvyq9VPP/2U2rRpkx5++OH05JNPpvz8/DRp0qRlLgfl4YEHHkj5+flp6tSpae7cuYv9K9oX582bl/bbb7+0++67p7lz56YPP/wwNWrUKA0ePDiltHA/btu2bTr11FPTW2+9lVIq2X7StWvX1KhRo3TZZZel5557Lr3//vupf//+qXnz5qlz587pscceS08//XTq1KlT6tSpU265m2++OTVt2jTdfvvt6ZVXXkn33Xdf2n777dORRx6Zm6ddu3bpH//4xwr1y7Rp03LTvv/++7T99tunAw44ID355JPp4YcfTrvssks65JBD0vz581NKKfXp0ye1b98+/etf/0ovvfRS6t27d2rQoEH67LPP0pQpU9J1112X8vPz0wsvvJC++eabpbaha9eu6YQTTii2LvPmzcvN9+6776aGDRumQYMGpZRS6tevX9p6663T+PHj0/Tp09Mdd9yR8vPz04MPPpgmTJiQUkrp+uuvT40bN04DBgxIzz33XLrwwgtTgwYN0pgxY1JK/3vPa9WqVfrXv/6VHnzwwTR37tzUrl27tM0226QzzzwzPf/88+nWW29NDRo0SEOGDEkpLfws2GuvvdK+++6bnnzyyfTiiy+mSy65JLf9lFJ6+eWXU35+fvrwww+XqzZdu3ZNPXv2zEwbNWpUKigoSP/4xz/S2LFj07XXXpsaNGiQ7rrrrpRSSr/88ktq2bJlOu2009KLL76YnnjiidShQ4d08MEHp5RS+vTTT9O+++6bunTpkt56663022+/lbpd9qWl98uq3JfUqOLUqCTy8/PTrbfeWmwNi9qWUkrDhw9P+fn56fXXX0/z5s1LBx10UNpzzz3Tb7/9lr755pvUp0+fVFhYmN566600ZcqUNH/+/NS1a9fUpk2bdN9996Wnn346de3aNbVs2TJ9++23ub5q2LBh2nfffdPYsWPTv//978z75cCBA9OLL76YzjnnnJSfn5/efPPNlFJKkydPTltvvXU69dRT0wsvvJCeffbZdMwxx6SGDRvmxoD9+/dP22677Qr1y++/U5955pmpsLAwjRgxIo0ZMyaddNJJqWnTprn331dffTU1aNAgDRo0KL3yyivpzjvvTE2bNk1XX311Simlt956K+Xn56dLL700ffDBB8vdttIQGqaU3nnnnZSfn5/++9//lniZK6+8MrVq1SrNnTs3N+2ss85K++233zKX/ec//5latmyZ8vPzc1+kzjzzzPTRRx/l5pk9e3Y65phj0vPPP59ZtkuXLqlbt24ppf8NHvr27Zt7/LHHHkv5+fnptttuy0275ZZbUsOGDXNtbdeuXWrXrl2aM2dObp7zzjsvtWrVKs2fP3+xL2KnnHJK6tSpU+a5jhkzJuXn52fa/HudOnVKF1988WLTd9xxx3TFFVcss5+Ko1blU6vfW9HQUJ1WTp2++eab1LZt23TwwQcvMdBZFrUq31otWLAgNW7cOOXn56f27dunr776apnLLIlalV+tzjzzzNS9e/eUUhIassoVfVlc0r9HH300N+8HH3yQGjVqlG655Za03377pQMOOCATXv3+S2tJ9pOuXbvm9rMi/fv3XyxUeuihh1J+fn4u9LrkkkvS0KFDM8/lhhtuSI0bN15ie5anXxb9En3llVemli1bZqZ98sknqaCgIPfe0Llz58yPJ3PmzEmXX355+vjjj5e43iXp2rXrEuvSpEmTzLyXX3552nrrrdNTTz2VGjZsmHm/+31I9/PPP6emTZumAQMGZNZxyimnpH322Sel9L/30ssvvzwzT7t27dLee++dmXbUUUelww47LKWU0tdff526du2aPvnkk8w8u+yyS65fyjo0nD9/fmrbtm0655xzMvNdeeWVafvtt09z5szJffEt+gKfUkqvvfZauuGGG3KvveLCyNKwLy29X1blvvT7tqhR8f2yOtSoJJZWw6LxYEoL3xsOOOCAtO+++6Zbb701NWrUKL3//vu5x38f0hXV69VXX81N++2339LOO++cLr300sxzevrpp3PzFL1fLtr/c+bMSc2aNUvXXXddbt3HHHNM5vUxceLElJ+fnx555JFi27M8/bLod+rPPvss84NNSgu/Kxx00EHpb3/7W0oppaFDh6bmzZtnfii5//770+jRo5e43vLmmoYRudOzUiluW73vvvvGzTffnLuO2OzZs+PJJ5+MU089dZnL7rXXXtGpU6cYO3ZsjB07Nl566aV4+OGH49FHH42rrroq9thjj6hevXrceuutkVKKSZMmxZdffhmffPJJfPXVV7Hppptm1te0adPc/9etWzciFl7/oMg666wT8+fPj1mzZuWuidC5c+fMKY677rprjBo1Kr788suoXr16Zv2vvfZa7LbbbhERudMYtt9++6hatWq8/PLLUVBQkDsdrkjRDRny8vKK7YPlvZ6XWpVfrcqSOpV/nb777rs4+uijIyLi6quvXuK+tixqVb61mjdvXgwdOjR+/fXXGDRoUBx88MExevTo2GCDDZbZV7+nVuVTq+eeey6efPLJePTRR5fZJ7Ay3X777cXeGGDRm200atQojjnmmLjqqquiSpUq8dBDDy3xsgMRy95PGjRoEBER9erVW2ysVq1atdzjEZG7/EzRpQiKbqw0ZcqU+Pzzz+OLL76IF198sczuGr+k59O8efNYc801c8+nfv36sfHGG8fLL78cHTp0iBYtWsR9990X33//fbRr1y522WWXzCnBpdWyZcs466yzFpv++/465ZRT4sknn4xevXpFixYt4ogjjljiOt9555347bffYqeddsqcMrzTTjvF448/Hj///HNuWv369Rdbfuutt878/ac//Sk+/fTTiFh4A4KRI0fG/Pnz48svv4wvv/wyPvroo5gxY0a51ebLL7+MH374IXbeeefFns/NN98cn376aWy11VaxzjrrxAknnBB77LFH7LzzztG6detS3am0pOxLy7Yq9qVFqdGyreoaLcsxxxwTe+yxx2LTF72ZUaVKleLSSy+NAw44IK688so44YQTonHjxktc52uvvRY1a9aM5s2b555zpUqVonXr1otd7mxZ741Vq1aNOnXqxK+//hoRC68HvvPOO8fs2bPj008/jS+//DLefPPNiIhyq2PRJYOKe68vOnV6m222iVmzZsXee+8df/3rX6Ndu3ax//77L/d3vbIgNIz/3c3n66+/XuI83333XeZL3l/+8pdo1KhRPP7447HjjjvGM888E3PmzCnxHWerV68eHTp0iA4dOkRExAcffBCnnXZa/OMf/4i//vWvUalSpXjqqafisssui8mTJ8e6664bW2+9dayxxhqLrau4u0oteh2t4qy//vqZv4sudD9t2rTMNQgjIn7++ee444474o477lhsPd9//328+uqruWuzFBk3blzUqlUrZs6cudgyM2fOXO47oalV+dSqrKlT+dbps88+i2OPPTbmzp0bw4cPXyycKQ21Kt9aVa1aNXf9wW222SZ22WWXGDVqVPTq1WupbSyOWpV9rd5444246KKL4uSTT44NNtgg5s2bl7uO44IFC2LBggXL/SMXrKiGDRuW6AL4e+65Z9x0002x2WabRb169ZY677L2kyLF3ayuevXqmS8tRftG0Q8Zn332WZx//vnx1ltvxZprrhkNGzYsk7t9Ls3PP/8c77zzTrFfOn/44YeIWPjlfoMNNoiHH344nn322ahUqVLsvPPOcdllly3XTflq166d+RFkSWrUqBEdOnSIESNGxI477rjUL3xFoeBBBx1U7OM//vhj7n216EeXRf3+R5RKlSplrkl77733xvXXXx9Tp06NDTbYIAoLC6N69eql+hGqNIqezymnnFLs4z/88EM0atQo7rjjjrjxxhvjwQcfjDvvvDPWWmutOO6446JHjx5l2h770rKtin1pUWq0bKu6Rsuy4YYblui9saCgIPLz8+ODDz6InXfeeanz/vzzzzFz5sxin/Pvn09x742/H48u+t44b968uPzyy+O+++6LefPmxeabb567YUl5vze2bt262Md//fXXaNGiRQwaNCiGDx8ew4YNi0GDBsUmm2wSf//735f7ppcrSmgYC19wDRo0iLFjx8Zhhx222OMzZ86MTp06xaGHHpq5tfU+++wTgwYNiosvvjgeffTR2GmnnZa5Mx588MFRWFgYffr0yUxv3Lhx9OzZM84555z46aefYvr06XHqqafGgQceGD179sx9CezevXuJL+q+NIv+Yhnxv4veF7ez1apVKzp37lzsQGa99daL2rVrF3vzjHr16sWkSZMy03766aeYMWNGsb8ElIRalU+typo6lV+d3n333Tj22GOjVq1aceeddy5zwLQsalU+tXr55Zdj/vz5mRuWrLvuurHBBhtkBqqloVZlX6sPPvggvv766+jXr1/069cv81jHjh1jv/32W2w6rE5SSnHppZfGFltsERMmTIjbbrstunXrtsT5l7WfLK8FCxZEz549Y7311ovHHnssttxyy8jLy4uhQ4fGK6+8stzrXZZatWpFu3bt4sQTT1zssbXXXjsiFn5pPOmkk+Kkk06KL774Iv7zn//EjTfeGDfccENcfPHF5da2zz//PO66664oKCiIwYMHxx577BEbb7xxsfMW/Zg+ePDgxX4siYjYZJNN4scff1yudrz66qu5H0cOO+ywqFOnTkQsPJK7vBQ9n759+0ajRo0We3zzzTePiIU/bF1//fUxZ86ceOONN2LkyJFx9dVXR8uWLRc7erK82ZdW332piBqt/jUqiQceeCA++uij+Mtf/hIXX3xxjBo1aolnvtWuXTvWX3/9GDx48GKPrejZckOGDInRo0fHtddeGzvssEPUqFEjvv/++3jwwQdXaL1LU7t27ahcuXLcc889xf6QVBQ8t2/fPtq3bx/Tp0+PsWPHxuDBg+O0006LF198sdzD6eL4+fz/O+KII+KZZ57J3JGpyJAhQ2L27NmLHW671157xcyZM+O5556L559/Pvbbb79lbmfjjTeOf/7zn8V+8I8fPz7WW2+9qFOnTnz44Ycxd+7cOO6443Jfwn755Zd47733yiT5HjNmTGY9Tz31VGy44Ya5D/FFbbPNNvHll19G06ZNc//WXXfduOaaa2L8+PFRq1atzGNFvzBsv/328e6778Z3332X2U7VqlVX6NQDtSr7WpUHdSr7On311VfRo0ePqFu3btx9990rHBgWUauyr9VDDz0U5513XiY4mzBhQkyePDlzV7/SUquyrVXjxo3j/vvvz/w7++yzI2Lhl/flOSIUVqZRo0bF66+/HldccUV069Yt+vfvn/nB9vdHyi5rP1leU6dOjYkTJ8YhhxwSW221Ve7L0AsvvBAR5XfURtHzadCgQe751K9fP66//vp4//33Y8GCBbHXXnvFiBEjIiJiiy22iBNOOCEKCwvjm2++iYjlv2TO0qSU4oILLogtttgi7r777qhbt25cdNFFucd/f0pls2bNokqVKvHzzz9navPee+/FsGHDVqiNb7/9dlSuXDmOP/74XGA4adKkmDRpUrnVZYsttoh11lknvv/++8zz+e6772LAgAExZ86ceOGFF6J169YxderUqFatWrRu3TrOO++8iPjfEfUr80hv+9LquS8tSo1W/xoty48//hhXXnllHHbYYXHdddfFJ598Erfeemvu8eJqOHXq1Fh77bUzdRw1alQ88cQTK9SWt99+O5o2bRodO3bMnfny4osvRkT51nD+/Pnx66+/Zp7Ps88+G/fcc09Urlw5dzmjiIUh4+677x7dunWL6dOnx4wZMyJi5dfRkYb/3/777x9jxoyJ448/Po444oho3bp1zJ49O5544ol45JFH4sQTT4xmzZpllqlbt260bds2rrzyyqhRo8YyD6+NWHiY/osvvhgHHHBAHHPMMdGwYcOYM2dO/Pe//4277rorLrvssqhUqVI0aNAgKleuHFdddVUccsghMWXKlBg2bFj88ssvufPwV8Qnn3wSZ5xxRuy7777x+uuvx4MPPhh9+/Ytdt6ePXvG4YcfHr1794599tknZs2aFQMHDoxp06ZlrvXwe3vuuWcMHjw4jj322DjllFPiu+++i6uuuioOO+ywFTo8Wq3KvlblQZ3Kvk79+vWL6dOnxwUXXBDffPNNbgAQsTDkKe7ohJJQq7Kv1THHHBP/+te/4qSTToquXbvG1KlTY8CAAbH55pvH/vvvv9xtV6uyrVVRkLiooh+68vPzY5NNNlnh5wDL6/333y/2Gl8RC6/T9MMPP8SVV14ZBx10UDRr1iz+8pe/xCOPPBIXXnhh7tpIa621Vrz33nvx6quvxnbbbVdu44S6devGRhttFMOHD4911lkn8vLyYtSoUbnrN/36669LfC4r4phjjomHHnooTjjhhOjSpUvk5eXFsGHDYty4cXH++edHpUqVorCwMG688caoXr16bLnllvH222/HG2+8kTvqpuiUyCeeeCLatGmzzP3+l19+ibfffrvYx+rUqRObb7553HXXXfHmm2/GXXfdFTVr1ozzzjsvTjjhhHj44Ydjn332yR2J9/TTT0e1atViyy23jMMPPzwuvfTS+Omnn6Jx48bx3nvvxQ033BD777//Ch1Z0rRp09xpeB07doyvvvoqbrzxxliwYEGZvE8Xp0qVKtGzZ8+4+uqrY8GCBdGyZcsYP358XHPNNbH11lvHeuutl3tOvXr1ih49ekTVqlXj9ttvj7XWWitatWoVEQtr8/HHH8crr7yyxEtflIR9adlWxb60KDVatlVdo2X5+uuvl/jeuOmmm0bdunXjkksuiSpVqsSpp54aa621Vhx55JExcODA6NSpU9SrVy/WWmutmDlzZjz99NPRtGnTaNeuXTRs2DB69OiRO6Pl0Ucfjfvvvz+uueaaFWpvkyZN4uabb44777wzttpqq3j77bdj8ODBkZeXF7Nnz16hdS9Jo0aNYtddd43evXvHiSeeGPXr14/nn38+br755jj99NMjYuF1cwcMGBDnn39+7LHHHjFt2rQYMmRIbLvttrkMZa211orXX389mjdvvlKOyhYa/n+VKlWKG264Ie6+++4YPXp03HvvvZGXlxdbbbVV9O/ff4mH8O+7775x2mmnRdeuXUv0gV6vXr0YPXp0DBkyJEaOHBnff/99VKtWLRo3bhxDhgyJnXbaKSIW/jJw2WWXxaBBg+K4446L9ddfP3bdddfYe++9o3///rmUeXkdeOCBMWvWrOjVq1dssMEGcckll8SBBx5Y7LzNmjWLESNGxHXXXRe9evWKGjVqxHbbbRdnnHFG7hfL4tSoUSNuvfXWuOSSS6J3795Ru3bt6NKlS5x22mkr1Ha1KvtalQd1Kts6zZs3L8aMGRPz58+P3r17L/b4ueeem7sxSmmpVdnvUwUFBXH77bfHNddcE6ecckpUrVo12rdvH2eeeeYyr+O3NGpVMd7/oCwcc8wxS3zs3Xffjb59+0aVKlVyXzRq1KgRffr0iV69esUDDzwQBxxwQBx11FFx+eWXR48ePeLf//53ue0neXl50b9///jHP/4Rp5xyStSuXTuaN28eQ4cOjeOOOy7efvvtaNOmzXKvf0k23njjuPPOO+Pqq6+O3r17R+XKlaNZs2Zx22235S6Fc+6550b16tVjyJAhMWXKlNh4443j7LPPzp2y2KpVq2jZsmX07ds3Dj744LjwwguXus3XXnstDjnkkGIf23333eOcc86Ja6+9Nvbff//YZpttIiKiXbt20b59+7jssstixx13jL/85S+x2267xdChQ+P999+PIUOGxLnnnhvrrbde3HXXXbnr0x5//PHRs2fPFeqj1q1bx1lnnRUjR46MUaNGxYYbbhgHHHBATJ8+PZ544olyO6Lm6KOPjjXXXDNGjBgRN910U9StWzf3WRSx8EvvzTffHFdffXWcddZZMXfu3GjWrFkMHz4898X48MMPj7PPPjuOPfbYuO2223L9WVr2pWVbFfvSotRo2VZ1jZbltttui9tuu63Yx/7+97/HBhtsEE888URcccUVufCyV69e8dhjj8UFF1wQt99+e3Tq1ClGjx4dp5xySpxyyinRo0ePuOWWW+Lqq6+Ofv36xa+//hr169ePa6+9ttibrpTGcccdF9999130798/5s6dG/Xr14/LL788br311njrrbeiS5cuK7T+Jbnuuuvi+uuvj0GDBsXPP/8cm2yySZx33nm5m2W1aNEirrnmmhg2bFg8+uijUb169dhpp50yN7Q59thjY8iQIfHqq6/Giy++WC43Nl1UXiqvTwpWW+3bt49dd901dwoAqy+1qhjUqeJQq4pDrQAAYNVypCEAALDSFN2dfGkqVaq0yq+/9X/RvHnzljlP5cqVl3o3aFYe+9LqT40qvpRSzJ8/f5nzlfcRf6vKH/NZAQAAq6U+ffos8w6V7l6+ajRu3HiZ89x+++256w6yatmXVn9qVPG9+uqrceSRRy5zvnHjxq2E1qx8Tk8GAABWmq+++ip++umnpc5Tp04dNyNaBd57771lzlO/fv2oVavWSmgNy2JfWv2pUcU3Y8aM+PLLL5c53+9vrvdHITQEAAAAADKcOA8AAAAAZAgNAQAAAIAMoSEAAAAAkCE0BAAAAAAyhIYAAAAAQIbQEAAAAADIEBoCAAAAABlCQwAAAAAg4/8BeWJCNazlZJEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_PSO_HotelCancel_xgboost,\n",
    "                                                  cost_funct=cost_funct)\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Particle Swarm Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.52\n",
      "Average Test and Validation Cost: 0.86 Average External Test Cost: 0.91\n",
      "Hotel Cancellations dataset - Ant Colony Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1)#,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_ACO_HotelCancel_xgboost,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Ant Colony Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.4\n",
      "Average Test and Validation Cost: 0.88 Average External Test Cost: 0.93\n",
      "Hotel Cancellations dataset - Simulated Annealing\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=200,random_state=1,n_jobs=-1)\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_SA_HotelCancel_xgboost,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Simulated Annealing'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.41\n",
      "Average Test and Validation Cost: 0.86 Average External Test Cost: 0.92\n",
      "Hotel Cancellations dataset - Genetic Algorithm\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_GA_HotelCancel_xgboost,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Genetic Algorithm'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Intel(R) Extension for Scikit-learn* enabled (https://github.com/intel/scikit-learn-intelex)\n"
     ]
    }
   ],
   "source": [
    "patch_sklearn()\n",
    "model_use = LogisticRegression(random_state=1,n_jobs=-1)\n",
    "\n",
    "fsObj = FeatureSelection(columns_list = linear_features_all,\n",
    "                         data_dict = data_dict_for_linear,\n",
    "                         x_validation_dataframe = x_validation_linear,\n",
    "                         y_validation_dataframe = y_validation_linear,\n",
    "                         model = model_use,\n",
    "                         cost_function = cost_funct,\n",
    "                         cost_function_improvement = 'increase',\n",
    "                         average = None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.6930952380952381\n",
      "Best combined performance on test and validation data for iteration 1: 0.6930952380952381\n",
      "Best combined performance on test and validation data for iteration 2: 0.725\n",
      "Best combined performance on test and validation data for iteration 3: 0.756271186440678\n",
      "Best combined performance on test and validation data for iteration 4: 0.756271186440678\n",
      "Best combined performance on test and validation data for iteration 5: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 6: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 7: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 8: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 9: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 10: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 11: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 12: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 13: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 14: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 15: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 16: 0.7615976233701931\n",
      "Best combined performance on test and validation data for iteration 17: 0.764\n",
      "Best combined performance on test and validation data for iteration 18: 0.764\n",
      "Best combined performance on test and validation data for iteration 19: 0.764\n",
      "Best combined performance on test and validation data for iteration 20: 0.764\n",
      "Best combined performance on test and validation data for iteration 21: 0.7872514619883042\n",
      "Best combined performance on test and validation data for iteration 22: 0.7872514619883042\n",
      "Best combined performance on test and validation data for iteration 23: 0.7872514619883042\n",
      "Best combined performance on test and validation data for iteration 24: 0.7872514619883042\n",
      "================= Best result: 0.7872514619883042 =================\n",
      "================= Execution time in minutes: 664.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Particle Swarm Optimization\n",
    "best_columns_PSO_HotelCancel_logistic = fsObj.ParticleSwarmOptimization(iterations = 25,                                                                   \n",
    "                                                                       swarmSize = 75,\n",
    "                                                                       run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_PSO_HotelCancel_logistic.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_PSO_HotelCancel_logistic, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 1: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 2: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 3: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 4: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 5: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 6: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 7: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 8: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 9: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 10: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 11: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 12: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 13: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 14: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 15: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 16: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 17: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 18: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 19: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 20: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 21: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 22: 0.6061453853738875\n",
      "Best combined performance on test and validation data for iteration 23: 0.6061453853738875\n",
      "================= Run time exceeded allocated time. Producing best solution generated so far. =================\n",
      "================= Best result: 0.6061453853738875 =================\n",
      "================= Execution time in minutes: 1254.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Ant Colony Optimization\n",
    "best_columns_ACO_HotelCancel_logistic = fsObj.AntColonyOptimization(iterations = 25,\n",
    "                                                                N_ants = 75,\n",
    "                                                                run_time=1200,\n",
    "                                                                evaporation_rate=0.8,\n",
    "                                                                Q=0.2)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_ACO_HotelCancel_logistic.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_ACO_HotelCancel_logistic, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for iteration 0: 0.5644444444444444\n",
      "Best combined performance on test and validation data for iteration 1: 0.5696793381384174\n",
      "Best combined performance on test and validation data for iteration 2: 0.5853146430352831\n",
      "Best combined performance on test and validation data for iteration 3: 0.5863837433273799\n",
      "Best combined performance on test and validation data for iteration 4: 0.6863388423456264\n",
      "Best combined performance on test and validation data for iteration 5: 0.6819483694523936\n",
      "Best combined performance on test and validation data for iteration 6: 0.7052742410530978\n",
      "Best combined performance on test and validation data for iteration 7: 0.7061374060800186\n",
      "Best combined performance on test and validation data for iteration 8: 0.7056052925624682\n",
      "Best combined performance on test and validation data for iteration 9: 0.7052742410530978\n",
      "Best combined performance on test and validation data for iteration 10: 0.7076994956954682\n",
      "Best combined performance on test and validation data for iteration 11: 0.7071580468530485\n",
      "Best combined performance on test and validation data for iteration 12: 0.7058366182816198\n",
      "Best combined performance on test and validation data for iteration 13: 0.7067943566444475\n",
      "Best combined performance on test and validation data for iteration 14: 0.7067943566444475\n",
      "Best combined performance on test and validation data for iteration 15: 0.7067943566444475\n",
      "Best combined performance on test and validation data for iteration 16: 0.7060798772371963\n",
      "Best combined performance on test and validation data for iteration 17: 0.7191447007006747\n",
      "Best combined performance on test and validation data for iteration 18: 0.6990285951859143\n",
      "Best combined performance on test and validation data for iteration 19: 0.7035554747165902\n",
      "Best combined performance on test and validation data for iteration 20: 0.7046171745007991\n",
      "Best combined performance on test and validation data for iteration 21: 0.7042958237382718\n",
      "Best combined performance on test and validation data for iteration 22: 0.7042958237382718\n",
      "Best combined performance on test and validation data for iteration 23: 0.7037243173579419\n",
      "Best combined performance on test and validation data for iteration 24: 0.7042958237382718\n",
      "Best combined performance on test and validation data for iteration 25: 0.7060670962910837\n",
      "Best combined performance on test and validation data for iteration 26: 0.7054036781009521\n",
      "Best combined performance on test and validation data for iteration 27: 0.7059462252168858\n",
      "Best combined performance on test and validation data for iteration 28: 0.7042822270613786\n",
      "Best combined performance on test and validation data for iteration 29: 0.70552454917515\n",
      "Best combined performance on test and validation data for iteration 30: 0.7048247741773123\n",
      "Best combined performance on test and validation data for iteration 31: 0.6972188235315063\n",
      "Best combined performance on test and validation data for iteration 32: 0.6903458951345103\n",
      "Best combined performance on test and validation data for iteration 33: 0.6927905506688916\n",
      "Best combined performance on test and validation data for iteration 34: 0.7020128993813204\n",
      "================= Best result: 0.7191447007006747 =================\n",
      "================= Execution time in minutes: 413.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Simulated Annealing\n",
    "best_columns_SA_HotelCancel_logistic = fsObj.SimulatedAnnealing(temperature = 1500,\n",
    "                                                               iterations = 35,\n",
    "                                                               n_perturb = 75,\n",
    "                                                               run_time=1200,\n",
    "                                                               n_features_percent_perturb=1,\n",
    "                                                               alpha=0.9)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_SA_HotelCancel_logistic.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_SA_HotelCancel_logistic, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best combined performance on test and validation data for generation 0 : 0.5902881218067784\n",
      "Best combined performance on test and validation data for generation 1 : 0.5892286062789036\n",
      "Best combined performance on test and validation data for generation 2 : 0.5920887836221107\n",
      "Best combined performance on test and validation data for generation 3 : 0.5920859882590694\n",
      "Best combined performance on test and validation data for generation 4 : 0.5909966228730517\n",
      "Best combined performance on test and validation data for generation 5 : 0.5910900161625708\n",
      "Best combined performance on test and validation data for generation 6 : 0.5910196922949658\n",
      "Best combined performance on test and validation data for generation 7 : 0.6043108387297329\n",
      "Best combined performance on test and validation data for generation 8 : 0.599796681212791\n",
      "Best combined performance on test and validation data for generation 9 : 0.6006203096781776\n",
      "Best combined performance on test and validation data for generation 10 : 0.5945118750591034\n",
      "Best combined performance on test and validation data for generation 11 : 0.6004267312288498\n",
      "Best combined performance on test and validation data for generation 12 : 0.5932572367513596\n",
      "Best combined performance on test and validation data for generation 13 : 0.5924081799380838\n",
      "Best combined performance on test and validation data for generation 14 : 0.5950755099511233\n",
      "Best combined performance on test and validation data for generation 15 : 0.5928221090089069\n",
      "Best combined performance on test and validation data for generation 16 : 0.5921427081380298\n",
      "Best combined performance on test and validation data for generation 17 : 0.5923124011263176\n",
      "Best combined performance on test and validation data for generation 18 : 0.5920071901819142\n",
      "Best combined performance on test and validation data for generation 19 : 0.5947125316411573\n",
      "================= Best result: 0.6043108387297329 =================\n",
      "================= Execution time in minutes: 898.0 =================\n"
     ]
    }
   ],
   "source": [
    "#### Genetic Algorithm\n",
    "best_columns_GA_HotelCancel_logistic = fsObj.GeneticAlgorithm(generations=20,\n",
    "                                                             population=75,\n",
    "                                                             run_time=1200)\n",
    "\n",
    "with open(r\"D:\\\\Results\\\\ResultPickle\\\\best_columns_GA_HotelCancel_logistic.pickle\", \"wb\") as output_file:\n",
    "    pickle.dump(best_columns_GA_HotelCancel_logistic, output_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.02\n",
      "Average Test and Validation Cost: 0.79 Average External Test Cost: 1.0\n",
      "Hotel Cancellations dataset - Particle Swarm Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model_use = XGBClassifier(n_estimators=500,random_state=1,n_jobs=-1,tree_method = 'gpu_hist')\n",
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_PSO_HotelCancel_logistic,\n",
    "                                                  cost_funct=cost_funct)\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Particle Swarm Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.17\n",
      "Average Test and Validation Cost: 0.6 Average External Test Cost: 0.86\n",
      "Hotel Cancellations dataset - Ant Colony Optimization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_ACO_HotelCancel_logistic,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Ant Colony Optimization'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.01\n",
      "Average Test and Validation Cost: 0.71 Average External Test Cost: 0.71\n",
      "Hotel Cancellations dataset - Simulated Annealing\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_SA_HotelCancel_logistic,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Simulated Annealing'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "External Test recall: 0.15\n",
      "Average Test and Validation Cost: 0.6 Average External Test Cost: 0.95\n",
      "Hotel Cancellations dataset - Genetic Algorithm\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_tree_cost,external_tree_cost = get_metrics(model=model_use,\n",
    "                                                  data_dict=data_dict,\n",
    "                                                  problem_type = 'classification',\n",
    "                                                  dependent_variable=dependent_variable,\n",
    "                                                  features=best_columns_GA_HotelCancel_logistic,\n",
    "                                                  cost_funct=cost_funct)\n",
    "\n",
    "print('Average Test and Validation Cost:',round(np.mean(list(test_tree_cost.values())),2),'Average External Test Cost:',round(np.mean(external_tree_cost),2))\n",
    "metric = 'Precision'\n",
    "title = 'Hotel Cancellations dataset - Genetic Algorithm'\n",
    "print(title)\n",
    "make_plot(metric=metric,\n",
    "          title=title,\n",
    "          test_tree_cost=test_tree_cost,\n",
    "          external_tree_cost=external_tree_cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "unpatch_sklearn()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}