4.-Correct model selection, training and tunning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d12c5c12",
   "metadata": {},
   "source": [
    "# Model Selection\n",
    "Diferent algorithms will be run to find the one with the best metrics and results for this problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "c6f9733d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Setup\n",
    "from __future__ import division, print_function, unicode_literals\n",
    "\n",
    "# Common imports\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "# to make this notebook's output stable across runs\n",
    "np.random.seed(42)\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97408076",
   "metadata": {},
   "source": [
    "# csv loading \n",
    "\n",
    "load train and test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "fd01c428",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pathlib\n",
    "pathlib.Path.cwd()\n",
    "\n",
    "#Loading train set\n",
    "path = pathlib.Path.cwd() / 'process sets' / 'xtrain_sel.csv'\n",
    "x_train = pd.read_csv(path, index_col= None)\n",
    "path = pathlib.Path.cwd() / 'process sets' / 'ytrain_sel.csv'\n",
    "y_train = pd.read_csv(path, index_col= None)\n",
    "\n",
    "\n",
    "#Loading test set\n",
    "path = pathlib.Path.cwd() / 'process sets' / 'xtest_sel.csv'\n",
    "x_test = pd.read_csv(path, index_col= None)\n",
    "path = pathlib.Path.cwd() / 'process sets' / 'ytest_sel.csv'\n",
    "y_test= pd.read_csv(path, index_col= None)\n",
    "\n",
    "\n",
    "\n",
    "#Preventive reset index and squeeze to avoid bugs\n",
    "y_train.reset_index(inplace=True)\n",
    "y_train.drop(\"index\", axis= 1, inplace=True)\n",
    "y_train = y_train.squeeze()\n",
    "\n",
    "\n",
    "\n",
    "x = x_train.copy()\n",
    "y = y_train.copy()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79a37bec",
   "metadata": {},
   "source": [
    "# Defining cross validaton score function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "12a159d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Crossvalidate and display score function based in mean squared error\n",
    "\n",
    "def cross_score(model,x,y):\n",
    "    from sklearn.model_selection import cross_val_score\n",
    "\n",
    "    scores = cross_val_score(model, x, y,\n",
    "                         scoring=\"neg_mean_squared_error\", cv=10)\n",
    "    rmse_scores = np.sqrt(-scores)\n",
    "    \n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "  \n",
    "    \n",
    "    \n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d30619cb",
   "metadata": {},
   "source": [
    "# Def model functions \n",
    "\n",
    "Since some models are stochastical and doesnt show the same result in every run, we run them n times and average the resulting evaluation metrics to choose the best\n",
    "\n",
    "Models:\n",
    "- Linear Regressor\n",
    "- Decision Tree Regressor\n",
    "- Random Forest Regressor\n",
    "- Support vector machine regressor\n",
    "- KG-Boost Regressor\n",
    "- K-n Neighbor regressor\n",
    "- Lasso Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "ca91811b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Number of times to run each model\n",
    "nloop= 10\n",
    "\n",
    "\n",
    "#Defining Linear Regression function\n",
    "\n",
    "def lin_loop(x,y):\n",
    "    \n",
    "    #Rmse acumulation variable\n",
    "    rmseacum = 0\n",
    "    \n",
    "    #Model import\n",
    "    from sklearn.linear_model import LinearRegression\n",
    "    \n",
    "    #Loop of model run\n",
    "    for i in range (nloop):\n",
    "        model_reg = LinearRegression()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "# Defining Decision Tree Regression function\n",
    "\n",
    "def dtr_loop(x,y):\n",
    "    #Rmse acumulation variable\n",
    "    rmseacum = 0\n",
    "    \n",
    "    #Loop of model run\n",
    "    from sklearn.tree import DecisionTreeRegressor\n",
    "    \n",
    "    for i in range (nloop):\n",
    "        model_reg = DecisionTreeRegressor()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "#Defining Random Forest regression function\n",
    "\n",
    "def rfr_loop(x,y):\n",
    "    \n",
    "    rmseacum = 0\n",
    "    from sklearn.ensemble import RandomForestRegressor\n",
    "   \n",
    "    for i in range (nloop):\n",
    "        model_reg = RandomForestRegressor()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "\n",
    "\n",
    "#Defining Support Vector Regression function\n",
    "\n",
    "def svr_loop(x,y):\n",
    "    #Rmse acumulation variable\n",
    "    rmseacum = 0\n",
    "    from sklearn.svm import SVR\n",
    "   \n",
    "\n",
    "    #Loop of model run\n",
    "    for i in range (nloop):\n",
    "        model_reg =  SVR(kernel=\"linear\")\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    \n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "#Defining XGBOOST Regression function\n",
    "def xg_loop(x,y):\n",
    "    \n",
    "    rmseacum = 0\n",
    "    import xgboost\n",
    "    print(xgboost.__version__)\n",
    "    from xgboost import XGBRegressor\n",
    "   \n",
    "    for i in range (nloop):\n",
    "        model_reg =  XGBRegressor()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "#Defining K-n Neighbors Regressor function\n",
    "\n",
    "\n",
    "def kn_loop(x,y):\n",
    "    \n",
    "    rmseacum = 0\n",
    "    from sklearn.neighbors import KNeighborsRegressor\n",
    "   \n",
    "    for i in range (nloop):\n",
    "        model_reg =  KNeighborsRegressor()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#------------------------------------------------------------------------------------------------------------\n",
    "#Defining Lasso Regressor function\n",
    "def lasso_loop(x,y):\n",
    "    \n",
    "    rmseacum = 0\n",
    "    from sklearn.linear_model import Lasso\n",
    "   \n",
    "    for i in range (nloop):\n",
    "        model_reg =  Lasso()\n",
    "        model_reg.fit(x, y)\n",
    "        model=  model_reg\n",
    "        \n",
    "        \n",
    "        \n",
    "        \n",
    "        #Scores acumulator using cross validation function\n",
    "        scores = cross_val_score(model, x, y, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "        rmse_scores = np.sqrt(-scores)\n",
    "        rmseacum= rmseacum + rmse_scores\n",
    "        print(\"it\", i)\n",
    "    \n",
    "    #Scores average metrics\n",
    "    rmse_scores = rmseacum/nloop\n",
    "    print(\"Model:\", str(model))\n",
    "    print(\"Scores:\",  rmse_scores)\n",
    "    print(\"MAE:\", round( rmse_scores.mean(),3))\n",
    "    print(\"RMEE:\",round(  rmse_scores.std(), 3))\n",
    "    print(\"MAE / data mean:\", round( (rmse_scores.mean()/ y.mean()),3))\n",
    "    print(\"RMEE / data std:\",round(  (rmse_scores.std()/ y.std()), 3))\n",
    "    \n",
    "    \n",
    "    return rmse_scores\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98facff8",
   "metadata": {},
   "source": [
    "# Models Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "dfdd419d",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: LinearRegression()\n",
      "Scores: [58937.3798169  28709.72573867 31374.80687159 35240.08627964\n",
      " 40750.63842641 23427.99333511 27727.78355793 30789.63938308\n",
      " 46884.17401914 26799.92249634]\n",
      "MAE: 35064.215\n",
      "RMEE: 10352.741\n",
      "MAE / data mean: 0.194\n",
      "RMEE / data std: 0.13\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: DecisionTreeRegressor()\n",
      "Scores: [53016.80365556 41765.75365772 43728.985807   40190.6320584\n",
      " 38365.75029055 32130.80698462 36492.45215896 37118.78287098\n",
      " 47083.01263986 34323.73421758]\n",
      "MAE: 40421.671\n",
      "RMEE: 5935.546\n",
      "MAE / data mean: 0.223\n",
      "RMEE / data std: 0.075\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: RandomForestRegressor()\n",
      "Scores: [38847.30013617 26278.35474188 25109.70298525 31496.02620255\n",
      " 29674.81932385 23314.26846102 26808.67880035 27666.81193781\n",
      " 44949.02329306 24046.49537378]\n",
      "MAE: 29819.148\n",
      "RMEE: 6612.881\n",
      "MAE / data mean: 0.165\n",
      "RMEE / data std: 0.083\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: SVR(kernel='linear')\n",
      "Scores: [97641.76655465 81528.91956797 91795.83002306 83628.39365988\n",
      " 91208.03359781 63782.65159702 69885.01618053 69404.99238288\n",
      " 84314.90652896 72013.86097555]\n",
      "MAE: 80520.437\n",
      "RMEE: 10717.995\n",
      "MAE / data mean: 0.444\n",
      "RMEE / data std: 0.135\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: KNeighborsRegressor()\n",
      "Scores: [53146.93402847 37907.6320009  40497.64705161 37117.25065722\n",
      " 47228.37842193 24540.21584184 31336.05679161 33985.93587704\n",
      " 46410.67188599 28413.18639042]\n",
      "MAE: 38058.391\n",
      "RMEE: 8520.079\n",
      "MAE / data mean: 0.21\n",
      "RMEE / data std: 0.107\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: Lasso()\n",
      "Scores: [58923.26438256 28710.16995858 31369.71142671 35254.61762575\n",
      " 40784.35319978 23420.76990877 27736.23440963 30778.33958464\n",
      " 46851.89763912 26808.36295764]\n",
      "MAE: 35063.772\n",
      "RMEE: 10347.851\n",
      "MAE / data mean: 0.194\n",
      "RMEE / data std: 0.13\n",
      "1.5.0\n",
      "it 0\n",
      "it 1\n",
      "it 2\n",
      "it 3\n",
      "it 4\n",
      "it 5\n",
      "it 6\n",
      "it 7\n",
      "it 8\n",
      "it 9\n",
      "Model: XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
      "             colsample_bynode=1, colsample_bytree=1, enable_categorical=False,\n",
      "             gamma=0, gpu_id=-1, importance_type=None,\n",
      "             interaction_constraints='', learning_rate=0.300000012,\n",
      "             max_delta_step=0, max_depth=6, min_child_weight=1, missing=nan,\n",
      "             monotone_constraints='()', n_estimators=100, n_jobs=4,\n",
      "             num_parallel_tree=1, predictor='auto', random_state=0, reg_alpha=0,\n",
      "             reg_lambda=1, scale_pos_weight=1, subsample=1, tree_method='exact',\n",
      "             validate_parameters=1, verbosity=None)\n",
      "Scores: [30387.295421   27298.11990199 28363.90360942 31941.36867512\n",
      " 30562.89447288 25498.63025746 29507.72777175 27533.69875104\n",
      " 53567.39667649 26983.9383318 ]\n",
      "MAE: 31164.497\n",
      "RMEE: 7692.583\n",
      "MAE / data mean: 0.172\n",
      "RMEE / data std: 0.097\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "#LR RMSE\n",
    "lin_mod = lin_loop(x,y)\n",
    "#DTR RMSE\n",
    "dtr_mod = dtr_loop(x,y)\n",
    "#Random Forest RMSE\n",
    "rfr_mod= rfr_loop(x,y)\n",
    "#SVR RMSE\n",
    "svr_mod= svr_loop(x,y)\n",
    "#K-n RMSE\n",
    "kn_mod= kn_loop(x,y)\n",
    "#Lasso RMSE\n",
    "lasso_mod = lasso_loop(x,y)\n",
    "#Xg boost RMSE\n",
    "xg_mod = xg_loop(x,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "b53aa5c6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best MAE model:  random forest reg MAE:  0.1645763042170381\n",
      "Best RMSE model:  dt reg RMSE:  0.07470784601465019\n"
     ]
    }
   ],
   "source": [
    "#Selecting best\n",
    "\n",
    "#model  results\n",
    "modelos =[lin_mod,dtr_mod, rfr_mod, svr_mod, kn_mod, lasso_mod, xg_mod]\n",
    "#model names\n",
    "modelos_nombres= [\"linear reg\",\"dt reg\", \"random forest reg\", \"SV Reg\", \"Kn Reg\", \"Lasso Reg\", \"XgBoost\"]\n",
    "\n",
    "#Initial proposed high Mae and RMSE to be replaced by a better one with models in the next cycle\n",
    "BestMAE = 10000000\n",
    "BestRMEE = 10000000\n",
    "\n",
    "\n",
    "#Finding best model with this cicle (normalized by expected mean and standar deviation)\n",
    "for n,i in enumerate (modelos):\n",
    "    #Normalizing results\n",
    "    MAE = (i.mean()/ y.mean())\n",
    "    RMEE =(i.std()/ y.std())\n",
    "    \n",
    "    #Saving top model bases in MAE and RMSE\n",
    "    if MAE < BestMAE:\n",
    "        BestMAE= MAE\n",
    "        ModelMAE= modelos_nombres[n]\n",
    "    if RMEE < BestRMEE:\n",
    "        BestRMEE= RMEE\n",
    "        ModelRMEE= modelos_nombres[n]\n",
    "        \n",
    "    \n",
    "    \n",
    "#printing\n",
    "print(\"Best MAE model: \", ModelMAE , \"MAE: \", BestMAE)\n",
    "print(\"Best RMSE model: \", ModelRMEE , \"RMSE: \", BestRMEE)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf8f139d",
   "metadata": {},
   "source": [
    "# Choosed Model \n",
    "\n",
    "\n",
    "The choosed model to work with is Random Forest Regression, because it has low MAE and RMSE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "27f24bd7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "f82d51e3",
   "metadata": {},
   "source": [
    "# Hyperparamater Tuning of Random Forest Regression\n",
    "\n",
    "- We can use grid search \n",
    "- We can use randomized search\n",
    "\n",
    "We will use randomized search approximation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3349012",
   "metadata": {},
   "source": [
    "# Grid Search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "930267e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-7 {color: black;background-color: white;}#sk-container-id-7 pre{padding: 0;}#sk-container-id-7 div.sk-toggleable {background-color: white;}#sk-container-id-7 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-7 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-7 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-7 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-7 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-7 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-7 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-7 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-7 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-7 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-7 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-7 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-7 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-7 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-7 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-7 div.sk-item {position: relative;z-index: 1;}#sk-container-id-7 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-7 div.sk-item::before, #sk-container-id-7 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-7 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-7 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-7 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-7 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-7 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-7 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-7 div.sk-label-container {text-align: center;}#sk-container-id-7 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-7 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-7\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "             param_grid=[{&#x27;max_features&#x27;: [2, 4, 6, 8],\n",
       "                          &#x27;n_estimators&#x27;: [3, 10, 30]},\n",
       "                         {&#x27;bootstrap&#x27;: [False], &#x27;max_features&#x27;: [2, 3, 4],\n",
       "                          &#x27;n_estimators&#x27;: [3, 10]}],\n",
       "             return_train_score=True, scoring=&#x27;neg_mean_squared_error&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" ><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "             param_grid=[{&#x27;max_features&#x27;: [2, 4, 6, 8],\n",
       "                          &#x27;n_estimators&#x27;: [3, 10, 30]},\n",
       "                         {&#x27;bootstrap&#x27;: [False], &#x27;max_features&#x27;: [2, 3, 4],\n",
       "                          &#x27;n_estimators&#x27;: [3, 10]}],\n",
       "             return_train_score=True, scoring=&#x27;neg_mean_squared_error&#x27;)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-16\" type=\"checkbox\" ><label for=\"sk-estimator-id-16\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(random_state=42)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-17\" type=\"checkbox\" ><label for=\"sk-estimator-id-17\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(random_state=42)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "             param_grid=[{'max_features': [2, 4, 6, 8],\n",
       "                          'n_estimators': [3, 10, 30]},\n",
       "                         {'bootstrap': [False], 'max_features': [2, 3, 4],\n",
       "                          'n_estimators': [3, 10]}],\n",
       "             return_train_score=True, scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "param_grid = [\n",
    "    # try 12 (3×4) combinations of hyperparameters\n",
    "    {'n_estimators': [3, 10, 30], 'max_features': [2, 4, 6, 8]},\n",
    "    # then try 6 (2×3) combinations with bootstrap set as False\n",
    "    {'bootstrap': [False], 'n_estimators': [3, 10], 'max_features': [2, 3, 4]},\n",
    "  ]\n",
    "\n",
    "forest_reg = RandomForestRegressor(random_state=42)\n",
    "# train across 5 folds, that's a total of (12+6)*5=90 rounds of training \n",
    "grid_search = GridSearchCV(forest_reg, param_grid, cv=5,\n",
    "                           scoring='neg_mean_squared_error', return_train_score=True)\n",
    "grid_search.fit(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "4d8a5a89",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-8 {color: black;background-color: white;}#sk-container-id-8 pre{padding: 0;}#sk-container-id-8 div.sk-toggleable {background-color: white;}#sk-container-id-8 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-8 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-8 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-8 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-8 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-8 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-8 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-8 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-8 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-8 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-8 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-8 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-8 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-8 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-8 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-8 div.sk-item {position: relative;z-index: 1;}#sk-container-id-8 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-8 div.sk-item::before, #sk-container-id-8 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-8 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-8 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-8 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-8 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-8 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-8 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-8 div.sk-label-container {text-align: center;}#sk-container-id-8 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-8 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-8\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomForestRegressor(max_features=8, n_estimators=30, random_state=42)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-18\" type=\"checkbox\" checked><label for=\"sk-estimator-id-18\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(max_features=8, n_estimators=30, random_state=42)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomForestRegressor(max_features=8, n_estimators=30, random_state=42)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Best Params\n",
    "grid_search.best_params_\n",
    "\n",
    "\n",
    "#Best estimator\n",
    "grid_search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "b84e29e5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sqt mean error:  36699.870050839854 {'max_features': 2, 'n_estimators': 3}\n",
      "sqt mean error:  32793.10325435628 {'max_features': 2, 'n_estimators': 10}\n",
      "sqt mean error:  31458.089978869604 {'max_features': 2, 'n_estimators': 30}\n",
      "sqt mean error:  35653.668075661975 {'max_features': 4, 'n_estimators': 3}\n",
      "sqt mean error:  32411.88977480537 {'max_features': 4, 'n_estimators': 10}\n",
      "sqt mean error:  31166.5177205135 {'max_features': 4, 'n_estimators': 30}\n",
      "sqt mean error:  34213.17861855719 {'max_features': 6, 'n_estimators': 3}\n",
      "sqt mean error:  31621.453023347487 {'max_features': 6, 'n_estimators': 10}\n",
      "sqt mean error:  30385.830853619733 {'max_features': 6, 'n_estimators': 30}\n",
      "sqt mean error:  34023.0020723454 {'max_features': 8, 'n_estimators': 3}\n",
      "sqt mean error:  31131.855019379138 {'max_features': 8, 'n_estimators': 10}\n",
      "sqt mean error:  29988.117689734587 {'max_features': 8, 'n_estimators': 30}\n",
      "sqt mean error:  36649.80807552118 {'bootstrap': False, 'max_features': 2, 'n_estimators': 3}\n",
      "sqt mean error:  32187.773926057336 {'bootstrap': False, 'max_features': 2, 'n_estimators': 10}\n",
      "sqt mean error:  36273.248715285066 {'bootstrap': False, 'max_features': 3, 'n_estimators': 3}\n",
      "sqt mean error:  30545.97718594391 {'bootstrap': False, 'max_features': 3, 'n_estimators': 10}\n",
      "sqt mean error:  34845.09576564536 {'bootstrap': False, 'max_features': 4, 'n_estimators': 3}\n",
      "sqt mean error:  31079.176022426047 {'bootstrap': False, 'max_features': 4, 'n_estimators': 10}\n"
     ]
    }
   ],
   "source": [
    "# Grid search all possible combinations\n",
    "cvres = grid_search.cv_results_\n",
    "for mean_score, params in zip(cvres[\"mean_test_score\"], cvres[\"params\"]):\n",
    "    print(\"sqt mean error: \", np.sqrt(-mean_score),  params)\n",
    "    \n",
    "    \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45b7948f",
   "metadata": {},
   "source": [
    "# Random Search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "4ef3eb3d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-9 {color: black;background-color: white;}#sk-container-id-9 pre{padding: 0;}#sk-container-id-9 div.sk-toggleable {background-color: white;}#sk-container-id-9 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-9 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-9 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-9 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-9 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-9 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-9 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-9 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-9 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-9 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-9 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-9 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-9 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-9 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-9 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-9 div.sk-item {position: relative;z-index: 1;}#sk-container-id-9 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-9 div.sk-item::before, #sk-container-id-9 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-9 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-9 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-9 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-9 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-9 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-9 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-9 div.sk-label-container {text-align: center;}#sk-container-id-9 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-9 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-9\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>RandomizedSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "                   param_distributions={&#x27;max_features&#x27;: &lt;scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec39ab6a0&gt;,\n",
       "                                        &#x27;n_estimators&#x27;: &lt;scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec7118640&gt;},\n",
       "                   random_state=42, scoring=&#x27;neg_mean_squared_error&#x27;)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-19\" type=\"checkbox\" ><label for=\"sk-estimator-id-19\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomizedSearchCV</label><div class=\"sk-toggleable__content\"><pre>RandomizedSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "                   param_distributions={&#x27;max_features&#x27;: &lt;scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec39ab6a0&gt;,\n",
       "                                        &#x27;n_estimators&#x27;: &lt;scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec7118640&gt;},\n",
       "                   random_state=42, scoring=&#x27;neg_mean_squared_error&#x27;)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-20\" type=\"checkbox\" ><label for=\"sk-estimator-id-20\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(random_state=42)</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-21\" type=\"checkbox\" ><label for=\"sk-estimator-id-21\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(random_state=42)</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "RandomizedSearchCV(cv=5, estimator=RandomForestRegressor(random_state=42),\n",
       "                   param_distributions={'max_features': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec39ab6a0>,\n",
       "                                        'n_estimators': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7fbec7118640>},\n",
       "                   random_state=42, scoring='neg_mean_squared_error')"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from scipy.stats import randint\n",
    "\n",
    "#Parameter distibution limits\n",
    "param_distribs = {\n",
    "        'n_estimators': randint(low=1, high=200),\n",
    "        'max_features': randint(low=1, high=8),\n",
    "    }\n",
    "\n",
    "\n",
    "forest_reg = RandomForestRegressor(random_state=42)\n",
    "#running random search\n",
    "rnd_search = RandomizedSearchCV(forest_reg, param_distributions=param_distribs,\n",
    "                                n_iter=10, cv=5, scoring='neg_mean_squared_error', random_state=42)\n",
    "rnd_search.fit(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "81fb5980",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29995.680519249894 {'max_features': 7, 'n_estimators': 180}\n",
      "31144.8907908471 {'max_features': 5, 'n_estimators': 15}\n",
      "30263.984734829122 {'max_features': 3, 'n_estimators': 72}\n",
      "30842.33175925967 {'max_features': 5, 'n_estimators': 21}\n",
      "30063.12349033762 {'max_features': 7, 'n_estimators': 122}\n",
      "30182.633404015352 {'max_features': 3, 'n_estimators': 75}\n",
      "29897.267059738264 {'max_features': 3, 'n_estimators': 88}\n",
      "29971.637256400685 {'max_features': 5, 'n_estimators': 100}\n",
      "29997.552885294186 {'max_features': 3, 'n_estimators': 150}\n",
      "36157.172712934494 {'max_features': 5, 'n_estimators': 2}\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'max_features': 3, 'n_estimators': 88}"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Showing all parameters\n",
    "cvres = rnd_search.cv_results_\n",
    "for mean_score, params in zip(cvres[\"mean_test_score\"], cvres[\"params\"]):\n",
    "    print(np.sqrt(-mean_score), params)\n",
    "\n",
    "#Showing the best estimator\n",
    "rnd_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "0aea94a3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.22193843032575272, 'OverallQual'),\n",
       " (0.18547357122156197, 'GrLivArea'),\n",
       " (0.09975365810044305, 'GarageCars'),\n",
       " (0.08871356645158868, 'ExterQual'),\n",
       " (0.0774430698079306, 'YearBuilt'),\n",
       " (0.07517701956936046, '1stFlrSF'),\n",
       " (0.07100224858900141, 'TotalBsmtSF'),\n",
       " (0.045195660085231695, '2ndFlrSF'),\n",
       " (0.04452941178622579, 'BsmtFinSF1'),\n",
       " (0.03593994616455523, 'TotRmsAbvGrd'),\n",
       " (0.02985089055098579, 'KitchenQual'),\n",
       " (0.02498252734736268, 'BsmtQual')]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Showing future importances for the best model\n",
    "feature_importances = rnd_search.best_estimator_.feature_importances_\n",
    "feature_importances\n",
    "\n",
    "attributes = list(x.columns)\n",
    "sorted(zip(feature_importances, attributes), reverse=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3f8ddfc",
   "metadata": {},
   "source": [
    "# Evaluate your model in the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "876fd5d2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean square error total sum 689434591.4845568\n",
      "Root mean square error total sum 26257.0865003061\n"
     ]
    }
   ],
   "source": [
    "#Loading test set\n",
    "\n",
    "# dependent features (independent feature)\n",
    "xt = x_test.copy()\n",
    "\n",
    "# y feature (expected feature)\n",
    "yt = y_test.copy()\n",
    "\n",
    "#Preventive reset index and squeeze to avoid bugs\n",
    "yt.reset_index(inplace=True)\n",
    "yt.drop(\"index\", axis= 1, inplace=True)\n",
    "yt = yt.squeeze()\n",
    "\n",
    "\n",
    "#mean square error\n",
    "\n",
    "from sklearn.metrics import mean_squared_error\n",
    "final_model = rnd_search.best_estimator_\n",
    "final_predictions = final_model.predict(xt)\n",
    "final_mse = mean_squared_error(yt, final_predictions)\n",
    "final_rmse = np.sqrt(final_mse)\n",
    "print(\"Mean square error total sum\", final_mse)\n",
    "print(\"Root mean square error total sum\", final_rmse)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee56e465",
   "metadata": {},
   "source": [
    "# Dataframe with results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "c262f557",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error max  =  142.9\n",
      "Error mean  =  12.1582191780822\n",
      "Error std  =  18678.567688787796\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>OverallQual</th>\n",
       "      <th>GrLivArea</th>\n",
       "      <th>2ndFlrSF</th>\n",
       "      <th>BsmtFinSF1</th>\n",
       "      <th>GarageCars</th>\n",
       "      <th>TotRmsAbvGrd</th>\n",
       "      <th>TotalBsmtSF</th>\n",
       "      <th>1stFlrSF</th>\n",
       "      <th>YearBuilt</th>\n",
       "      <th>ExterQual</th>\n",
       "      <th>KitchenQual</th>\n",
       "      <th>BsmtQual</th>\n",
       "      <th>Predicted</th>\n",
       "      <th>Expected</th>\n",
       "      <th>Error abs</th>\n",
       "      <th>Error %</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.593095</td>\n",
       "      <td>0.426634</td>\n",
       "      <td>0.114812</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.134861</td>\n",
       "      <td>0.377220</td>\n",
       "      <td>0.074074</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.5</td>\n",
       "      <td>206379.500000</td>\n",
       "      <td>204750</td>\n",
       "      <td>1630.0</td>\n",
       "      <td>0.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.777778</td>\n",
       "      <td>0.604401</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.172927</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.416667</td>\n",
       "      <td>0.301800</td>\n",
       "      <td>0.646573</td>\n",
       "      <td>0.029630</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.5</td>\n",
       "      <td>280484.647727</td>\n",
       "      <td>257000</td>\n",
       "      <td>23485.0</td>\n",
       "      <td>9.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.444444</td>\n",
       "      <td>0.484528</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.163359</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.215057</td>\n",
       "      <td>0.518336</td>\n",
       "      <td>0.392593</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>159430.681818</td>\n",
       "      <td>145000</td>\n",
       "      <td>14431.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.444444</td>\n",
       "      <td>0.417112</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.056697</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.177741</td>\n",
       "      <td>0.446217</td>\n",
       "      <td>0.414815</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>130897.727273</td>\n",
       "      <td>144900</td>\n",
       "      <td>14002.0</td>\n",
       "      <td>9.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.589312</td>\n",
       "      <td>0.340436</td>\n",
       "      <td>0.087704</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.174141</td>\n",
       "      <td>0.438472</td>\n",
       "      <td>0.140741</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.5</td>\n",
       "      <td>208157.397727</td>\n",
       "      <td>239000</td>\n",
       "      <td>30843.0</td>\n",
       "      <td>12.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.555556</td>\n",
       "      <td>0.470245</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.173281</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>0.206547</td>\n",
       "      <td>0.503056</td>\n",
       "      <td>0.229630</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.5</td>\n",
       "      <td>166941.113636</td>\n",
       "      <td>181500</td>\n",
       "      <td>14559.0</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   OverallQual  GrLivArea  2ndFlrSF  BsmtFinSF1  GarageCars  TotRmsAbvGrd  \\\n",
       "0     0.555556   0.593095  0.426634    0.114812        0.50      0.416667   \n",
       "1     0.777778   0.604401  0.000000    0.172927        0.50      0.416667   \n",
       "2     0.444444   0.484528  0.000000    0.163359        0.25      0.250000   \n",
       "3     0.444444   0.417112  0.000000    0.056697        0.50      0.333333   \n",
       "4     0.666667   0.589312  0.340436    0.087704        0.50      0.250000   \n",
       "5     0.555556   0.470245  0.000000    0.173281        0.50      0.333333   \n",
       "\n",
       "   TotalBsmtSF  1stFlrSF  YearBuilt  ExterQual  KitchenQual  BsmtQual  \\\n",
       "0     0.134861  0.377220   0.074074   1.000000     0.666667       0.5   \n",
       "1     0.301800  0.646573   0.029630   0.333333     0.666667       0.5   \n",
       "2     0.215057  0.518336   0.392593   1.000000     1.000000       1.0   \n",
       "3     0.177741  0.446217   0.414815   1.000000     1.000000       1.0   \n",
       "4     0.174141  0.438472   0.140741   0.333333     0.666667       0.5   \n",
       "5     0.206547  0.503056   0.229630   1.000000     1.000000       0.5   \n",
       "\n",
       "       Predicted  Expected  Error abs  Error %  \n",
       "0  206379.500000    204750     1630.0      0.8  \n",
       "1  280484.647727    257000    23485.0      9.1  \n",
       "2  159430.681818    145000    14431.0     10.0  \n",
       "3  130897.727273    144900    14002.0      9.7  \n",
       "4  208157.397727    239000    30843.0     12.9  \n",
       "5  166941.113636    181500    14559.0      8.0  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbwAAAGoCAYAAAA991BSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAACAhUlEQVR4nOzdd5gV5dnH8e/MnN6277KVjh2l2OKKPaixJVEjJprEjoqCooCKiIqKChYIxkRN8mJEsSQxakyMRpGIqEREUSxI2d7L6W3m/eOwyy7sLgts3/tzXVzA2TlznlmFH8/Mc9+PYhiGgRBCCDHAqb09ACGEEKInSOAJIYQYFCTwhBBCDAoSeEIIIQYFCTwhhBCDgqm3B9ATqqq8vT0EIYTodhkZ7t4eQp8mMzwhhBCDggSeEEKIQWFQ3NIcSA4fdyBlJaWdOjY7N4fPPt3UzSMSQoj+QQKvnykrKWXma7M7dezDZz3QzaMRQoj+Q25pCiGEGBQk8IQQQgwKcktzIFMgM9PT6cPlmZ8QYiCTwBvIDDr9vA/kmZ8QYmCTW5pCCCEGBQk8IYQQg4Lc0hzIVCjybqfEW0xtqAZ/zI9hGFg0C26Lmwx7JvnuAlJsqb09UiGE6HYSeAOQN+JlXcXHcDO88PVzALgtHlxmF6qiUh+uZ3vjNiJ6BIA0WzqjkkdDWm+OWgghupcE3gASjUdZW76Gj8vXohs6bINzzv4xBe6h2Ey2VscahkF9uJ4tDZv5rv5bPir/EKbBL16/kKsPv47C3EkoitJLVyKEEF1PnuENEOX+cpZ/9Qc+LPuAMSkHcvlhV8NKGJNywG5hB6AoCim2FMZnTeTCA6ZwzeHXw7vwv8pP+OmrZ3Pe387kfxWf9PyFCCFEN5HAGwA+r97Aik3LicajXDDmIn404mySrcl7dQ6n2ZkIvEu+5P7jH+bbuq85/eWTuepfv6LUV9It4xZCiJ4ktzT7McMw+LDsA/5b+j5D3cM4a+S52E32fT+hAgU5mYlfW4Dj4K/RV/jrxlfgHeAjwNh5uBSqCyH6Ewm8fsowDFaXrGJt+RoOSTuUycPORFX2c8LeRqF6fbief2/7J1vP2MKQn2ZzxvCzSLMnVrdIoboQoj+RW5r9kGEYvFv8DmvL1zA2/QhOH/aj/Q+7diRbk/np6Av50fBzaIjUs/yrP/Bp5ToMw9jzm4UQog+RGV4/tKbsv6yr+JhxmeM5Of+0bl9NqSgKB6UdTL67gDe3vsHb29/i+/rN4OrWjxVCiC4lM7z+5gj4oHQ1B6cd2iNh15LL4uKnoy/glILTKPJuh2vhra1v9tjnCyHE/pDA60f+s/1tOBuGuocxeegZvVInpygK4zIncMnBv4IG+PkbF3LvmruI6bEeH4sQQuwNCbx+4tu6b7j8n5dCNZwz8jw0VevV8aTZ0+FpuOTgX/H4p4s5/9VzqPCX9+qYhBCiIxJ4/YAv4uVX/7gYm8kKfwZrG4XkvSIGi058nKWnPMn6yv9x8spCVpes6u1RCSFEmyTw+jjDMLjhnWvZ3PAdv/vhH6Gxt0e0uwsPmMKb5/+HZGsy5796Dos/eTDR2kwIIfoQCbw+bun6x3jt+79x57H3UJg7qbeH064DUw/inxe8y3mjfsoDH93Lxa+fT02wpreHJYQQzSTw+rBVxe+y4MO7OHfkT5h6+PW9PZw9cpldPHHqUzw46RFWF6/ilJWFfFS2treHJYQQgARen1Xk3c5V//oVY1IO4JGTl/abnQsUReFXh17OGz/9N2bNzHl/O4PffPq4FKoLIXqdBF4fFIqFuOzNS4jqMf5w+rO4zP2vwntsxhG8fcH7nD7sR8xfcweX/uMi6kK1vT0sIcQgJp1W+hjDMJi96mY+q/qU/zvjeUYmj+7tIbVPgcxMz56POwr+OfkfHPz5SP7+y38ycchR3T82IYTYhQReH/Onjc/w3Kbl3DThFk4ffmZvD6djbTSbbk+Zv4w/f/Anzvnr6dx57N1cPfa6fnObVggxMMgtzT5kbdmH3L76Vk4t+CG3HHlbbw+nS2U7s+FJOG3o6dz539v45T+mUBWo6u1hCSEGEQm8PqLcX8bl/7yEPHc+T5z2VK93UukWIfjj6X/m3uMe4D9Fb3PCC8fw5pY3entUQohBQm5p9gGReITL3rwEX8THi2f/jaS93K2831AgKysp8etMCP+kikuDF8H/gDeBSOvDZYNZIURXksDrA25fPYtPKj7iqR/+iYPSDu7t4XSfXZ75xfU4/y19n4/Gf4jnmCR+OPR0hiUNb/66bDArhOhKckuzl/1+wxP8aePTXD9uOueM+nFvD6dHaarGpLwTueiAn6MpKi99+wKvf/8q/qi/t4cmhBiAZIbXi974/jXuWD2bM4efze1Hz+vt4fSaPHc+vzzkctaWrWFt+Rq2NHzPpLyTQBZxCiG6kMzwesn/Kj5h6r8vZ3zWBJad+vuBuUhlL5hUE8flHs8vD76cdHsG/9r2D7gS3i9+r7eHJoQYIGSG1wu+qf2aX7xxIZmOLIoeLmLY1CG9PaQ+I82exs8OuJivajfyRv1r/PTVszml4DTuOGY+h6Qf2tvDE0L0YxJ4Pez7+u/46atnoyoaz5/1MsfeNKHTxdswOBZyKIrCwWmH8sbS15j3+r08+r+HOXnlcfx49PncMP4mDk47pLeHKIToh+SWZg/a1riVn/ztbGJ6lJfP+XvfbhvWF8TgunE38PHPP+O6cTfy5pY3OPGFY/n56xfwfvF70pBaCLFXJPB6yDe1X3PeX88kEPPz4jmvckDqgb09pH4j2ZbCncfezaeXbmT2UXfwv4pP+OmrZ/ODFRN4Yv1SKgIVvT1EIUQ/IIHXA9aWfchZfzmNSDzCy+e+xqHph/X2kPqlFFsqN028lf9d+iVLT3mSNFs68z64jbF/HMN5fz2Tpz//nYSfEKJd8gyvm/1989+47t9XkuPK5YWz/8JQz7DeHlL/0ZndGDLAOMTgg0NW80Hpam57/xYmZB3JpPwTOSHvJCZkHYlFs/TMeIUQfZoEXjeJxqMsWDufZesfZ0LWkTx75krS7Gm9Paz+ZS92YwB4+LIHMA42+GTUR3xS9hGLP3kw0a5sO1AEFAMlQChxvLQuE2JwkcDrBqW+Eq5+6zLWlq3h14dewd3H3Y9Vs/b2sAa+Spj5TCIgQ7EQRd7tbPdupSipiOpRO3dmSLWlkePK5Ys/bGB95f84KO0Q+e8jxCAggdeFdENnzMVDaTyyIfF09FX4w11P8Qee6u2hDTo2k43RKWMYnTIGgHA8TLm/jDJfKaX+EjbXfwtnww9fOhGTauLA1IM5POMIDss4nMMzjuDgtEOxm+y9fBVCiK4kgddFvqzZyOxVN9N4fAMF7qGcNvR0Un6Qssf3DYa6ur7AqlkZ6hnW/AzVMAwWXboQsiGWHeOL7A18kbMBHDveoANVQClQlviRpQ7h84+/6ZXxCyH2nwTeftrWuJUHP7qPl755AY81Cf4KF9xzkezm3ccpigJ1MHP5zmeEhmHgjTRSESinIlBBRXI5FbnlBGIBACr0co57biIHph3MgakHcWDqwRyUejDDkoZjUuWPkhB9nfwp3Ucbqtbzuw1P8JdvX0JTNK4bdyPTxk3ngDnDJOz6KUVR8FiT8FiTGJ1yAJAIQV/UR2WgnL889TKjf30AX1Rv4LXNf8MgUfhu1ayMTjmgOQRHJY9mWNJwhnmG4zA7OvpIIUQPksDbC43hBl7//u+s2PQsH5Z9gMPk5NJDfs0N424i25XT28MT3UBRFNwWN26LG96DN979e+ILZiCdxEa2mWG+yNzAF1kbYJcqiiHObIZ5hjM8aQTDk0aQ585niDObbGc2Q5w5OM3Onr4kIQYtCbw9KPeX8c72f/OvrW/y9vZ/EY6HGZ40gvk/uI+LD/rFwN2dXOyuE2USoViI+nAd9eE6Xvv9q5x43clsafied7b/m4pA+W7Huy0esp3ZpNhSSbGmkGRNJtmWQrI1mWRr4me3xY3L7MZpduI0u3b8nPj1YN9lQ4i9IYHXgV+/+Qte//5VIPEv9UsP/jU/HXMh4zInyG1L0SabycYQUzZDnNm8tvpVnn//zzu/aCYxA/QA7sTPXncjXncj2MHkMZE1dAj14Xr8UV/nPjBKotYwAoRb/LqNHx6bh/m33dcqMF0W1y5B6kJVpAGTGJgk8DpwTPaxjMscz8kFp3FI2qEScmLv7G3h/FkP8GnllwBE4hEawg00hOvxRhrxx/z4o378UR++iA9/1M+8Bbdx5EVHE9EjROORHT9HiegRIvEIUT1CVI8SiUeIG3EaaWTGu9fvcRwOkwPHjlDcObPcGYouiwunaedM02Vx4zA5MGsWzKoJk2rGvOOHSTVh0SzNr5lUU4uvmZuPt2gWNEWTP2OiW0ngdeDqw6/j8HEHcm/JXb09FDEYdKaV2i5OuOWkTh0X1+M88rOHErNMC2Dd8XMbPwKWAAFLgGpb1c7j2/rRHX97xEmUhBg7ft7xQ1VUcrJyUVUNk6KhKRom1YSmmnb8WkNTTGiqhmnHz83HNL+utTi+6XV1x/EtX1fRVFOL85h2nD8RyDE9RtzQiRtxdD1O3IgT02PoRrz59bgeJ27Emn+98qUVBIIBUEj8UOnw1ymbU/n61a3d8A0e3BRD9lgRQggxCMjNeiGEEIOCBJ4QQohBQQJPCCHEoCCBJ4QQYlCQwBNCCDEoSOAJIYQYFCTwhBBCDAoSeEIIIQYFCTwhhBCDggSeEEKIQWFQ9NKsqvL29hCEEKLbZWS4O33sQP17saPvgczwhBBCDAoSeEIIIQYFCTwhhBCDggSeEEKIQUECTwghxKAggSeEEGJQkMATQggxKEjgCSGEGBQk8IQQQgwKEnhCCCEGBQk8IYQQg4IEnhBCiEGh25pHP/nkk7zzzjtEo1GmTJnCUUcdxezZs1EUhdGjRzNv3jxUVWXlypU8//zzmEwmpk6dykknnUQoFOKWW26hpqYGp9PJwoULSU1NZf369SxYsABN0ygsLOT6668HYOnSpbz77ruYTCZuu+02xo4d212XJYQQop/qlhne2rVr+fTTT1mxYgXLly+nvLyc+++/n+nTp/Pcc89hGAZvv/02VVVVLF++nOeff56nn36axYsXE4lEWLFiBWPGjOG5557jvPPOY9myZQDMmzePRYsWsWLFCj777DM2btzIxo0b+eijj3jxxRdZvHgx8+fP745LEkII0c91S+CtXr2aMWPGcN1113HNNddw4oknsnHjRo466igAJk2axAcffMCGDRsYN24cFosFt9tNQUEBmzZtYt26dRx//PHNx65Zswafz0ckEqGgoABFUSgsLGTNmjWsW7eOwsJCFEUhJyeHeDxObW1td1yWEEJ0L8NAqazs7VEMWN1yS7Ouro7S0lJ++9vfUlxczNSpUzEMA0VRAHA6nXi9Xnw+H273zr2LnE4nPp+v1estj3W5XK2OLSoqwmq1kpyc3Op1r9dLampqd1yaEEJ0D8NArayAQACDzN4ezYDULYGXnJzMiBEjsFgsjBgxAqvVSnl5efPX/X4/Ho8Hl8uF3+9v9brb7W71ekfHejwezGZzm+cQQoj+RK2sQPF5MVStt4cyYHXLLc0JEybw/vvvYxgGFRUVBINBjj32WNauXQvAqlWrmDhxImPHjmXdunWEw2G8Xi+bN29mzJgxjB8/nvfee6/52AkTJuByuTCbzWzfvh3DMFi9ejUTJ05k/PjxrF69Gl3XKS0tRdd1md0JIfoVtaIcxTcwdyDvSxTDMIzuOPGDDz7I2rVrMQyDGTNmkJeXx9y5c4lGo4wYMYJ7770XTdNYuXIlL7zwAoZhcPXVVzN58mSCwSCzZs2iqqoKs9nMokWLyMjIYP369dx3333E43EKCwuZMWMGAEuWLGHVqlXous6cOXOYOHFiq7EM1K3shRD9n1JRgeprbP69oWrow0fs07kyMjp/d2ug/r3Y0feg2wKvLxmo/2GFEP3brmEHEnj7q6PvgRSeCyFEL2gr7ET3ksATQogeplRWStj1Agk8IYToQUplJaq3obeHMShJ4AkhRA9Rqqok7HqRBJ4QQvQApaoKtbG+t4cxqEngCSFEN1OqqyXs+gAJPCGE6EZKTQ1qQ12njlWLi/Bcfinu667q5lENTt22PZAQQgx2Sk0Nan3nmtmrW77Hc8NU1Opq1Jpq0HVQZU7SleS7KYQQ3UCp7XzYaV9/hefaK1Grq9Hdbhqf/j8Ju24g31EhhOhiSl0tal3nws70+Qbc11+DWl+PnpJC47LfE5t4VDePcHCSwBNCiC6k1Neh1tZ06ljTJx/hvvFaVJ8PPT2DxmW/Jz7mwG4e4eAlgSeEEF1Eqa9LPH/rBPN/38d9840owSDx7Bwaf/sU+rDh3TzCwU0CTwghuoDSUN/5sHv7LVyzbkaJRIgXDE2EXW5eN49QyCpNIYTYT0pjA2p1VaeOtbz+d5z33Y2i68RGjsL7+DKM1LRuHqEACTwhhNgvSmMDalVlp461vvIizoceACB20CF4H1mCkZTUncMTLUjgCSHEPlK8jZ0OO9uz/4fjN48BED1iHN6HHwWnqxtHJ3YlgSeEEPtA8TaiVlbs+UDDwP7Uk9if+T0A0aOPwfvAw2Czd/MIxa4k8IQQYi/tVdgteQT7ij8DEJl0Ir577geLpZtHKNoigSeEEHuh02Gn6zgeegDbX18GIPzD0/HPvQtM5o7fZ5K/lruLfGeFEKKTFJ+3c2EXi+FcMB/rm28AEDrnPAK33gaa1uHbDIsVPTunK4Yq2iCBJ4QQnaD4vKgV5Xs+MBLBdedtWN77DwChC6cQmH4zKEqHbzPsDvQh2dJDsxtJ4AkhxB50emYXCuKacyuWDz8AIPirywleNXXPYed0oWcN2eNxYv9I4AkhREd8vkTYGUbHx/n9uG+ZgfnTdQAEpl5P6NJf7/H0uicZIyMDgGAsSE2wmnAshNVkI82ejt0kqzm7isydhRCiPT4fWmX5HsNOaWzEc8PU5rDz33Rr58IuJbVV2BV7t6MbOg6zE93QKfZuJxgL7v91CEACTwgh2tbZsKutxX391Zi+3Iihqvhun0f4gp/t8fR6RmarlmI1wWqsmg2LZkFRFCyaBatmoybYuf6ce+vMl0+lNtS5XR0GCgk8IYTYld/fubCrrMQz9QpM336DoWn45y8gctY5HZ9bUYhnZWN4WrcUC8dCmNXWJQtm1Uw4FtqnS9iTTyo+4rhzjuyWc/dV8gxPCCFa8vvRKsr2GHZqaQnuaVPRSkswzGZ8Cx4kevykDt9jqFpiJaZ99+dyVpONqB7Fou0sSo/qUawm275dRyfUVHXP7LGvkhmeEEI06WzYbduK55orEmFns+F9+LE9h51mSmwB1EbYAaTZ0wnHQ0TiEQzDIBKPEI6HSLOn7/Pl7NEgWxQqgSeEENDpsNO+/QbP1CtQqyrRnU68j/6G2FFHd/gew2xBz8vvsKWY3WQnz12AqqgEon5URSXPXSCrNLuQ3NIUQohAoHNht/Fz3DOmoXq96J4kvI/9hviBB3X4HsNqS3RP2UOXFWgKvfy9GrroPAk8IcTgFgiglZfuMexM/1uH+5bpKIEAeloa3seWER85qsP3GA5n4pmdFJT3CRJ4QojBq5NhZ17zAa7ZM1EiYeJZWXiX/BY9v6DD9+guD0ZmpoRdHyLP8IQQg1Nnw+7dd3DdOiMRdnn5eJ94es9hl5yKkZUlYdfHyAxPCDH4BIOdCjvLm2/gvPculHic2PAReB9fhpGe0eF79PQMjKTkLhxsNxpkeSyBJ4QYXIJB1LI9h531r6/gePA+FMMgdsCBeB9dipGc0v4bFAU9MwvD5e7iAYuuIoEnhBg8doSdYugdHmZ9/jmcjy0CIHrY4fgWP9ZhkBmKmlic4nB06XBF15LAE0IMDqHQnsPOMLD94Skcv/8tANGJR+F9cHG7xeKwo3tKTi5YrV09YtHFJPCEEANfKIRaWrLHsLMvW4L92T8BEDnueHwLFnYYZIbJnAg7s7ndY/qytIy0PR80gEjgCSEGts6Ena7jWPQgtldeBCB88qn477q3wyAzLNZE2HWioLyv+sMzf+7tIfQoCTwhxMDVmbCLxXDefw/WN14DIPyjs/HPmdthkBl2R+KZndq/K7tiRqy3h9CjJPCEEANTOLznZ3bRKM677sD6zr8BCJ3/MwIzZnYYZIbLjZ45MGrsIvFIbw+hR0ngCSEGnnA4MbPT4+0fEwrhun0Wlg9WAxD8xS8JXjutwyDTk1Iw0rtx94IeFtWjvT2EHiWBJ4QYWDoTdoEA7ltvwrzu48Rvr76W0C8v6zjsUtMwUlK7erS9SmZ4QgjRX3Ui7BSvF9dNN2D+YgMA/htvJnzRxe2fU1HQMzIx3J6uHm2vi+oSeF3ivPPOw+1OFGrm5eVxzTXXMHv2bBRFYfTo0cybNw9VVVm5ciXPP/88JpOJqVOnctJJJxEKhbjllluoqanB6XSycOFCUlNTWb9+PQsWLEDTNAoLC7n++usBWLp0Ke+++y4mk4nbbruNsWPHdtdlCSH6qs6EXX0d7huvw/TN1xiKQmDWbYTP/Un751QU4lnZ4HR2w4B7n8zwukA4HAZg+fLlza9dc801TJ8+naOPPpo777yTt99+myOOOILly5fz8ssvEw6HufjiiznuuONYsWIFY8aMYdq0abz++ussW7aMO+64g3nz5rFkyRLy8/O56qqr2LhxIwAfffQRL774ImVlZUybNo2XX365Oy5LCNFXRSJ7DruqKtw3Xotpy/cYmoZ/7nwik89o93hD1RL72Nls3THiPkGe4XWBTZs2EQwGueyyy4jFYtx0001s3LiRo446CoBJkybx3//+F1VVGTduHBaLBYvFQkFBAZs2bWLdunVcccUVzccuW7YMn89HJBKhoCDRpbywsJA1a9ZgsVgoLCxEURRycnKIx+PU1taSmjqw7rULIdoRiaCWFHcYdmpZGe5p16CVFGOYzfjuuZ/oCSe1e7yhmRI1dh3sUD4QyAyvC9hsNi6//HIuuOACtm7dypVXXolhGCg7Hgg7nU68Xi8+n6/5tmfT6z6fr9XrLY91uVytji0qKsJqtZKcnNzqda/XK4EnxGDQmbAr2p4Iu4oKDKsV3wMPEz3mB+0eb1isiZmdaeAvcRhsz/C6pWpy+PDhnHPOOSiKwvDhw0lOTqampqb5636/H4/Hg8vlwu/3t3rd7Xa3er2jYzs6hxBigOvEbUztu2/xXHNFIuwcTryPLO047Gx29Ny8QRF2AA89cn9vD6FHdUvgvfTSSzzwwAMAVFRU4PP5OO6441i7di0Aq1atYuLEiYwdO5Z169YRDofxer1s3ryZMWPGMH78eN57773mYydMmIDL5cJsNrN9+3YMw2D16tVMnDiR8ePHs3r1anRdp7S0FF3XZXYnxEDXFHbx9juFaF99ifu6q1Fra9DdHhofX0Zs3Ph2jzecrsRtzH7ePWVveH3e3h5Cj1IMYw+bQu2DSCTCnDlzKC0tRVEUZs6cSUpKCnPnziUajTJixAjuvfdeNE1j5cqVvPDCCxiGwdVXX83kyZMJBoPMmjWLqqoqzGYzixYtIiMjg/Xr13PfffcRj8cpLCxkxowZACxZsoRVq1ah6zpz5sxh4sSJrcZTVTW4/qMKMaBFo4nbmB2EnemzT3HdfCOq34+ekor38WXER41u93jdnYSRmdkdo+1RGRmdv7ulzFPgfah8obEbR9TzOvoedEvg9TUSeEL0vGAsSE2wmnAshNVkI82ejt3U/jY7ndKZsPtoLe5ZN6GEQugZmTQueQJ96LB2j9dTUjFSB8auAXsVeHcosBYqnxs8gTd45u5CiB4TjAUp9m5HN3QcZie6oVPs3U4wFtz3k3Yi7Mzvr8I980aUUIh4Ti6Nv32q47BLzxgwYbfX4kD/3ehhn0jgCSG6XE2wGqtmw6JZUBQFi2bBqtmoCVbv2wmj0T0+s7P8+1+45sxEiUaJDx1G4xNPJZ7JtWVHQbmRlLxv4xkIJPCEEGL/hWMhzGrrveTMqplwLLT3J4vFEmEXa79I2vL3v+G88zaUeJzY6DE0Lvt9u8/kDEUlnp0LLcqcBqVBGHiDY+2tEKJHWU02onoUi7azcDuqR7Ga9rJrSSyWuI3ZQdhZX3wB5+IHE4cfcijexUswPG33vTQ0U6LGroNdzAeNQRh4MsMTQnS5NHs64XiISDyCYRhE4hHC8RBp9r3YWqcTYWf70zPNYRcdP4HGx5a1H3ZmS6LGTsIuQQJPCCH2n91kJ89dgKqoBKJ+VEUlz13Q+VWae7qNaRjYf/sbHL/9DQCRY4/Du+jxdps8G1ZbIuzM5ja/PigNwsCTW5pCiG6RCL38vX9jU9hF22l7ZRg4Hl2EbeUKACInnoxv/oJ2+14adgf6kOxBVVDeKToSeEII0Wvi8Y7DLh7H8eB92F79KwDh08/Ef/u8dluBGS43emZWhxu7DloywxNCiF4Sjyee2bUXdrEozrvnYX3rnwCEzvspgVtmtztz05NSMNL34pnhYKMz6BJgkF2uEKJP2tPMLhLBNXc2llWJHrvBKb8gOG16uzM3PTUNI0V66nYozqBbxSGBJ4ToXU1hFwm3/fVQEPfsmZjXfghA8PKrCF5+Vfthl5GJ4UnqrtEOHDLDE0KIHrSnsPP7cM+cjnn9pwAErruR0C8ubfvYHd1T2lupKXYhMzwhRH/SLQ2ae8oewk5pqMc9Yxqmr74EwH/LbMI/uaDNYw1VS6zEtPeTa+8LBuEqzUGW70IMHN3SoLmn6DpqWWn7YVdTjfvaqzB99SWGquKbO7/9sNNMiRo7Cbu9ozPoEmCQXa4QA0eXN2juKbqemNmF2+6rqVaU47n2Kkzfb8YwmfDdcz+RM89q81jDbEHPy2+3Bk90QAJPCNFfdGmD5p6yp7ArLsJ9zRVo27dhWCz4Fi4ievKpbR7b3D2lnRo8sQfyDE8I0V90WYPmntJ0G7O9sNvyPZ4bpqJWV2PY7XgfeoTYhCPbPNZwOBPP7KSgfN/JMzwhRH/RJQ2ae0pT2IXafr6ofb0Jz7VXolZXo7tceB9b1m7Y6e6kxI4HEnb7R25pCiH6i/1u0NxTDKPDsDN9vgH39Vej1tejJyfjXfokscPGtnmsnpLa7j53Yi8pJEJvEJFbmkL0Y/vcoLmn7Cns1n2M+5YZKMEgeno6jY8/gT58RPPXQ7Ew9eE6IrEwypAcUjwO+lic918qgy7wZIYnhOgeTWEXDLT5ZfMHq3HfdCNKMEh8SDaNTzy1W9iV+0vRMdDyhhFzOftP2UV/YAJ0OHzcgb09kh4jgSeE6Hp7Crt3/o1r1s0okTDxgqF4f/tUoryghfpwHWazA3LzwensP2UX/UhqbiplJaW9PYweI4EnhOhahoFaXtZu2FneeA3X3DkosRixkaNofOL36FlDdjsuTAxy8zBsO1ed9vmyi/5EBWWQRYA8wxNCdJ2msAv42/yy9ZUXcT70AACxgw7B+8gSjKTdGz0bFiukDiOiKbQsKe/TZRf9jQqqMrgCb3BdrRCi++wh7GzP/l9z2EWPGEfjkmVth53Njp6bR5p7SP8pu+iPJPCEEGIfGAZqRXnbYWcY2H//Wxy/eQyA6NHH4H1kCThdux/qdKHn5IKq9p+yi/5qEAae3NIUQuyfprDz+9r8mn3Jo9hXPAtAZNIJ+O55oM3el7o7abcauz5fdtGfSeAJIfqLPrE1UEdhp+s4Hn4A219eBiB82mT8d84Hk3n3Q1NSMVLTunu0oiUVFAZXt5rBFe9CDBB9Ymsgw0CtrGg77GIxnPfc1Rx2oXPOwz/vnrbDLiNTwq43KINvhje4rlaIAaIvbA2kVlag+Ly7fyEaxTV3DtY3XwcgdOEUArPvAG2XTsU7dig3PLsvXBE9QG5pCiH6g3AshMPsbPWaWTUTiLa9QrKrqRXlbYddKITrtluxrPkvAMFfXU7wqqm7NXqWHcr7ABUUCTwhRF/Xm1sDtRt2fj/uW2dg/t86AAJTryd06a93O8zQTImVmLJpa+9SQR1kO04MrngXYoDora2B2gs7pbERz43XNoed/6Zb2w472aG8z1DNKps/+I7s3JzeHkqPkcAToh/qjRo1paLtZ3ZKbS3u66/GtPELDFXFd/s8whf8bLfjmgrKZYfyvkE3dM45+8d89umm3h5Kj5H/84Top3qyRk2pqED1Ne7+emUlnhumom3biqFp+O+6l8ipP9ztOMPpSvTLHGS30Po0BTR5hieEEDu1F3ZqaQnuaVPRSkswzGZ8Cx4kevyk3Y5rq6Bc9AEKKIPsHyASeEKIdrUbdtu24pk2FbWqEsNmw7twMbGjjt7tOCko78MU2S1BCCGAxO3KtsJO++5b3Ddci1pXi+504lv0GLHDx+12nJ6RKTV2fdkgLDyXwBNC7EaprET1Nuz2urbxC9wzpqF6G9E9SXgfXUr8oIN3ebNCPHMIuHZvDi36EAk8IcRgp1RVtRl2pk//h3vmjSiBAHpaGt7HlhEfOarVMVJQ3o/IMzwhxGCmVFWhNtbv9rr5ww9wzZ6JEg4Tz8rCu+S36PkFrY6RgvJ+Roe4Hu/tUfSowTWfFUK0S6mubjvs3n0H1y0zEmGXl4/3iad3DzspKO9/4hCJR3p7FD1KZnhCCJSaGtSGut1et/zzHzjvmYcSjxMbPgLv48sw0jNaHWNYbejZObs3hxZ9WwzCeri3R9GjJPCEGOSU2hrU+trdXrf+7RUcC+9DMQxiBxyI99GlGMkprY4xHM7EM7tB9ixoQIgnmpAPJhJ4QgxiSm0Nal0bYff8czgfWwRA9LDD8S16DMPtbnWM7vIkCsol7Polt90z6G5pdtszvJqaGk444QQ2b97Mtm3bmDJlChdffDHz5s1D13UAVq5cyU9+8hMuvPBC/vOf/wAQCoWYNm0aF198MVdeeSW1tYk/jOvXr+eCCy7goosuYunSpc2fs3TpUs4//3wuuugiNmzY0F2XI8SAo9TV7h52hoHtD0/tDLsJR+J97De7h11yKkZWloRdP+atbsQXbWPz3gGsWwIvGo1y5513YrMltiq5//77mT59Os899xyGYfD2229TVVXF8uXLef7553n66adZvHgxkUiEFStWMGbMGJ577jnOO+88li1bBsC8efNYtGgRK1as4LPPPmPjxo1s3LiRjz76iBdffJHFixczf/787rgcIQYcpa4Wtbam9YuGgf2JpTh+9wQAkeOOx7vosd1KDPT0DIw06Z7S74WgLrz7c9uBrFsCb+HChVx00UVk7uift3HjRo466igAJk2axAcffMCGDRsYN24cFosFt9tNQUEBmzZtYt26dRx//PHNx65Zswafz0ckEqGgoABFUSgsLGTNmjWsW7eOwsJCFEUhJyeHeDzePCMUQrStzbDTdRyLHsS+/I8AhE85Dd8DD4HV2uKNCnrWEIyk5B4bq+hGQagPSeDtl1deeYXU1NTm0AIwDKO5wNHpdOL1evH5fLhb3CZxOp34fL5Wr7c81tWia8OeXhdCtK3NsIvFcN53N7aXVwIQ/tHZ+OcvAJO5+RBDUYkPycFwtb61KfqxIDRGGojpsd4eSY/p8kUrL7/8MoqisGbNGr766itmzZrVatbl9/vxeDy4XC78fn+r191ud6vXOzrW4/FgNpvbPIcQYndthl00inP+XKxvvwVA6KcXErjpFlB3/lvYULVEQXnL2Z7o/4KJn+rD9aR388bBfUWXz/D+/Oc/8+yzz7J8+XIOOuggFi5cyKRJk1i7di0Aq1atYuLEiYwdO5Z169YRDofxer1s3ryZMWPGMH78eN57773mYydMmIDL5cJsNrN9+3YMw2D16tVMnDiR8ePHs3r1anRdp7S0FF3XSU1N7epLEqLfazPswmFcc25pDrvgL35J4OZbW4edyZwoKJewG3h2rFcp95f17jh6UI+UJcyaNYu5c+eyePFiRowYweTJk9E0jUsuuYSLL74YwzCYMWMGVquVKVOmMGvWLKZMmYLZbGbRosRqsfnz5zNz5kzi8TiFhYUcfvjhAEycOJGf/exn6LrOnXfe2ROXI0S/0mbYBYO4b70J8ycfARC4aiqhX13eatWlFJQPcPWJn4q9RRyaflivDqWnKIZhGL09iO5WVSXP9cTgpNTXodZUt37N58V1842YN3wGgP+GmwhP+XmrYwy7I1FQrkr3wf4kI6Pzj3QUlwK3wH2FD3LF2Gu6cVQ9q6PvgRSeCzFAtRl29XW4p1+P6etNGIpC4NbbCJ/3EwBCsTD14TqCVg3VlUeaHsauyq4HA5YfbJqNIm9Rb4+kx8g/34QYgJSG+t3DrroK97VXJcJO0/DfeXersCv3lxJ1uzDnDkPHoNi7nWAs2BvDFz0kz51PkXd7bw+jx0jgCTHAKA31qNVVrV5Ty8rwXHMFpi3fY5hM+O59gMjpZzZ/vT5ch5oxBDVzCIqiYNEsWDUbNcHqXU8vBgiX28V3n3zLa+//jcPHHdjbw+kREnhCDCBKY8PuYVe0HffUy9FKijEsVnwPPkL0xJNbvEnBn+JCTWm9NN2smgddc+HBxOf1MfaYw7EPdVBWUtrbw+kR8gxPiAFCaWxAraps9Zq2+TvcN1yLWluD4XDgfehRYuMnNH/dUFT0rCGYdCtRPYpF27mfXVSPYjXZemz8oud5rEkEYwEw7/nYgUBmeEIMAG2G3aavcF97FWptDbrbTePjy1qHXVNBudNJmj2dcDxEJB7BMAwi8QjheIi0QVKQPFh5LEmJXyT17jh6igSeEP1cW2Fn+mw97uuvRm1sQE9Jwfub3xE/ZGetlWEyo+fmwY4G73aTnTx3AaqiEoj6URWVPHcBdpOs0hzIPBZP4hfJvTqMHiO3NIXoxxRvY6uwC8XCRD94h7w770ENh4lnZOB9/An0YcObjzEs1kRBuan1H/9E6OX32NhF7/NYd0ztknt1GD1GAk+IfkrxNqJWVjT/PhQLE3z7VYYveBg1GiMyJIuNC+ZAiorSsA2LyUpSUjaW3DwpKBcAuMwuVEVFT9J7eyg9Qv6vF6If2jXsAPQ3/8aIex5CjcYIF+Tz3eL7qUi20BCux2ayE7Pb2OaOEdTDvTRq0deoiorb7B40MzwJPCH6GcXn3S3sLH//G7n3PYgSjxMaMZyixQ9R6zZhNznQ0TE8HpScPKwmu9TWiVY81iRZtCKE6HsUnxe1orzVa9YXX8B1390ohkHgwAMoWvwg8ZRkIvEIYKCkpRPLSKy2lNo6sSuPJUlmeEKIvqWtmZ3tT8/gXPwgAOEjxvHNfXcQdtqaN12uS7Lhytq5YEVq68SuPBYPuNnxD6SBTQJPiP7A50uEXdPmJoaB/be/wfHb3wAQOeYH+B9ZQmbGSFRFJRQP4cgfQ1JGHqqiSm2daJfHmgQKlPiKe3so3U62BxKir/P50CrLW4Wd49FF2FauACBy4sn45i8AS6JLiqFqia197HaCsSA1wWrCsRBWk400e/o+1dZ11XlE99qb7YE0s4Ye02EY8CvgT8AWyM7N4bNPN3XPAHtAR98DCTwh+jK/H62ibGfYxeM4HrwP26t/BSB8+pn4b5/XXFNnaKZEjV0X7lAejAUp9m7Hqtkwq2aiepRwPCSF6X3QXu2HpyjMfG029eF6nvr8t0wedgaHpR/Ow2c9QGVlYzeOsnvt8354c+bMafdr999//76PSIhesOssxWl24o/6++6sZdewi0Vx3j0P61v/BCB03k8J3DK7uabOMFsSrcJMXVteWxOsxqrZmvtsNv1cE6yWQvUBwG1OBERjuP+GXGd1+AzvzDPP5Mwzz6ShoYERI0Zw/vnnc8ABBxCJDPyHm2JgaZql6IaOw+wkEA2wrvxjAtEADrMT3dD71v5vgUDrsItEcN0+qznsglN+TuDWOTvDzmpLtArr4rADCMdCmNXW3YVltefAoakabrObhkhDbw+l23X4p+P4448H4A9/+ANXXnklABMmTODXv/51949MiC606yzFH/XhNLvwx3x4rJ4en7V0+EwsEEArL90ZdqEg7tkzMa/9MPHey64keMXVoCgAGA5n4pndjt93NavJJjspDHAeaxKN4YEfeJ1apRkIBFizZg0+n4/333+faDTa3eMSokvtOksJx8LYTXbCsZ1dR3pq1hKMBdlc9y0l3mIq/ZWUeIvZXPdtYnYZDLYOO78P94xpzWEXuO5Gglde0xxuusvTrWEHyE4Kg4DH4qExMshvaTZZsGABK1as4Gc/+xkrV67kkUce6e5xCdGlmmYpO39vJRgLYjXtXNzRU7OWEm8xNcEaTKoJh8WBSTVRE6yhrOo71LKdYac0NOCZNhXz+k8B8M+cRegXlzafR09OxcjK6tawA9lJYTDwWJLwRb3oxsDuqdmpG/4jR45kxowZbN++nQMOOID0dPmXnehf0uzpFHu3A4mZnNPsoiZYTZo9HcMwWq087G4VgXKcZidmLTHjNGtm3HETjSVfomSOB0CprcF9w7WYNn+Hoar4b7uTyI/Obj6Hnp6BkZTc7WNtIjspDGweaxK6oeOLDOwV7Z0KvGeffZa33nqLhoYGfvzjH7Nt2zbuvPPO7h6bEF2maZZSE6wmEPXjMDuYMORI/FE/gagfq8m237OWTteqGQYGO6uB1HAYa2UVUT3xr2u1ohz3Ddeibd+GoWn45i8gesppiYMVBT0zC8PV+eXnQuxJ0754A/22Zqduab7++uv88Y9/xO1288tf/pLPPvusu8clRJdrmqWMTBlNnjufFFtqq9/vb9i1XAXa0arPLGc2gViQSDyCEgqhlpYSigZJt2egFhfhvuaKRNhZLPgWLmoOO0NRiQ/JkbATXS5px754jQN8pWanAq+pNl3Z8azAYrF0dLgQg07LVaCKomDRLFg1W5s7E+S680izpUAohFFchBGPk2z1kFsZxDP1CrTyMgy7He+ix4gel1gpbahaouzA4ejpSxODgHvHDK9hgK/U7NQtzR/96Ef8/Oc/p7S0lCuvvJJTTz21u8clRL8SjoVwmJ2tXjOrZgJR/27H2k12RjqG4q3YSNSagsVkJX1bJek3z0Ctr0d3ufAtXkLssLEAGCZzoqDcbN7tXEJ0BbNqxmFyDPhbmp1qLRaLxdi2bRvffPMNw4cPJycnB4/H0xPj6xLSWky0pSv7QxZ7i9ANvVWtWiQe2bGicZfFHuEwamkJih4HwPT5Blw3TUP1+dCTkvA+9hviBxwE7Cgoz84BTdu3ixSDyj710mxyJRCC7HcGbi/NDm9pVlVVsWXLFi6++GI0TePAAw/EbDZz2WWXdfkghehJe/PMrTM6Xau2a9it+xj3jdcmwi49ncYnntoZdnZHYmYnYSe6gR7Tmfna7OYf1MOI8SP7ddjtSYe3ND/77DP+9Kc/sWXLFu68804Mw0BVVQoLC3tqfEJ0i67uD7nrKtA2V31GIq3CzvzBalxzbkWJhIkPyca75An0vMRnGy43emb319gJ0awucacirsfR1IH5j6wOA+/UU0/l1FNP5b333uOoo47CbrdTUVFBVlZWT41PiG6xN8/cOqvDWrVdw+6df+OadztKLEY8vwDv48sSHVMAPSkFQ2pdRU+rhYgeodhXxFDPsN4eTbfo1CrNzz//nMceewxIdF353e9+162DEqK77dp5Bbqx00pT2MVjAFj+8TquuXNQYjFiI0bS+MTvd4ZdapqEnegdtYmftjR837vj6EadCrx33nmH2bNnA/D444/zzjvvdOughOhuPdYfcpews77yEq6770TRdWIHHYx32e8w0tJ3FpSnpHbt5wvRWTWJn75v2Ny74+hGnSpLUBSFSCSCxWIhGo0yCPaMFQNcp5657a9otFXY2Z5bjmPJo4kvHX4E3kWPgdOFoajoWUPA6WzzNLLbuOgR3sSfi4E8w+tU4F100UWcffbZjBkzhu+//54rrriiu8clRLfr1v6Q0ShqSTHhsJ/6UC0pf1pO6rMvJL501NF4H1gEdnuioDw7B2xt30ptudu4w+wkqkcp9m6X5s2iWwzzjGBL/SCf4V1wwQWccsopFBUVkZ+fT2qq3HYRA0eXz6BahF25r4T8Z54j48W/AFB/zEQC9z6AzW5PFJRn50AHnYtkt3HRk0Ykj+Sb2kFalrBs2TKuvfZabrrppua2Yk0WLVrUrQMToifsaQa112EYizXfxqwP1jD0N8+Q9to/AGg86QSKb52OSoAsS0Yi7PawQ3l3rCYVoj3Dk0bwr63/GLClCR3+aTv55JOBxC1NIQaijmZQTVsKdfp2YiyGWlKMEotCLMaQBxeR+u93Aag/YzIVM27ApKoETHqiL6a65zVjstu46Ekjk0YR1aMUebczLGl4bw+ny3UYeJs2bWLTpoE7vRWDR3sztY5mUHt1O7FpZheLQjSKa97tWP7zLgB1Pz6XymuvBlUlYjOjZGV1Kuxg9338enLfPjH4jEweBcD3Dd8NyMDrsJdm023Lzz77DJvNxrhx4/j888+JxWL9qhZPemkObi1vW+4aGjXB6nZ7YDaFYcvb+YZhEIj6GZkyeucHxOOJmV00AqEQrttuxbLmvwCUX/QTqi/7FSbVTNhlw5/i2usFJ7JKU3TWfvXSVAAHcAvwBvBR4uXs3P7VW7Oj70GnmkdffvnlPP30082/v+yyy3jmmWe6ZnQ9QAJvcOuosXPL25Z7E4Z57vxEEPkqoGgbVkMlJW4hfc5szP9bB0Dgmuuo//nPqQ/XEXDZ0DKGSFiJbrU3gacoSqKHZguGYbBk/aMcknYIpxT8EICHz3qAysr+s4tCR9+DTq3SrK2tpbGxEY/HQ11dHfX19V01NiG6XUe3LTuqx+vodmIwFqS4fgvuqgbMhgm9sR73bfMwb/oGAP+MmYQvnIINyMwei+FJ6unLFmKvKYpCijWFulBdbw+lW3Qq8K655hp++tOf4nK58Pl83Hfffd09LiG6zJ4WfrRXj9dRGBbXb8Vd1YAlZqA1NJA36w5s323GUBT8c+YSOftcUBTimUPA5eqxaxVif6XaUinxlfT2MLpFpwJv8uTJTJ48mZqaGjweD2bZiFL0I/uz8KPNMIzHMYq3Y9Y1TNU15N16G9bt2zFUlW23Tsdz9rmJ7inZOWCX25eif0mxpfJV7ZfE9BgmtVMR0W906mo+/vhj5s+fTzwe5/TTTycnJ4cLLrigu8cmRJfo0jZi8ThqaQk2XUEpKyH/1rlYysrQzWaK7riVwHE/wK2ZEmFntXb9xQjRzTyWxO13b8RLii2ll0fTtToVeI8++ijPPvss06ZN45prrmHKlCkdBl48HueOO+5gy5YtaJrG/fffj2EYzJ49G0VRGD16NPPmzUNVVVauXMnzzz+PyWRi6tSpnHTSSYRCIW655RZqampwOp0sXLiQ1NRU1q9fz4IFC9A0jcLCQq6//noAli5dyrvvvovJZOK2225j7NixXfPdEf1KR6sZ97aNWJvnUiyJ0oNImLSKRlJmzMJSXYNus7LtrtupO+JgslwZiRq7FndB9neVpazSFD3JbUks+vBFB17gdaoYSFVVkpOTURQFq9WKs50mt03+85//APD8889zww03cP/993P//fczffp0nnvuOQzD4O2336aqqorly5fz/PPP8/TTT7N48WIikQgrVqxgzJgxPPfcc5x33nksW7YMgHnz5rFo0SJWrFjBZ599xsaNG9m4cSMfffQRL774IosXL2b+/Pn7+S0R/VFX7mDe5rnqtxAp2owSCaN99y0Z11+HpbqGuMPO5nvvpP6IQzEsNoo8UBwqb/7c/R1XV+/MLsSeuMyJZ86+yMBb3d6pGV5BQQGLFi2ivr6e3/3ud+Tk5HR4/KmnnsqJJ54IQGlpKenp6bz77rscddRRAEyaNIn//ve/qKrKuHHjsFgsWCwWCgoK2LRpE+vWrWtuUD1p0iSWLVuGz+cjEolQUJB47lJYWMiaNWuwWCwUFhaiKAo5OTnE43Fqa2ul3+cAsDczm67sObnbudCwVDXQEI1h2VyG5+bpqD4fEZeT7++bh3XsROL4UXLycZisrTqy7O+4pJem6GmuHTM8b9TXyyPpep2a4c2bN4+cnBwmTJiA3W7nnnvu2eN7TCYTs2bN4p577mHy5MkYhtFcwOt0OvF6vfh8PtzunTUTTqcTn8/X6vWWx7parHbb0+uif9vbmU04FsKs7ryNGIqFqA5W8U3t1xR7i/ZqRtTqXLqOubwCc1SHdR+RMn0aJp+PSLKHz+6bzdY8N58FviOclYXFbENRFCyaBatmaw7rluOCxMKZcCy092PZh/cLsbesmhWzahm8M7xrrrlmnwrNFy5cyMyZM7nwwgsJh8PNr/v9fjweDy6XC7/f3+p1t9vd6vWOjm1aMdrWOUT/trczm5alB6FYiDJ/CRgKydbk5rDs7EKV5nMpJsxl5SjhMNaPPiL/rvvRIhFC6al8ed/tRHOHYHZY2Gb2Yo/68Nh21to11fntby9M6aUpeoPDZCc0AG+bdyrw3G43//73vxk+fDjqjh6Aw4e332ftr3/9KxUVFVx99dXY7XYUReHQQw9l7dq1HH300axatYpjjjmGsWPH8uijjxIOh4lEImzevJkxY8Ywfvx43nvvPcaOHcuqVauYMGECLpcLs9nM9u3byc/PZ/Xq1Vx//fVomsZDDz3E5ZdfTnl5Obquy+3MAWBvdwloWXpQG6oBQ8HAINWette3AdPs6RQ3bMVc1QiRGLbV75O/4EHUaIzAkAy+uu92wlkZRJOTCDktNNQU8W3tJhRVIcWaim1HSDXdht2fXpjSS1N0F9Wk8vBZD7T9xauhobGBL2/YSHZux4+w+pM9thbz+Xz88pe/xOFw7HyTovB///d/7b4nEAgwZ84cqquricViXHnllYwcOZK5c+cSjUYZMWIE9957L5qmsXLlSl544QUMw+Dqq69m8uTJBINBZs2aRVVVFWazmUWLFpGRkcH69eu57777iMfjFBYWMmPGDACWLFnCqlWr0HWdOXPmMHHixFbjkdZi/U9H7cDaC62mZ37f1H5NsjWZVHsath0zoTZ7YLbHMIgUfU9DfSmOt95i6EOPo+g6oaEFrLvrRuLp6eiZmQRsJop9RcTjOiZNI8sxBAOddHsmisK+bzHUznXJKk2xJ/vbWqzJC18/R9EX26lc2H9aijXZ516azz77LM888wyapjF37lwmTZrULQPsbhJ4/U9HDZ/39Jf9voRlM8NALStFCQaw/u0VHAvvQzEMYmMOoHrRI2zVGilzK9g8aVQGKvBH/aQ7Msh25RCMBvCFvdgsDg5NP0xCSfS4rgq8v333Ct9+/Q2V8wdW4HV4S/O1117jzTffxOfzceutt/bbwBP9z/4Ui7d3GzDdnkGxt6j9mVLLsHv+OZyPJXYLiR52OL5Fj2FOSiYn4xCMaDUVgXJ8ER/5ngKyXTnYTDaSrckYTqO5R6cQ/ZVVs8IAfEzcYeA1lQukpqYSjUZ7akxCAHtfLN76fa3DMt2eQXWwqs3NXAFqAlXopUXYw3HyVr6K8/eJ7a+iE47E++BiDE8SenYOdouFUaQwKmU0xZ7dZ5KyoEQMBIMy8FrqxC5CQvSK9p5xtQzLYm9Rm6s+S7zFGEYcV3Uj9rBB2tN/xPP8SwBEflCIb8FCDLcHPScXTK3/uMiCEjFQWU02sDLg+ml2eCXfffcdN998M4ZhNP+6SdPmsEL0ppbP+nadubW8rdjeqs+ihq2MCNqxBiNk/ua3pPz1VQAaJhUSv/fhRNgNyQZN2+2zu7RHpxB9iE1LTO8awg2k2dN6eTRdp8PAe/TRR5t/fdFFF3X3WITotKZZ3bbGrZhVM5mOrOaib9i9BGHXerZQLESlv5z6bV/SoGZyyLLlpPzrHQAafngqW268mvyk5ETYtdjxfFf7ettViL7MqiUanzdE6gdP4DW1AhOiL2k5q1MNFQWFMn8J2c5cbCZbm/V6LW8/xvU42xu3YauqIT/uYcyi35D6308AqDvnLEquuxItKWmPYSfEQNVUztMYbujlkXStTrUWE6IvadmFxWa2oSoqZtVCXbgWaHvhSNPtR1VRKfUWkVTrZ2jczZGLniFrR9gV/fRMSq6/ioDHTlLBQRJ2YtCy7rilWR+u792BdDEJPNHvtOwvmWJLJRwPgwGhaIhIPEI4HiLNnr7b+5puP+YHzBSE7Yy4cwGeDz8GYOvPf8KGi85AT09nyNDD5TmcGNSanndXBip6eSRda+AsvxGDRsvncTaTjRxXLhX+cnQjvqO4vP2FI2pFOc6aRvJun4dz41cAVF5zJb7zzyUvPYPsnIN68lKE6JM8Fg8ARTseAwwUEnii30mzp7O57hsCsSBxPY6majjNDkamjOlwZqZUVKAWb2fErDuxfL0JQ1Eov/F6as46HV+am5zMYT13EUL0sg57aQLqrSrF3qIeHFH3k8ATfU7nekcqidpQo6lGtOPnbUpFBdrWzbhvvA7T95sxNI3SW6ZTecrxKDnZ5CTnym1MMajoMb3N1mIPn/UAlZWNnPHyyWxr3NrzA+tGEniiT+lMXV1NsBqPNYl0R0bz+yLxSLu7ISgVFZi++xr3tKloxUUYJhO+e+7Hdupk8rJzwGLZ7T1NY5GmzWKwOiDlIP659Y1We5n2d7JoRfQpLVdg7rqZapO92RRVqazE9NUXuK+5PBF2Fiu+Bx8h8sMz0HPzOgy7vdmAVoiBZmzmEdSEaij1lfT2ULqMBJ7oU9rcuTzQeufypkUrLbVViqBUVmJevw7P1CvQKiowHA68jywhctIpibAztX+DozPBK8RANi5jPAAflX/YyyPpOhJ4ok9pGWahWIitDd9T7CvGF/FS6ithc903OM1OwvFECYJhGG2WIiiVlZg//hD3dVeh1tSgu900Pr6M6PEnJPpiqh3/r783s0ghBqKxGUeQZE3mvaL/9PZQuowEnuhT0uzpzWFW6i2lMlhF3IgxxJWNqqjUhOqoCdY0F5EHov7dShGUykosq9/Dff3VqA0N6CkpeH/zO6LHFHa6e0pnZ5FCDFSaqnF87gn8p+htdEPv7eF0CQk80ae07IhS4tuG0+Qkz52PzZS4vegw2anwlzUXkY9MGU2eO7912L3zFu7p16H6/egZmTQu+z3RI4/ByMzs9DhaBm97s0ghBrrTh59Jmb+Uj8s/6u2hdAkJPNHnNIVZjjufLEdWcyNbAAWl3RmaUlmJ9R9/xz3zRpRQiHhOLo2/fYrYhCMx0vauAW7L4G1rFinEYHDm8LOwaTZe+XZlbw+lS0jgiT4ryzEEf9RPNB7FMAyi8Sj+qJ8sx5DdjlUqK7H95UVcs2eiRCLEhw6j8bdPEztiPEZyym7HJ1ZhFrG57tvmxTC7am8WKcRg4bK4OWP4j3jl25fw79KQvT+SwBN9QlsBlOvOI82eRkyPEYgEiOkx0uxp5LrzWr1XqajAtmI5znm3o8TjxEaPofGJp4gddjiG29POZ0nJgRCd8evDrqIhXM/L3/T/WZ4Unote0bKo2wAawnXEDb25VVh9qJaRKWMYmTK6w+JvpaIC+x+fwrloIQCxQw6l8ZHfED/gQLC3PSNrWXIAtLuHnhACjh5yDIelH87vNizjFwf/ElXpv/MkxUj0ZRrQqqq8vT0E0ULLbipm1cy3td9Q5CtmWNJQXGYXUT2KPxog15XLqJTR7Z5HqajAsewxHE8sBSA6bgKNi5egjxoNVmu779tc9y0Os7NV9wjDMAhE/Yzs4POE6OsyMtydPlYza+ixNlZfKsCuqXAocD7wImTX5/DZp5v2Y5Tdq6PvgczwRI/bdYbVEKnDY3ERjPpxW9xYNAuGYVDhL2s38JTycpyLF2L/49MARI75Ad6HHkEfMQrM5t2ObzmjrN0xm/RYd97ulJIDMdi010uzzWMNnT9tfAYugbI5pd08su7Tf+emot/avahbwayYCcejLV7pYDVmeTmuBXftDLsTT8b7yFL0UWPaDbuWz+ySLElsb9xKY7hRSg6E6ARVUflBTiE1oWoY19uj2XcSeKLH7VrUnW7PpDHaiKqorVZjJluSd1vIopSW4rrjVmwvPAdA7ckn8M1ds/HnZ4Omtfl5u7YJ81g9FHiG0RBpkJIDITppTMoB5Lry4BRoDDf09nD2iQSe6JTOLOPvrF2LutPsabjMLpxmZ/NqTKfFhYHeaiVlxTcf47jlBmyv/hWA2h+dTvm98wnnDqHYX9zumNpqE+a2uEm1pkjJgRCdpCgKJ+efCg5Y9MmDvT2cfSLP8MQedWbLnvbe19YKy6ai7ppgNYGoH4fZwbG5x+GP+ptXbZYHymkI1VHiK8Wm2cioCTL2rsU41iQ6PlT/9FwqZ91KPCMdC0BcaXeVZcsd0pvIMzsh9l6Wcwj8D36vPcHPD7qUMakH9PaQ9ooEntijfVnG3xSShgG+qBe/v5QtDd9zaPphpNhSm4u6W0qxpTa/LxQNEoj60RQNfXsRhy54hpT1XwJQPuV8am6+Cb1F9xSzaibQTmGs0+zki+rPQTdwWly4LG4UBfLcBfv9vRFi0HkH3D9wM+Pd63n1vDfR1LYfJfRFcktT7NG+7BxQE6zGMKAmVIVu6HgsHkyKiS+qP+/wdmhTuMaMGCoqGWV1nHznE6TtCLvvf3khlbfeTCi59dLjphnbrrde60K1VAeryHIMwWl14Y/5qQiUk27PkNuYQuwLP9xT+AAfl6/lmS9+19uj2SsSeGKP9mXngHAshC/qxaxamheLOMwO0I3mPeXaei7YFK521UrS1mJ+cOuDpH+1GYB1l53L2ktPx++w0Bhu2K2xs9Ps3K2DyhfVn2MY4LF6yHHlMjplDEM9wwZEmyQhessFYy7i5IJTWfDhfLY1bu3t4XSaBJ7Yo33ZOcBqsuGP+lvNDJsWo4RjoXbbexlANB4hZ2s1p922lJTNReiKwrtXnMm2X08hNXPYjluqiXG0XGXpj/p327QVwBdp3XhA9rUTYv8oisLDJzyGoqjc/O6N9Jf+JRJ4Yo/2ZeeApjAMRAPNpQbheBizaqY2XMf/Kj6hLlSHbugoioJu6NSF6tjWsJWGtf9m3PQ7cW8rQdc01tz4c2p+/StMbg+ptjQsmgWPNRmbydZqlWVbt16dZif+iK/Va7JgRYj9l+fO585j72ZV8X947qvlvT2cTpFFK6JT2lpksqfjD00/jC+qP6cx1IDT4sJtdlPk3Y7L7KYiUIHT5KAuWIfNYqPKX4lDs5P3VRHH3fYI1opKdLOZL26fxtZTxpHuSSfVloZtR1C1tUjFarLhjXjxx3yEY2GsJitmxQyqQiQewayaiepRwvGQLFgRogv88pDL+Ot3LzPvg9s5ZehpDHFm9/aQOiS9NEW3almaUOYvoyFUh6qaqApUEI5HCUT9eMwusl25pK37ghPnLsFR10jcZqPo/vk4f3YlxcEydENvVVYQiUdQFZU0e3rz+UOxENsat5BiS8NushOMBfFHfRySfhhxI95c8gCJdoFtNaMWoj/rkl6ae7Jrr81UYCqwGXi+9aHZuT3fd1N6aYpeYzfZm0Npa8MWQrEgQ5xDyLBn8XXtRkLxCMGIl/Ff1nD8HY9j9fqJOR0UP/QA5Scew0iTiTR7OsXe7QCtZmnp9oxW9YE1oRrMqhVQCMVC2HYEWtyIk+fO361pdWfrCYUYiPaml+aefFy+lvfM/+Gs/zuXA1MPan794bMe6JLzdxUJPLFX2ism7+j4ppAJRPz4oj4awg1kOYfgMLtQCDDkg4+ZtORdLP4gUY+LdQtmYT71BKxa4n/PXQvVrSZb8+9b1gfqhk6KLQVN1ch25QA7d0EA2RZIiO4yIetIvq7dxNvb32KoZ1if/QekLFoRnbYvG6c2hYxu6ESNKGE9TCgWZHPdd5T5Sjhg1edc9ci/sfiDBFM8fPjwHVQXHknYiLRaBdrW7uMtF6mEYiEawg0UeYso9hYR2rEKs+UClX2pJxRC7JmqqEwedgahWJBVxf/p7eG0SwKvn+vKHpd7smsTZotmwarZmuvq2tIUMnXhWtIdmehGHAPQdZ2jV33Hzxb9DUs4SigjlX8tnMHHo2y4LO5O3WZsqg8MxUKU+kpwW9wY6BhAibeYxnBjq/KJfaknFEJ0ToYjk4lDjuLz6g0U7XgE0ddI4PVj+zLj2h/7MkMygCLvdjbXbSYUC+E0uYjGIkz458dMWfImlpiONyuND5fei/3YEzk65ziSbMmduiXSVB9Y4S/HsqPAPdWeSqYjEx2dhkhDq+Dcl3pCIUTnHZt9HB5LEm9t+ycxPdbbw9mNBF4/ti8zrv2xtzOkYCxIKBYkGA3iMDnwR33Uh2o57rVP+PVTazHFDSpzUnj54evQDp9AKBZmU82XrCl5n+/qvt1jcDc924vpUaJ6FE3VGOYZwbCk4YxKHk2qNaVVcO5LPaEQovMsmoVTC35IbaiGj8vX9vZwdiOLVvqxcCyEw+xs9VpHTZT3V1urJRvDDdhMdjbXfbvbkv9QLITHmoTL4qbcX8YXles55eWP+clziT8IFUMzeeGeXxPI8RCs/AxNNWFWzTjMbkp8JYRiAUamjOkwkOwmO0OThu9WttBeEO9tPaEQYu+MSB7JmJQD+bDsg0TJQh8iM7x+rKefSe06Q4rEI4CBRbOgKhol3mJKfCWoioZu6Gxr3II/6qfMV0aVv5wzVqxpDruS0TmsuP8KajNdRGNR6sMNWFQLKJDpzMRpdhDYsSJ0T88p5ValEH3LyfmnoKkmOIs+1XZMAq8f2/Uv+sZwI9sat9IQqu+2BSwtV0vaTDY81mQsmoX6cB0uiwuzovFN3deUekuoDzbwSekHlPtKOGbpSk5+/n0Avjkgi5cWXo2Sk0OSNRnd0AlFQzRGGojGYzSE69F1nbgepyHcsMfnlHKrUoi+xWVxc3zuCTACXvrmhd4eTjO5pdmPtaxPqw3VUhuqIcsxBLfFvd9F1W3V2wGtXmsI1ZNqTyMUC7G5bjN14TrqQjUkW5I4YsgEYkaUzyo+Y9qzX3LwGx8DsGlsPn+c/WOGJKeRbkunIdKA2+JG0zTc5iTsZjsxPUqJv5g0WwaBWAC3JWuPtXNyq1KIvuXwjCN4+4N/Me+D2zh16A9JsfX+/U2Z4fVzTX/RJ1mTGOoZhsfq2e8FLG2t/vyy+nM+KH6fLXWbaQg3EIgGqA3XUh2oYmvDFhoi9QQifiyKFX/Mx/f1m6luKOX6Jz9i4o6w+/KoUbz30K0kpedTH6mn1F+KaiigqngsSUT1CLF4FMOAWDxGRA/j0OxSOydEP6QqKvwd6kJ13LNmXm8PB+iGXprRaJTbbruNkpISIpEIU6dOZdSoUcyePRtFURg9ejTz5s1DVVVWrlzJ888/j8lkYurUqZx00kmEQiFuueUWampqcDqdLFy4kNTUVNavX8+CBQvQNI3CwkKuv/56AJYuXcq7776LyWTitttuY+zYsbuNaTD00txc9y0OsxNFUZpfa+oyMjJl9F6dq9hb1GoRSCgW4pPyTzCrJoYlDSOmxwjHw7gsbr6r+xaTakIxDL6s/RKrZiHVlkajt5rzF77IYR98DcDWU47h/26YjF+LkmJPJcnsZnjyKHRDp8hbRLI1GVXV8IYbwIB0RwYeaxJJ1qR2+2jKjE6I1nqkl+beUIBTgeOAZyA73v29NXu0l+arr75KcnIyDz30EHV1dfz4xz/mwAMPZPr06Rx99NHceeedvP322xxxxBEsX76cl19+mXA4zMUXX8xxxx3HihUrGDNmDNOmTeP1119n2bJl3HHHHcybN48lS5aQn5/PVVddxcaNGwH46KOPePHFFykrK2PatGm8/PLLXX1J/ULTApbOrFTck5arP0OxEF/XbaI6UI5dcxJx52DVrADE4lEcJjuReARF0ch2ZmNgoIXjXHT3Cg5al9i4df0Px/Ovm36MWVEJBIqxqjacpsS+eIqqkOPOwdANbCYbB6YeCLRuDt1WH03Z7UCI/dOVvTQ7EolH+OPGpzBda6bs9tJu/7yOdPktzdNPP50bb7yx+feaprFx40aOOuooACZNmsQHH3zAhg0bGDduHBaLBbfbTUFBAZs2bWLdunUcf/zxzceuWbMGn89HJBKhoKAARVEoLCxkzZo1rFu3jsLCQhRFIScnh3g8Tm1tbVdfUr/QlSsVW3YwKfOXEIwGcZmTUBSo8JcTjocxqSb8ER9pjgySrSmoiorb4sEUDHPB7b9vDrtvLzyTNbdfDZqGqikckn4Y2a4crJoFb8xHtjOXIY5sDHR8Ye9uY5cFKUL0b4navMnUhmrgB707li6f4TmdiZmBz+fjhhtuYPr06SxcuLD5VpvT6cTr9eLz+XC73a3e5/P5Wr3e8liXy9Xq2KKiIqxWK8nJya1e93q9pKb2/sPRntZeg+V9CQan2ckX1Z9TFajCpllRFDBrJsyag2g8wnc132BgEIgHKXAVsKF6A75oIxkhK5ff/QK5XxUBsPGX51E/+zZygzWYNBMxI0q2M5e6UC1xPU7MiDXvb5duz6Qh0kBtsIZAPFGoXhOsbhF6cvtSiP6qqTbvmxM28X39d4xIHtUr4+iWRStlZWVceumlnHvuuZx99tmo6s6P8fv9eDweXC4Xfr+/1etut7vV6x0d29E5BquWJQNNW/LsbY/NYCxIdbCKLMcQTIpGJBYmEouQYkvBYXawtX4LX9ZspMi7nVRbKp9Xb8AbakCpqeXXt/2B/B1ht/rys6i69RZUVUNXdAwMsp252Ew2Umyp+HfM1ppmdIoCo5JHYTFZyHJkkWpL7fZWaUKInnNy/ikQg1vem9FrtXldHnjV1dVcdtll3HLLLZx//vkAHHzwwaxdmyg4XrVqFRMnTmTs2LGsW7eOcDiM1+tl8+bNjBkzhvHjx/Pee+81HzthwgRcLhdms5nt27djGAarV69m4sSJjB8/ntWrV6PrOqWlpei6Pihnd7vanx6bTe3KPFYP+UkF5CUVMCJ5JFbNSqmvBJvZyRDnEMakHkh1sIZybympjRHm3v0vhm6pQVfg1StP5b8XnUB1qIZQLMDBaYeQYktpDrjEs7k0Um1prW5T+qP+Hm2VJoToOS6LG96C90veY8WmZ3tlDF1+S/O3v/0tjY2NLFu2jGXLlgFw++23c++997J48WJGjBjB5MmT0TSNSy65hIsvvhjDMJgxYwZWq5UpU6Ywa9YspkyZgtlsZtGiRQDMnz+fmTNnEo/HKSws5PDDDwdg4sSJ/OxnP0PXde68886uvpxusbd7yu2tllvylAfKCMfCqIqKgsqoPazYbLlgJcWaSpm/BJNipiHcgMfqwWq3EtPjeKwetjVuJbnGy433vkVGeQNxVeHFqafx31MOZphmae6W4o/6d7vdOjJl9G7X3NOt0oQQPex/cOzVxzHvg9s5peA0spxDevTju7wsoS/qS2UJbe263bTqsKtCb3Pdt6iKRnmgFLNqwayaicQjNIYb+EHe8R1+TlslCZWBCrY1fI/VZEc3DBrDDehGjMav1nPDvJfJqAkQM6k8fe2JfH3yOPwxP8fkHE++J49ANECmI7NTpRG7fjZICYIQe2NvyhIURemRVZotPXzWA6z5Zh0nvvADTht6Os+cvrzLP6Oj74EUnvewntjhwGqyURWsxLxjyxxFUVAVFZfFvcfPabnaMxgNUhmowBtpJN2WgYaGgY7D5MD46ktuvuMlMmoCRMwaC64dz9tHplMXaSDHlU+WM7N5B4POlkZIT0whBr6RyaO55cg5vPb933j9+7/36GdL4PWwnth1O82ejjfSCEai+DwajxKOh8l0ZO3xc5pWe0biEb6v/w7DMBiRNApNNVMfbiDZmkLypu+47rYVJNcHCVo1ls3+EVUnHovd7MCsWhiePAIAfzSAw2TvdGBJCYIQg8PUw6dxaPpYbnlvOlWBqh77XOml2cO6skAc2n8eONQznJpgNbFoDKvJSo4rF1VRMSvmPZ7TbrJjM9kYnXpAqx6WwzxD0T5ey2k3P4rNHybotPLc/F+gHDmO7EiAMWYXmY4MGqM+fOFGzGYbiqK1Ki/ozGfL7UshBjazZuY3p/yOH750Aje/O40/nbGiVZeo7iLP8HpYVz7D6+hcwH59zq6tyrbUb0F9721Om/0o1lAUn8fOA7NPpnJ0PgemHYrLYsdhcTE2/XBqgzVYTJZufU4phNhdn2sttisFaJk4xwCnQ9KqJL59vqhLPqJHW4uJju1rgXhbM7mWzwNh950EOvqcYCxIibeYCn8ZKApZjiHkuvOav2412fBGvPijPhrDDShvvsbZ8/+AJRqnLsnGHbN+QG1BEm6gKlBGIOZmhDmxS0MgHsRt9exxhwMhRO/pqdZiHTEMg5e+eYFtx2ztkYJ0eYbXC1oWiOe58zsVdm3V1TWEGzp8Htje5wRjQTbXfUOJrwSLZsWqWinxFrO57tvmWj2n2cn2xq2EYiEcr73GufOewRKNU5XuZPbtx/Ndlg2r2Y4/FmB7YxGaotEY8RKOh3CYHLLDgRBijxRF4fThP4I4XPf2VcT0WLd+ngReP9DWyk7DgK2NW/mu9hvKfKWEdoRJZ54H1gSrCcSCOM0OrCYrFpMFl8WFP+pvXsXpj/op8Awj/+//ZvLdT2OK65QNcbPw7nPwF+SQZE2i0l+Oaig0RGr5vu5bqgOVpNszSLIm7bYTuzfipTZct9edX4QQA5vb4obXYF3FJzyy7qFu/SwJvH5g15WdoViIqkAlVs2CpmmEYiFKvMU0hhtpDDfs2JC1/WAJx0LE9Xirc8aNOBW+cr6p/ZpibxENoXqGvvgaR8xbjKYbVA7NZNn9l1KUBA3hOrbWf8f3jVvwxXyM8BzAsOQRRI0Ypd6SNndi3964lSRL0l53fhFCDAIb4YIxF7H4kwdZV/Fxt32MBF4/0LSys0ldqLa5PVeOKw+b2YaOTmWgAjCwaJYOg8VqsqGpWvM5w/Ewpb4SDAWSrckEogGsyx4lb+58FMOgbHQuv3ngF/jS7TSE6glHI1QFq3CYbKgGZDhTsZsc5Dhz+KLmC2qC1URiESoCFdSGammINFDQRZvTCiEGpvuPf4gcVy7X/vtKfFFft3yGBF4/sOuMyRf2YqCTYk3FZrKR7cxhVPJoFMBjTd5jUXuaPR2HyY4/GiAcC1PlryISj5BkTcJhcZL2yCMc+5sXASg7dCRP3X8ppWY/myo3YaATN2JoJjMecxJhI0a5v5wMRwYmxUSFrxzd0Em1p5HlyMKimXFo9sRtixbkmZ4QoiWPNYmlpzzJ1oYtzPvv7d3yGRJ4/cCuBdk2i4N0e2bz1jqQeHaHonS4WCSx+KWIUm8xNpMDt9lFibeI7+u/w6raSLEmk3r33Rz45HMAbBs/hqcW/Jxqc4QkczL+aAO+mA8NlUPTDyXZkUyqNZmaYC1RPUqZv4xcT95ugRuIB3d7prc/tYdCiIHp2JzjuG7cjSz/8g/8c+s/uvz8Enj9RMsVl4emH4aisFsLrizHkHaDpa2VnrWhGg5OP4zxWUcSiQVJmjWTQ557HYBvjjmIJ+f+mJyMURyYciA14RoynTlk23PJ9eSS48zBrFjwRXwoqFQHqjBpGgenHdLq882qGYfJIS3DhBCdMuuo2zkk7TBm/Oc6KgOVXXpuCbx+qL0WXLnuvHaDZdeVnv6YD6fZRU2whkjUz6R7nuTINz4BYH3hQayY+zOiJoVt9d9T4S/DbXaSac/AH/PhjfixqBZSrKlomsqI5OGk2FI4JvsHOHfZ7SCqR7FqVhRUirzb+b7+OyLxiBShCyHaZNWsLDv193gjXm57/5YuPbcUnvdT7bXgallsbgAKKt/Xb+bb2k04zE4MwKbZaIw0kGnPospbzJl3/YGsdz4F4D8njuThn48guWEzybYU6kP1RPQIoWiIdGcaue486kP1lPpKURSN0SkHMiH7SLKdOeS68yj2bgdo7rDSGK4HFDzWJEYkjWzuuiKEEO05KO1gpk+YycKPFvDO9l9wcsFpXXJeCbx+qr0emk1B2HQL0zCgPlRHVI/xZfWXuM1O4oqBrhsUV3/DxfOfJ2vddwC8dcbBPPerY3BGQ1QFq9AxUAwDtzUZk2KiNljHiKQReKxJ1ARrcGh2Ds8ch67rhHasBN21u4vN5MCiWaTrihBir1w/bjovf7OSW1fdzH+nfIxVs+73OSXw+rCWodbUfq6pFV0oFsBjTcZhdhLVoxR7t7e6Tdh0C7MmWI3NZEM34pT5y6jTErcig95qpt79V0Z9VQrAqz8+nBXnH0qKyUkoHsGiWjCpKgdnHMb2uu+JG1FsmoWoHkNBwWP24LF48Ed9pNszsWjWFi3NdgbZ5rpv21xII5u6CtG7VJPKw2c90NvDSFAgM9Oz++sjgUsg/9wM+KjjU2Tn5vDZp5s6PEYCr49q2RhaVTS2e7eDYVDgGUploIJQPIzLsrOuDVrPmpp2D2/a7XxL/feYVI2KQAU15Zu5++E1jN5SD8DzFx7BmiknkqRHsJttNEYVnCY3DpMTKzY8thRSFTMmVcNqsmMYMQKREMOSh1HgGUpMj1EVqCTZlrLbrK2rd4cQQnSNvtBLc08Mw+CFr5+j7txarrj9Gsxa+7u9dCa8ZdFKH9VykUl9pA6n2YHL4qI+XIdhGDhMdurCtc3H71rX1hQ0VpOVUl8JddF6AhE/eSErC+//LwdsqUdX4KUrTmLtxacRjIUo9ZVTE6wlx5WH3WIl3Z5JhjOdVGsaVrOFHFcuw5IKsJmdWCxWTGpiAYxZMycW0MQCu12HbOoqhNhXiqJwXO7x+KN+vqrduN/nkxleH9U0Q0v8Oryz8XM0iNVkJa7HCcfCQKLVWGWgonnm5DQ7CcVCbGvYglm1UOQtAh1MFRXcuvC/5Jb7iKsKv/v1RNZNHoPPV0yJr4QcVy52zUZVoBoVFZfFRV24DrfFwzD7MLwxHxbNhtPkwKnZienRxAazejSxE7q2+6rLfd0dQgghAPJc+aTZ0tlY8wVjM47Yr3NJ4PVRLW8FWk1WonoUBQWryUqKLZVtDVuxm+0Eo8FWtztrQ7W8X7yKDEcGTouTaDxGIOLFU1bD7fetJqPKT0xTuOvSEbx/lJv8aABF0UhzZKApJupCtaCo5LpySbOnMSblAOoj9ZhVMyNSRmEz2fBYPTSEGgjHwwRjiQBOt2TiMDvavBbZ1FUIsa8UReHgtEN5v+Rd6sP1JFuT9/lcEnhdpL1Vk/sqzZ7evMQ/2ZLSKtSa+mjaTHZKfcXYNCtZziEAbGnYjE2zYRhNzaENxtY7OGf+v0iu9RMxqSy+/hhWH2IjFAlSFagk1ZZOrjsPt9mFN9zI0OThpNsySLYmYTFZONh9CNXBRPux2lANtaFaqgJVHJJ+KBn2jOZSA7lNKYToDmNSxvB+ybtsb9xK8n7M8iTwukDLBSbtrZrcWy1vBUaNCLmuXAB0I45ZMTMyZTR2kx0FUBWNulAtxd7t1ASrGWLPwRfxoygqli++5KfTH8bZ4Cdk1Xh4+vF8fHASsWAtmqKiGwaVgXJQdLAbOC1uFMCkmTCbLNSF6qgMVJJsSaYsUIZVsZBqTSXLnkV9uB5N0UiyJsltSiFEt0m2pmDVrFQEKvbrPBJ4XWBPO4/vq87cCjSA7Y3bcFlcKGgoqHxavQ49HufQrX4umbscuy+E32bigZsL2XxwLqFgDZF4mGA0hDfaiFWzY9LMVAaryHbmEDdiKIAv4iPblY2Ghi/qw6xayPcMbe7hmWRL3tHlRW5XCiG6j6IoeCxJeCON+3UeWaXZBXbdrw56eDcARcHYUakXjAeoD9RR8Om3XDrnT9h9IRqdZmbNnMhXB2bgj/nxxhpRVRWbyY7FZMWmWYnGYtSHGvi27hti8SguSxJ2k43qYBWKqnRqZagQQnQXk2oibsT37xxdNJZBbV9rzfb2uV8wFqTEW0xFoBwMgyxnNuFYiAJ3AfWROrzRBryhBkZ/9DXTlqzBGtWp81h4YPZp+EdksbXmcyKxMDE9hlWzEjXiuM1OQvEw1aEa0qxppDpSiOgRKgJl5LsLMAxjxzW2Xhna2WsUQoiuEI6HdttmbG/JDK8L7EutWVu7F3S0C3gwFmRz3beUeIuxqlYsmpUSXwllgTKiepQUayqhWISx73/NrY8nwq4yxcqsW4/hyxwTxd5iVDQsqg3DgMZII4GID0NVcZtduEwOTJqGU3MRx6Ax1EhlsJJcVx4AKbZU/DsaVUs9nRCiJ8X0GPXhelKsqft1HpnhdYF9qTVr77lfibcYm8nWatYHsK78Y76q+RKTopLlHEKGMwun2UEgqrDdu42GUANpr/yNnz/xHppuUJpu48YbD6EkTSfoKyEY85NhyyCmxnEoTpLUJBpCDVT7K7CarKRa0rGabRiKQZYji2FJwyj3lRI34lhN1lYrQ6WeTgjRk4q9ReiGTo4rZ7/OI4HXRfa21qxlYXmTuB5nW+MW8t1D8UW8+L0lbKr9CofJzvbGbWiKhtVko8JfSSgeJs+dj27EicaiFKz4C6c/8S6qAVuyrFx93QjqUjVqfKXYzTY8ZjehWAizZiLVlkI4FsZpdqAbVgzAbDLhsSZh1izYTTYsigW3LZmYEcOjelAVtXllqBCi/+tTvTT35CfAAfCXX70MsbYPyc7dcxhK4PWStp77VQYqMKtWqoNVWDUrHlsS2xq28n39ZpItyQSiAXwRL5pmhhBoqNSH6zlm5XsUPvkaAFvyPVwzdTg1bo1Q1I/H6sFtSUJVVQJRH1mOIThMHioCpaBo5DizUVUFh8WBXbNjNdmwaonFKfmeoUwYcqSEnBADUH/opQlQH67nmS9+xxEZ4zj5r7tvE/TwWQ9QWdm51ZsSeL2kZWF5095xvogXh9mJRbM0N0lVFY1ILEQ9DVhMFhpC9YTDjXwfrsVucnLJK19R+Nw7AHw7IoU504/AsFtINmIEYgEsmo1RKaMTPTjtaXhMHsJ6hGRbMh5LEjnuXJIsHiJ6hLpQAxbFxDDPMFLtqTKjE0L0KsMweK/oHVRF5cghx+z3+STweklbz/2GJg2nzFvSqkWXqqro6GhAhnMIvnAjUT1CKBzmF//3CYVvfAHAlwdmsHD6cVSpDVgwiBHHYXaS48wjw5FBRI+QbPHwff13GKrCgSkHkeUcQlSP4bZ6MBRIc6STak1lRMrI/e4UI4QQ++uL6g18W/8Nx+eesN8rNEECr1ft+twvGAtS5i8lEA3gMDuI6lEcZjsOkwtdMWgMeclwZOHSHFy07N9Mejux99P/DsngnuvGodrNKGGVhkgD6bYMnGYnJpMJDAWzYiIQC5LlzCHNkY5JNeGPBRjpGcm4IROa24PJQhQhRF9QE6zh7aJ/U+AeypFDju6Sc0rg9SF2k51D0w/ji+rPaYw04twxQ9PjMXQDvqn9mng4wLmL/sKE9xNht/7IYTx89Xi8up/y2q8waWbcZg++aCMoBtnmIaCA2WQh1ZpGlmsIqqoSigSxalYMVZFVl0KIPsUX8fLKtysxqybOHH4WqtI1FXQSeH1Mii2ViUOOalWQPm7IkZR4i2n0VnL6fa8w9P2vAFh9TAFPXn0MDlcSFl8EXTFIsibjtLjwRXzkunPJducRjkXIVDNxWlyYFBVVMWFSNbyRRg7PGsfIlNG9fNVCCJEQjAV56duVBGJBLhxzEa4uuJXZRAKvD2qrxKG87BtOn72UoZ8kwu5fhQX88bJjiJhiKPEwGJDtyGZ02gG4LE5ieowUayoWxYyuxXGaHCgKqIpKOBpGRSPJlkySNak3LlEIIXYTiAZ4+duV1IVq+enoC8jez7q7XUng9QOhqlIOuOoGsj5L3MZ87dSR3HWmC1OwiHxzASoq4ViIuKFT7ivDpJpIsibhj/pJt6VT4BlGzIhTH6phiDMbw2LQGPWSbk2RTilCiD6hIVzPy9+upDHcyDkjf0yBZ1iXf4YEXg/Z1/3ylMpKUi8+H+eGRNj95cwx/P68EcSDFQQjjWxu+I4MexaaaiIWD9MY8WIzWTBjRlcNshzZeKweInoUs2YiHA8TM6KkWVMZN2SiPLMTQvS6ykAFL3/7IjE9ygVjLiLXndctnyOB1wP2db88tXg7Sb+4CNOXidKDZ84eypJT7OihclRVJRbXCQXrUVE5MPUgUMAwdOoCddQZdYxOHkWeJx+3xYM/5ufA1INQoEs2qBVCiP1lGAafV2/gne1vYTfZmXLgL0i3Z3Tb50ng9YA97Ze36+zPaXYS+2Yjw6+8FtOWbRiKwt8uPY6nJ0bBiBEnTCQex2VyEFHNRPUIVrONZGsy3oiXZGsqVrOVw9IPxxttoCHcABjYJOiEEH1EJB7h39v+yZe1GxnqHsaZI87GuUu7xa4mgdcD2uqbaVbNBKL+3WZ/3oiXsg//wcm3LsZWUoqhabx65Sm8c8oYopUfEwwFcFjs+A0/vrgft9mNw+zEbXaS58rHH/Nh15ygGKTaUogbMZKsSVhN1uYdGaT8QAjRq700M4ELgHTgXdi2aitPGEv26VSd6aHZRAKvG+w6YzOguW9mKBaiLlSLL+xFUVXKAuUYho7L4iLFmor2v485acb92Cpr0M1mtt11G2+PqcVmdTMmZQxry9dSH6rHYjJjVW04LA5sSmI3A504uh7HpzeS6RhCVbCKJGvyjvBL67Kd2IUQ/V9P9tJs2e/y+U1/Ztaqm3CZ3fz2tKc5/roTemQMIIHX5dp6XheKBQjFglg0K1WBSlRFJWbEiUZD1ASqGZEykmDYT/i9f3HKHUux13mJ2ayU3n8v3514BDnV66kIVHJoxlhqQjVUBisIRSMkWZMocA3DZrJi1WzYTHZ0IMWWTJo9A1+4EY/VQ6otDduOjVqbZpZCCNGT/FE/s1fdzAtfP0dh7iSeOO1pshxZPToGCbwu1tbzOo81mUg8Ql24Dh0dh9mBoih4VA9xPca2ui1kfvQpZ927HLs3QMRh473briD59JMJhasYnjyShkgj3pgPl9lDY7gRNIVMRwbDk4eTbs9EVRP71Y33DCfXnYfdZG/eQ2pvd2IXQogulQGnv3QS39R9zc0TZzFz4mw0VevxYUjgdbH2ntdF4xFSrSk4XHkoisLW+i2YVFNi4ck/n+ecx/+JNRAm6LTxfzeehu+o4WQ2fptYwKLHGJMyhspABR6rhyTrwbitHgqShhONB0m2eshwDeHQ9MNaPZtra0eGpn6ZQgjREzZWfw5XQk2ohpVn/5UT8k/qtbFI4HUxAyjybkc3dKwmKynWVFRFbZ5VNT3Ls5qsxCNhPK+/yY8e+QfWcBRfkoPlM86g6qgjCEXq8daFGJE0DAVwWdx4Iw0M9QzDF2kkP3kYadZUXBY3MSO2W9jBvu3ELoQQXSEaj/J20Vt8Ub0BSuA/V/6XLOeQXh1T13TkbMNnn33GJZdcAsC2bduYMmUKF198MfPmzUPXdQBWrlzJT37yEy688EL+85//ABAKhZg2bRoXX3wxV155JbW1tQCsX7+eCy64gIsuuoilS5c2f87SpUs5//zzueiii9iwYUN3XU6nBGNBQrEgwWgQk2IiFo+xrXEbjeF60uzppNnTCcdDROIRUhQXnlf+wgkLfp8IuzQPy285h28OzcdpdZNiS2ZE0nAcFjcRXachXE9D2EeqLYURKaMIRQOJjWJNVlKtqe2GWFObspEpo8lz50vYCSG6XU2whj9v+j++qN7AMdk/gOX0ethBN83wfv/73/Pqq69ityf+cr3//vuZPn06Rx99NHfeeSdvv/02RxxxBMuXL+fll18mHA5z8cUXc9xxx7FixQrGjBnDtGnTeP3111m2bBl33HEH8+bNY8mSJeTn53PVVVexceNGAD766CNefPFFysrKmDZtGi+//HJ3XFKn1ASr8ViTcFnc1IVqCcfC2DQrNpOjOWjy3AXU1peQsvJFhj3wO9RYnMYhabw252K2DrXjsTipDzeQZPOQ7kgUYH4X/IaD0w7BaXES1+OYNQs2zUbc0InGowSjAYq9Ra1q7Pa1s4sQQuyPL2s28ta2NzGpJs4f/TOGJQ3nQ/2D3h4W0E0zvIKCApYs2VlTsXHjRo466igAJk2axAcffMCGDRsYN24cFosFt9tNQUEBmzZtYt26dRx//PHNx65Zswafz0ckEqGgoABFUSgsLGTNmjWsW7eOwsJCFEUhJyeHeDzePCPsDeFYCLNqxmayke3KYVjycAo8Q1FaHGOPGIx84Q2GL1iEGosTHDaU9Q/djvPYUzGbrYTiYSLxECNTRmM1WVFQsGk2dEPHYXISjUcIRAIEY0HqQ3WE4mFyXHnNNXbBWLB5pahu6DjMzlZfE0KI7hDVo/xr6z94Y8vfyXJkcenBlzEsaXhvD6uVbpnhTZ48meLi4ubfG4aBoiT+2nc6nXi9Xnw+H273zm0fnE4nPp+v1estj3W5XK2OLSoqwmq1kpyc3Op1r9dLampqd1zWHllNtuZndE1aropUfF7sT/4Gx4P3oxgGgdEj+WT+DKpGZBM0QmQ4h+CNesmyZ+E2u4nGo/ijfoYmDSPDkYk/6sNt8RCMh4hEY6Q60hjqGdpccgCJWSbQYWcXIYToSvXhel7d/BcqAxUcPeRYjss9vsv2sOtKPbJoRVV3Xrjf78fj8eByufD7/a1ed7vdrV7v6FiPx4PZbG7zHL0h8fwuxLaGLbgsbjIdWWiq1rwqUmmox7HkERyPPwJAeOzhfDFvOlW5SdjtbhwoKCg4zA5SrWkEIoHmUoNcdx7VwSrS7OkMcWYT1aN8W/s1I5JHtQq7ljV27XV2EUKIrrS1YQuvff83DAx+POp8RiaP6u0htatHAu/ggw9m7dq1HH300axatYpjjjmGsWPH8uijjxIOh4lEImzevJkxY8Ywfvx43nvvPcaOHcuqVauYMGECLpcLs9nM9u3byc/PZ/Xq1Vx//fVomsZDDz3E5ZdfTnl5Obqu98rsrmWx+YjkUVQFK/m+4TuGeoaT5y7A0eDHsWghjqeeBCB61NFsvmsO2vACCkwadeHE875kWzK5plysJhsVgXJ0PY7NZMfWxmrLoUnDd6tjaTmb7GimKYQQXdJarBA4BagEnoe/1L3U5mF70/6rO/VI4M2aNYu5c+eyePFiRowYweTJk9E0jUsuuYSLL74YwzCYMWMGVquVKVOmMGvWLKZMmYLZbGbRokUAzJ8/n5kzZxKPxyksLOTwww8HYOLEifzsZz9D13XuvPPOnric3bQqNtegwDyUSDyCqqg4ahqx3XcXjhV/BsB77NE03v8Q3nQLdmuiAD3blPifwTAMaoM1GOjkuwuaa+ea+l+2vB3ZFLLQdo2d1N8JITqyp9ZiLduB7coX9XH9v6/mjS1/5yejz2fRVUtwzu/exs9dQTEMw+jtQXS3qipvt55/c923OMzO5ueUAIauEy3awoGPPYXzL68AUH9CIZ/O+BWVmS4sdjeZ9kw8Vk/zeyLxCBWBCrIcWa1mZ03huevzt45WYsoqTSEGn4yMzj/SURRlnwKvwl/Oz9+4kI3Vn3PXD+7lqrHXtvq7r7d19D2QwvMusNtiFV2H7dsYed8inG+9DUDt5FNYN/Ui9Px83CaVYCzE9satFHiG4ba4m2dhDpMDs2pudf72nr811di1paOvCSHEvthU+xUXv3Y+deE6nj3zBU4Z+sPeHtJekcDrAq1aeOkK6vat5N11L8mrPwSg7tyz2XjlhejZmZhsdjTDIG7EGeIcQkOkAW1HJ5am53Ty/E0I0dd8Uv4RU14/H5tm49Xz/sFhGYf39pD2mgReO/bmlmBTC6/ahlL0779j+PyFuD/6BIDqiy6g7Mpf0eA2sFl3Liixmqy4LW40RWVkyujmc0n/SyFEX/Nh2RqmvPZTMuwZvHzu38nvp38f9b1CiT5gXwq37TGF/PIgh9y5M+wqf30J3/3qAr5zRQiria4okXiEqB4hxZra5sytKTxVRSUQ9e94dif9L4UQveODktVc9Pcfk+3M5tUfv9lvww5khtemtrb4aXq9zedigQDat9/gnnE95i8S/TyLrv4V9T+7EEduPplGiOLGbVQFfKQ50hniyEFV1HZnbvL8TQjRF3xevYFfvPEz8t0FvHzua2Q6Mnt7SPtFAq8N7W3x09bCEcXnRft6E+4br8X0zdcYikLZjGnUn3U65BWgmM14sDAq9QAi8Qg2ky3Rgkwxy8xNCNF3JcOU135KkjWJF87+S78PO5DAa9OeWoRB4rZnfcX3qF9uZPSc+Zi2F2FoGv6586koPAwtNx/FvHO1ZdOeeDJzE0L0dZF4GH6e+Pnlc/5Ojiu3t4fUJSTw2rCnhSPBWJDybRtI+m4bw2+9A0tZObrZRP1d92CcdiZ4DKIaWFqcU1ZaCiH6A8Mw+OfWNyENnjn9WQ5IPbC3h9RlZNFKG/a0cKShaBPJmzYz4qbZibCzWtk+fy5lPxiHnptHmie7ed87wzCIxCOE4yHS7Om9fGVCCNGxDdXr+bruK3gHCnMn9fZwupTM8NrR5sIRw0CtKMf8v/8x4ra7MdXVEXc4KFkwn9C4I/Cluck0m7Fjlp3GhRB9WnZuTtu9NE8C3DBkS3aPj6m7SeB1lq6jlpVi+t8njL51LprXR9ztoviBBQQPPYRARioWq7X5cFlpKYToyz77dFNvD6HHSeB1RiyGWlaK+aMPcd98I0rATzQ5ia0L7yF+wEEEMlIJaXHy5JalEEL0WRJ4exKJJMLug/dx33oTSjhMPDOL2kceI5KTjC/NjcVqJU+aMwshRJ8mgdeRWAy1pBjLe+/gun0WSjRKPDcX75LfYsrNJzM3j0yLZc/nEUII0etklWZHolGsb76Oa84tibAbOozGZU8Rz81Hz80DCTshhOg3ZIbXAdvzf8Z51x0ohkFs9AF4H/sNelq6hJ0QQvRDMsNrh/b5BtyzbkqE3aGH4f3NkxJ2QgjRj8kMrx2G04memkbs4EPwzV+A4XJL2AkhRD8mgdcOfcRIatZ9gVZeiqGZ0HNyJeyEEKIfk1uaHVFVCTshhBggJPA6omkSdkIIMUDILc2OSNAJIcSAITM8IYQQg4IEnhBCiEFBAk8IIcSgIIEnhBBiUJDAE0IIMShI4AkhhBgUJPCEEEIMChJ4QgghBgUJPCGEEIOCBJ4QQohBQQJPCCHEoCCBJ4QQYlCQwBNCCDEoKIZhGL09CCGEEKK7yQxPCCHEoCCBJ4QQYlCQwBNCCDEoSOAJIYQYFEy9PYCB7rPPPuPhhx9m+fLlbNu2jdmzZ6MoCqNHj2bevHmoqsrKlSt5/vnnMZlMTJ06lZNOOolQKMQtt9xCTU0NTqeThQsXkpqayvr161mwYAGaplFYWMj1118PwNKlS3n33XcxmUzcdtttjB07dr/HHo1Gue222ygpKSESiTB16lRGjRrVr64hHo9zxx13sGXLFjRN4/7778cwjH51DQA1NTX85Cc/4ZlnnsFkMvW78Z933nm43W4A8vLyuOaaa/rVNTz55JO88847RKNRpkyZwlFHHdWvxi92MES3+d3vfmecddZZxgUXXGAYhmFcffXVxocffmgYhmHMnTvX+Ne//mVUVlYaZ511lhEOh43GxsbmXz/zzDPG448/bhiGYbz22mvGPffcYxiGYZxzzjnGtm3bDF3XjSuuuML44osvjC+++MK45JJLDF3XjZKSEuMnP/lJl4z/pZdeMu69917DMAyjtrbWOOGEE/rdNbz11lvG7NmzDcMwjA8//NC45ppr+t01RCIR49prrzV++MMfGt99912/G38oFDLOPffcVq/1p2v48MMPjauvvtqIx+OGz+czHn/88X41frGT3NLsRgUFBSxZsqT59xs3buSoo44CYNKkSXzwwQds2LCBcePGYbFYcLvdFBQUsGnTJtatW8fxxx/ffOyaNWvw+XxEIhEKCgpQFIXCwkLWrFnDunXrKCwsRFEUcnJyiMfj1NbW7vf4Tz/9dG688cbm32ua1u+u4dRTT+Wee+4BoLS0lPT09H53DQsXLuSiiy4iMzMT6H//H23atIlgMMhll13GpZdeyvr16/vVNaxevZoxY8Zw3XXXcc0113DiiSf2q/GLnSTwutHkyZMxmXbeNTYMA0VRAHA6nXi9Xnw+X/OtnqbXfT5fq9dbHutyuVod29Hr+8vpdOJyufD5fNxwww1Mnz69310DgMlkYtasWdxzzz1Mnjy5X13DK6+8QmpqavNfmND//j+y2WxcfvnlPP3008yfP5+ZM2f2q2uoq6vjiy++4LHHHuuX4xc7yTO8HqSqO/994ff78Xg8uFwu/H5/q9fdbner1zs61uPxYDab2zxHVygrK+O6667j4osv5uyzz+ahhx7qd9cAiVnSzJkzufDCCwmHw/3mGl5++WUURWHNmjV89dVXzJo1q9W/+Pv6+AGGDx/O0KFDURSF4cOHk5yczMaNG/vNNSQnJzNixAgsFgsjRozAarVSXl7eb8YvdpIZXg86+OCDWbt2LQCrVq1i4sSJjB07lnXr1hEOh/F6vWzevJkxY8Ywfvx43nvvveZjJ0yYgMvlwmw2s337dgzDYPXq1UycOJHx48ezevVqdF2ntLQUXddJTU3d7/FWV1dz2WWXccstt3D++ef3y2v461//ypNPPgmA3W5HURQOPfTQfnMNf/7zn3n22WdZvnw5Bx10EAsXLmTSpEn9ZvwAL730Eg888AAAFRUV+Hw+jjvuuH5zDRMmTOD999/HMAwqKioIBoMce+yx/Wb8YidpLdbNiouLuemmm1i5ciVbtmxh7ty5RKNRRowYwb333oumaaxcuZIXXngBwzC4+uqrmTx5MsFgkFmzZlFVVYXZbGbRokVkZGSwfv167rvvPuLxOIWFhcyYMQOAJUuWsGrVKnRdZ86cOUycOHG/x37vvffyj3/8gxEjRjS/dvvtt3Pvvff2m2sIBALMmTOH6upqYrEYV155JSNHjuxX/x2aXHLJJdx1112oqtqvxh+JRJgzZw6lpaUoisLMmTNJSUnpV9fw4IMPsnbtWgzDYMaMGeTl5fWr8YsECTwhhBCDgtzSFEIIMShI4AkhhBgUJPCEEEIMChJ4QgghBgUJPCGEEIOCFJ4LscPatWuZPn06o0aNan4tJSWFxx9/vMs/6+uvv6axsZEjjzxyj8eGw2HOOOMM3nnnnS4fhxCDiQSeEC0cc8wxPPLII93+Of/6179IT0/vVOAJIbqGBJ4QHYjFYvziF7/guuuu46CDDuKXv/wlTz31FLfeeivDhw9ny5YtGIbBI488QkZGBosWLeLjjz/GMAx+9atfccYZZ/DZZ5+xYMECDMMgKyuLuXPn8pe//AWz2cwhhxxCKBTikUceQdM08vPzufvuu4lEIsycOZPGxkYKCgp6+9sgxIAggSdECx9++CGXXHJJ8+9POOEEHn74Ya655hoyMjK49dZbyc7OBmD8+PHcfffd/PnPf+bJJ5/k+OOPp7i4mOeff55wOMyFF17Icccdx9y5c3nkkUcYOXIkf/7zn6murubHP/4x6enpHHbYYZx++uk899xzpKWl8eijj/KXv/yFyP+3d/+qqYNxGMe/ChmKDjr4B3GK4ioIwUFwcvECOmgXJ3G0ozp1zQU4KIjiBfQCBC/AQSKoi4uTg1DdRIpoJ0N74HCgwzmH5vmMSZZkefL+Et7n/Z1MJsPz8zOLxcLdxkpEvk+BJ/LJ70aauVwOx3EoFotfrr2fm06nxGIxVquVG5iXy4Xdbsfb2xupVAqAp6cnAPd73OFwYL/f02w2ATifzxQKBY7Ho9uQkM1mv7RuiMj36C9NkT9wHIfNZoNlWQwGA/f4crkEYD6fk06nMU2TfD7PeDxmNBpRLpdJJpNEo1G22y0AvV6PyWSCz+fjer0SDoeJx+N0u13G4zGNRoN8Po9pmjiOA8B6veZyufzt2xb5cfTaKPLJryPNe0dZv98nkUjw+PjoFn++vr4yHA55eHjAtm1CoRCz2YxqtcrpdKJUKhEMBnl5eaHdbuP3+4lEItRqNQzDwLZtUqkUnU6Her3O7XYjEAhg2zaWZdFqtahUKpimiWEY/+qRiPwY2jxa5BvuzQX3UaWI/P800hQREU/QCk9ERDxBKzwREfEEBZ6IiHiCAk9ERDxBgSciIp6gwBMREU/4AJx0cWPt3vcWAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Creating a base dataframe\n",
    "df = pd.DataFrame(xt)\n",
    "\n",
    "#Adding predicted, expected, error abs and error % columns\n",
    "df[\"Predicted\"] = final_predictions\n",
    "df[\"Expected\"] = yt\n",
    "df[\"Error abs\"] =  (round(abs((df[\"Expected\"] - df[\"Predicted\"]) )))\n",
    "df[\"Error %\"] = (round(abs((df[\"Expected\"] - df[\"Predicted\"]) )/ df[\"Expected\"],3))*100\n",
    "\n",
    "#Printing maximum error, error mean and error std\n",
    "print(\"Error max  = \" ,(df[\"Error %\"].max()))\n",
    "print(\"Error mean  = \" ,(df[\"Error %\"].mean()))\n",
    "print(\"Error std  = \" ,(df[\"Error abs\"].std()))\n",
    "\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "#Plot Predicted distribution\n",
    "plt.style.use('seaborn-dark')\n",
    "\n",
    "sns.jointplot(x=\"Expected\" , y =\"Predicted\", data = df, kind= \"reg\" , palette= \"viridis\", color=\"green\",  joint_kws={'line_kws':{'color':'red'}, 'scatter_kws': {'alpha': 0.1}})\n",
    "path = pathlib.Path.cwd() / 'results' / \"results.csv\"\n",
    "#Save to csv\n",
    "df.to_csv(path, index_label= False)\n",
    "df.head(6)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f095f5d",
   "metadata": {},
   "source": [
    "# Save Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "4bb11b17",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['/Users/kike/Python/Machine Learning/Proyecto Regresion/process sets/housing_model.sav']"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#TO SAVE\n",
    "import joblib\n",
    "path = pathlib.Path.cwd() / 'process sets' / 'housing_model.sav'\n",
    "\n",
    "filename = path\n",
    "joblib.dump(final_model , filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "f7ac1a99",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ready\n"
     ]
    }
   ],
   "source": [
    "#TO LOAD\n",
    "import joblib\n",
    "\n",
    "loaded_model = joblib.load(filename)\n",
    "result = loaded_model.predict(x)\n",
    "print(\"ready\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "83310434",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "46ceeae7",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}