main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Insurance Prediction Model\n",
    "Machine Learning Project - 2024"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier \n",
    "from sklearn import svm\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load and display train dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building_Painted</th>\n",
       "      <th>Building_Fenced</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Geo_Code</th>\n",
       "      <th>Claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>1240.0</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>75117</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>900.0</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>without</td>\n",
       "      <td>62916</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>4984.0</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>31149</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>600.0</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>6012</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>900.0</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>without</td>\n",
       "      <td>57631</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>550.0</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>without</td>\n",
       "      <td>33063</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>13004</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>480.0</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>94059</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>536.0</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>74243</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>NaN</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Building_Painted  \\\n",
       "0                  2012             1.0            1                N   \n",
       "1                  2012             1.0            0                N   \n",
       "2                  2013             1.0            1                V   \n",
       "3                  2016             0.5            0                N   \n",
       "4                  2012             1.0            0                N   \n",
       "...                 ...             ...          ...              ...   \n",
       "5007               2013             1.0            0                N   \n",
       "5008               2012             0.5            0                V   \n",
       "5009               2015             1.0            1                V   \n",
       "5010               2012             0.5            0                V   \n",
       "5011               2013             1.0            1                V   \n",
       "\n",
       "     Building_Fenced Garden Settlement  Building Dimension    Building_Type  \\\n",
       "0                  V      V          U              1240.0      Wood-framed   \n",
       "1                  V      V          U               900.0  Non-combustible   \n",
       "2                  N      O          R              4984.0  Non-combustible   \n",
       "3                  V      V          U               600.0      Wood-framed   \n",
       "4                  V      V          U               900.0  Non-combustible   \n",
       "...              ...    ...        ...                 ...              ...   \n",
       "5007               V      V          U               550.0         Ordinary   \n",
       "5008               N      O          R              1000.0   Fire-resistive   \n",
       "5009               N      O          R               480.0         Ordinary   \n",
       "5010               N      O          R               536.0   Fire-resistive   \n",
       "5011               V      V          U                 NaN      Wood-framed   \n",
       "\n",
       "     NumberOfWindows Geo_Code Claim  \n",
       "0            without    75117   non  \n",
       "1            without    62916   non  \n",
       "2                  4    31149   oui  \n",
       "3            without     6012   oui  \n",
       "4            without    57631   non  \n",
       "...              ...      ...   ...  \n",
       "5007         without    33063   oui  \n",
       "5008               4    13004   non  \n",
       "5009               3    94059   non  \n",
       "5010               4    74243   non  \n",
       "5011         without      NaN   non  \n",
       "\n",
       "[5012 rows x 12 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance = pd.read_csv(\"./train_Insurance.csv\")\n",
    "df_insurance.drop(axis=1, labels=['Customer Id'], inplace=True)\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Phase 1: Analyse Data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5012 entries, 0 to 5011\n",
      "Data columns (total 12 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   YearOfObservation   5012 non-null   int64  \n",
      " 1   Insured_Period      5012 non-null   float64\n",
      " 2   Residential         5012 non-null   int64  \n",
      " 3   Building_Painted    5012 non-null   object \n",
      " 4   Building_Fenced     5012 non-null   object \n",
      " 5   Garden              5008 non-null   object \n",
      " 6   Settlement          5012 non-null   object \n",
      " 7   Building Dimension  4935 non-null   float64\n",
      " 8   Building_Type       5012 non-null   object \n",
      " 9   NumberOfWindows     5012 non-null   object \n",
      " 10  Geo_Code            4939 non-null   object \n",
      " 11  Claim               5012 non-null   object \n",
      "dtypes: float64(2), int64(2), object(8)\n",
      "memory usage: 470.0+ KB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "YearOfObservation      0\n",
       "Insured_Period         0\n",
       "Residential            0\n",
       "Building_Painted       0\n",
       "Building_Fenced        0\n",
       "Garden                 4\n",
       "Settlement             0\n",
       "Building Dimension    77\n",
       "Building_Type          0\n",
       "NumberOfWindows        0\n",
       "Geo_Code              73\n",
       "Claim                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance.dtypes\n",
    "df_insurance.info()\n",
    "df_insurance.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['YearOfObservation', 'Insured_Period', 'Residential', 'Building Dimension']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_numeric_attributes(df:pd.DataFrame) ->list:\n",
    "  return list(df_insurance.select_dtypes(include='number'))\n",
    "\n",
    "def get_discrete_attributes(df:pd.DataFrame)->list:\n",
    "  return list(df_insurance.select_dtypes(include='object'))\n",
    "\n",
    "get_numeric_attributes(df_insurance)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAApAAAAIbCAYAAABcwq5LAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAACWKklEQVR4nOzdd1wUV9sG4HuX3ntVVBAL2ECNiiXRiGIjosaowdhbjBULGnslMfYSu1GjRhNbovHFrknsIthAKYJYqNKkLW3fP9DVFVDWBRad+/p+834ye+bsMzsBHp5zzoxIKpVKQURERERUSmJVB0BEREREHxYmkERERESkECaQRERERKQQJpBEREREpBAmkERERESkECaQRERERKQQJpBEREREpBAmkERERESkECaQRERERKQQJpBEREREpBAmkEREREQV5J9//oGnpydsbW0hEolw+PDhdx5z7tw5NG7cGFpaWnB0dMT27duLtFm3bh1q1KgBbW1tNG/eHFevXi374F/DBJKIiIiogmRkZKBRo0ZYt25dqdpHRkaia9euaNeuHYKCgjBhwgQMGzYMx48fl7XZt28ffHx8MGfOHNy4cQONGjWCh4cH4uPjy+s0IJJKpdJy652IiIiIiiUSiXDo0CF4eXmV2MbX1xd///037ty5I9vXt29fpKSkwN/fHwDQvHlzfPLJJ1i7di0AoKCgAHZ2dhg7diymTZtWLrGzAklERET0niQSCdLS0uQ2iURSZv1funQJ7u7ucvs8PDxw6dIlAEBOTg4CAgLk2ojFYri7u8valAf1cuuZ6DW5iQ9UHQK9sKrxbFWHQK9pmZul6hDoBU21fFWHQC80fXy43N+jrH4v+a3diXnz5sntmzNnDubOnVsm/cfGxsLKykpun5WVFdLS0pCVlYXk5GTk5+cX2+bevXtlEkNxmEASERERvafp06fDx8dHbp+WlpaKoqk4TCCJiIhIeArKpuKspaVVrgmjtbU14uLi5PbFxcXB0NAQOjo6UFNTg5qaWrFtrK2tyy0uzoEkIiIi4ZEWlM1Wztzc3HD69Gm5fSdPnoSbmxsAQFNTE02aNJFrU1BQgNOnT8valAcmkEREREQVJD09HUFBQQgKCgJQeJueoKAgREdHAygcEh8wYICs/ahRo/DgwQNMnToV9+7dw88//4zff/8dEydOlLXx8fHB5s2bsWPHDoSEhODbb79FRkYGBg8eXG7nwSFsIiIiEp6C8q8eFuf69eto166d7OuX8ycHDhyI7du3IyYmRpZMAoC9vT3+/vtvTJw4EatWrULVqlWxZcsWeHh4yNr06dMHCQkJmD17NmJjY+Hi4gJ/f/8iC2vKEu8DSRWCq7ArD67Crly4Crvy4CrsyqMiVmHnPL1bJv1o2tYrk34+NKxAEhERkfCoqAL5seAcSCIiIiJSCCuQREREJDwVsIL6Y8YEkoiIiISnjO4DKVQcwiYiIiIihbACSURERMLDIWylMIEkIiIi4eEqbKVwCJuIiIiIFMIKJBEREQmOlEPYSmECSURERMLDIWylcAibiIiIiBTCCiQREREJD4ewlcIEkoiIiISHNxJXChNIIiIiEh5WIJXCOZBEREREpBBWIImIiEh4uApbKUwgiYiISHg4hK0UDmETERERkUJYgSQiIiLh4RC2UphAEhERkeBIpbyNjzI4hE1ERERECmEFkoiIiISHi2iUwgSSiIiIhIdzIJXCBJKIiIiEhxVIpXAOJBEREREphBVIIiIiEp4CrsJWBhNIIiIiEh4OYSuFQ9hEREREpBBWIImIiEh4uApbKUwgiYiISHg4hK0UDmETERERkUJYgSQiIiLh4RC2UphAEhERkfAwgVSKIIawN23aBDs7O4jFYqxcuVKhYwcNGgQvL69yiUtVRCIRDh8+rOowiIiI6AOlVAVSKpWiQ4cOUFNTw/Hjx+Ve+/nnn/H999/jzp07qFq1qlJBFic/Px+rV6/Gtm3bEBYWBh0dHbRo0QIzZ85Eq1atZO3S0tIwZswYLF++HL169YKRkREAICsrCz/88AN+++03PHz4EAYGBmjXrh3mzp2LevXqlXm8qjB37lwcPnwYQUFBcvtjYmJgYmKimqA+QNeDbuOXPfsRfC8cCc+SsMpvFtp/2lLVYQlCK59eaPB1O2gZ6uLp9VCc/P4XpETFldi+Uf/2cPmmPQyrWgAAnoU+xqVVhxB57lZFhfzBspvSF1be7lAz1MXza/fxYNomZEfGvPUY60GdYDu6OzQtjJERHIXIGVuRHhQOAFA31ofd5D4w/qwRNKuYIy8pDUn/u4roJXuR/zxTrh+Lr9rBdqQndBxskJ+ehcQjFxH5/ZZyO9fKwmJgZ1iP6gENC2NkhkTh0azNyAgKK7G9SdeWsJ3yNbSqWiI7KgZPFu9E6pkAuTa2k/vBvF8HqBvpIf3aPTz8fgMkr13HBpc2QcvOUu6Yx347EbvuIADAwK0+LId5Qs+lFtQMdCGJjEHshkNIOvRPGZ555SCV8kbiylCqAikSifDLL7/gypUr2Lhxo2x/ZGQkpk6dijVr1pR58pibmwupVIq+ffti/vz5GD9+PEJCQnDu3DnY2dmhbdu2ctW16Oho5ObmomvXrrCxsYGuri4kEgnc3d2xbds2LFy4EKGhoTh27Bjy8vLQvHlzXL58uUxjVlRubm659m9tbQ0tLa1yfY+PSVZWNuo4OmDGpNGqDkVQmn3bDa6DO+Lk9G3Y/cUc5GZK8OUuX6hpaZR4zPPYJPzzwz782nUmdnWbheiLwfDa4gOz2lUqMPIPT5XvvGAztAsifDfidtfpKMjMhvNvsyB6y2dt9kVL1Jg7CI+X/Y6bHlOQEfwQzr/NgoaZIQBA08oEmtamiJq/E0HtJiJs/FoYt3OF43L57yObkZ6oNq0fnqw9iMC2E3D3q3lIORdUnqdbKZh4toLd7CF4umIvgjv7ICs4CrV2zYG6mVGx7fWa1IHDuklI3HsKwZ18kOJ/BTW3TIN2nWqyNtaje8BycDdET9+AEM+pyM/MRu1dc4pcxyc/7UGQ6yDZFr/t79fepy6yQh4iYsSPCO4wAYm/n4b9yvEwat+0fD4IVSooKJvtPaxbtw41atSAtrY2mjdvjqtXr5bYtm3bthCJREW2rl27ytoMGjSoyOudOnV6r9hKS+khbDs7O6xatQqTJ09GZGQkpFIphg4dio4dO8LV1RWdO3eGvr4+rKys8M033yAxMVF2rL+/P1q3bg1jY2OYmZmhW7duiIiIkL0eFRUFkUiEffv24bPPPoO2tjZ2796N33//Hfv378fOnTsxbNgw2Nvbo1GjRti0aRO++OILDBs2DBkZGdi+fTsaNGgAAHBwcIBIJEJUVBRWrlyJS5cu4ejRo/jqq69QvXp1NGvWDAcOHICTkxOGDh0KqVQqd57z5s2DhYUFDA0NMWrUKOTk5Mhe279/Pxo0aAAdHR2YmZnB3d0dGRkZste3bNkCJycnaGtro27duvj555/feo7r16+Hjo4O/ve//8nFcOjQIRgYGCAzs/Cvd19fX9SuXRu6urpwcHDArFmzZMnn9u3bMW/ePNy8eVP2H9P27dsBFB3Cvn37Nj7//HNZ/CNGjEB6errs9ZfD+EuXLoWNjQ3MzMzw3XfflXuiW1m0cfsE40YMhPtnrd7dmMpM46GdcHnNn4g4eQOJ9x7h2MQN0Lc0hmPHJiUe8+BUICLP3kRKVBySI2Px309/ICczGzaujhUY+YfHZng3PF65H8nHryEz5CHCxq2BppUJTDs1K/EY25GeiNt9CvH7ziIr9DEeTN2I/CwJLPu1BwBk3n+E+8N+QvLJ65A8jEPahTuI/mEPTDo0BdQKf/WoGemhmm8/hI9bg8RD/0HyMA6ZIQ+RfOJ6hZy3KlmN6I7E307g2e9nkB32GA+nrUdBtgTmfdsX336oJ1LP3UDchsPIDn+Mp0v3IPPOA1gO6iJrYznUEzGrf0fKiavICnmIqAmroGFlCmOP5nJ95adnIS8hRbYVZElkr8Wu3Y+nS/cgI+A+JA9jEb/1KFLPBcKkc4vy+SBUSVpQNpuC9u3bBx8fH8yZMwc3btxAo0aN4OHhgfj4+GLbHzx4EDExMbLtzp07UFNTQ+/eveXaderUSa7db7/99l4fS2mVyRzIgQMHon379hgyZAjWrl2LO3fuYOPGjfj888/h6uqK69evw9/fH3Fxcfjqq69kx2VkZMDHxwfXr1/H6dOnIRaL0aNHDxS8kdFPmzZNVmn08PDAnj17ULt2bXh6ehaJZdKkSXj27BlOnjyJPn364NSpUwCAq1evIiYmBnZ2dtizZw86dOiARo0ayX8YYjEmTpyI4OBg3Lx5U7b/9OnTsirnb7/9hoMHD2LevHkACoeD+/XrhyFDhsja9OzZU5aA7t69G7Nnz8aiRYsQEhKCxYsXY9asWdixY0eJ59i7d29069YNe/bskWuze/dueHl5QVdXFwBgYGCA7du3Izg4GKtWrcLmzZuxYsUKAECfPn0wadIk1KtXT/YfU58+fYp8XhkZGfDw8ICJiQmuXbuGP/74A6dOncKYMWPk2p09exYRERE4e/YsduzYge3bt8sSUqKyZlTNAvqWxnj43x3ZvpznWYgJioBtk1ql6kMkFqGOZwto6Ggh5kbJw4JCp1XNCppWJkj599Uwf/7zTDwPDINB0zrFHiPSUId+w5pIfe0YSKVI/fcWDJrULvG91Ax1kZ+eCeQX/ow3/rQRRCIRNG1M4fLPKjQJ2ITaGydB09asbE6ukhJpqEOvQU2kvfH5pf17E3qNi//M9ZrUkW8PIO18IPSbFLbXrGYFTStTuTb5zzORERQqa/OSzXc94XJ7J5z9l8NqlJcsoS+JmoEu8lLS39qGSm/58uUYPnw4Bg8eDGdnZ2zYsAG6urrYtm1bse1NTU1hbW0t206ePAldXd0iCaSWlpZcu/KeqlZmq7A3bdqEevXq4Z9//sGBAwewceNGuLq6YvHixbI227Ztg52dHUJDQ1G7dm306tVLro9t27bBwsICwcHBqF+/vmz/hAkT0LNnT9nXoaGhcHJyKjaOl/tDQ0Ph5eUFM7PCH0QWFhawtraWvdauXbt3Hu/i4gIA0NTUxLZt26Crq4t69eph/vz5mDJlChYsWICYmBjk5eWhZ8+eqF69OgDIqp4AMGfOHCxbtkwWv729PYKDg7Fx40YMHDiwxHP09vbGN998g8zMTOjq6iItLQ1///03Dh06JGszc+ZM2b9r1KiByZMnY+/evZg6dSp0dHSgr68PdXV12XkXZ8+ePcjOzsbOnTuhp6cHAFi7di08PT3x448/wsrKCgBgYmKCtWvXQk1NDXXr1kXXrl1x+vRpDB8+vEifEokEEolEbp9YIuGwOZWanoUxACAzMU1uf2ZiGvQsih/ie8m8TlV8fXgu1LU0kJORjT9HrMSzsKflFeoHT9PSGACQm5Aitz83IRWaL67Dm9RNDSBSV0NOMcfoOBY/XUDd1AB2E3sjbtcp2T7t6laAWIQq43ohatY25KVloJrv13DeNwc3P/eBNDfvfU+rUnv5+b35meclpkLbsfhpXxoWxshLlG+fm5AKDQsT2euFfZTcBgDitx1F5p0HyEt5Dv0mdVFl2jfQsDTB4/m/FPu+Jt1aQa9RLTyctr70J/ihKKNV2MX9ztPS0ir2d15OTg4CAgIwffp02T6xWAx3d3dcunSpVO+3detW9O3bV/Y7+6Vz587B0tISJiYm+Pzzz7Fw4UJZDlQeymwVtqWlJUaOHAknJyd4eXnh5s2bOHv2LPT19WVb3bp1AUA2TB0WFoZ+/frBwcEBhoaGqFGjBoDCeYuva9q06NyLN4eYFaXI8Y0aNZJV/QDAzc0N6enpePToERo1aoT27dujQYMG6N27NzZv3ozk5GQAhdW9iIgIDB06VO5zWLhwodxQPVD0HLt06QINDQ389ddfAIADBw7A0NAQ7u7usjb79u1Dq1atYG1tDX19fcycObPIZ/cuISEhaNSokdx/iK1atUJBQQHu378v21evXj2oqanJvraxsSmx3O7n5wcjIyO57cdVGxSKi4TFyaslxoVskW1idbV3H1SCpAcx2NlpBnZ3n4Obu06j8/KRMKtlW4bRftjMe7ZB8/Bdsk2k8f6fdWmp6evA6dfvkRn6CI+W7nv1glgEsaYGImduRcq5IKTfCEPo6BXQsbeGUav6JXdI7y1u8194fukOskIeImHXcTxa8AssB3eFSLNoPcmgZX3UWD4WD6euQ3boIxVEW87KaAi7uN95fn5+xb5lYmIi8vPzZcWZl6ysrBAbG/vOkK9evYo7d+5g2LBhcvs7deqEnTt34vTp0/jxxx9x/vx5dO7cGfn55bdQqEzvA6murg519cIu09PTZVWsN9nY2AAAPD09Ub16dWzevBm2trYoKChA/fr15eYXAiiSZdeuXRshISHFxvByf+3aJQ+jKHv869TU1HDy5ElcvHgRJ06cwJo1azBjxgxcuXJFlnRu3rwZzZs3L3Lc6948R01NTXz55ZfYs2cP+vbtiz179qBPnz6yz/fSpUvw9vbGvHnz4OHhASMjI+zduxfLli0rVdyK0tCQn4QtEomKTDV4afr06fDx8ZHbJ37+pFzioo9D+MkbiAl89UeVmlbhf+e65obIiE+R7dc1N0R88Nv/SCrIzUfKw8KV2nG3o2DdyAGNh3TCyenFDw8JTdLxa0h/bUhfpFn4va1hYYzc1z5rDQsjZNyNKraPvKTnkOblF6lQalgYyfUBAGI9bTjtmYn89GzcG7IE0rxXv9By4wr/2M56LTnJe5aG3KTn0Kxi/h5n92F4+flpvPH5qZsbITc+udhjchNSoG4u317Dwgi5Ccmy1wv7MJbrQ8PCCJl3I0uMJSMwFGINdWhWtYTkwatKvX6LenD8ZQYezduGZwfOlf7kBKi433nlNeK2detWNGjQAM2ayc9P7tu3r+zfDRo0QMOGDVGzZk2cO3cO7dsXP69WWeV2H8jGjRvj7t27qFGjBhwdHeU2PT09PHv2DPfv38fMmTPRvn17ODk5ySp379K3b1+EhYXhyJEjRV5btmwZzMzM0KFDh7cef+rUKbl5jgBQUFCAFStWwNnZWW5+5M2bN5GVlSX7+vLly9DX14ednR2AwmSqVatWmDdvHgIDA6GpqYlDhw7BysoKtra2ePDgQZHPwN7e/p3n6e3tDX9/f9y9exdnzpyBt7e37LWLFy+ievXqmDFjBpo2bYpatWrh4cOHcsdramq+868PJycn3Lx5U27Rz4ULFyAWi1GnTvFzcd5FS0sLhoaGchuHr+ltcjOykfIwTrY9C32C9PgUVG/16pZamvo6sHGpiacBis1nFIlEUCumuiJUBRnZyI6KlW1ZoY+QE5cM49avpt6o6evAwLUWnl+/X2wf0tw8pN+KgNFrx0AkglHrhngeECrXT729syHNzcO9QX6QSuQX3qVduwcA0Kn5athb3VgfGqYGkDxOKIvTrZSkuXnIuB0Bg9YNX+0UiWDYuiEybhT/mWcE3Ifh6+0BGLZxQXpAYfuc6DjkxCXJtRHr60DPpbasTXF069lDmp+PvGepsn0GbvVRa8dMPF68E4m7T7zPKX4YymgVtiK/88zNzaGmpoa4OPnbkcXFxb11uhlQOKq5d+9eDB069J2n5uDgAHNzc4SHh5f+81BQuSWQ3333HZKSktCvXz9cu3YNEREROH78OAYPHoz8/HyYmJjAzMwMmzZtQnh4OM6cOVMkgy9J37590aNHDwwcOBBbt25FVFQUbt26hZEjR+Kvv/7Cli1bilT0Xjdx4kQ0a9YMnp6e+OOPPxAdHY1r166hV69eCAkJwdatWyESiWTtc3JyMHToUAQHB+PYsWOYM2cOxowZA7FYjCtXrmDx4sW4fv06oqOjcfDgQSQkJMjmUs6bNw9+fn5YvXo1QkNDcfv2bfzyyy9Yvnz5O8/z008/hbW1Nby9vWFvby9XxaxVqxaio6Oxd+9eREREYPXq1XLzI4HCeZGRkZEICgpCYmJikTkaQGGSqq2tjYEDB+LOnTs4e/Ysxo4di2+++aZIiV2oMjOzcC80AvdCCytkT57G4V5oBGJiix/Cp7JxY6s/WozzQs0OjWFepyo6rxiJ9PgUhJ94dd+73r9Nh+vAV38stvH9ClWb1YFhVXOY16mKNr5fwc7NCSGHL6riFD4YMZuPouqEL2HSsSl061aD45pxyIlLRpL/q1uLOP8+B9aDO8u+frrxCKy83WHRuy10alWBw48joKarhfi9ZwAUJo/Oe2dDrKuNcJ+foaavCw0L48Kqm7jwV0/2gxg8878K+wVDYNC0DnTr2MFx1VhkhT9F2oU7+JjFbfoTFv06wOzLdtB2rIrqfqMg1tFG4r7TAIAaK8ejyrT+r9pvPQLDtq6wGtEd2jWrwNanL3Qb1kT89mOyNvFbj8BmXG8YdfgEOnWrw37lBOTGJSHl+BUAgF7jOrAc6gkdpxrQrGYF0x6fwm7OEDw7eB75qYVFBIOW9eG4Yybit/2N5GOXoG5hDHULY6gZ61fgp1NBVLAKW1NTE02aNMHp06dl+woKCnD69Gm4ubm99dg//vgDEokE/fv3f2s7AHj8+DGePXsmG/EtD+X2Z7mtrS0uXLgAX19fdOzYERKJBNWrV0enTp0gFoshEomwd+9ejBs3DvXr10edOnWwevVqtG3b9p19i0Qi/P7771i5ciVWrFiB0aNHQ1tbG25ubjh37pzcjcSLo62tjTNnzmDx4sX4/vvv5W4kfvnyZbkFPADQvn171KpVC59++ikkEgn69euHuXPnAgAMDQ3xzz//YOXKlUhLS0P16tWxbNkydO5c+IN22LBh0NXVxU8//YQpU6ZAT08PDRo0wIQJE0p1nv369cOSJUswe/Zsude++OILTJw4EWPGjIFEIkHXrl0xa9YsWVwA0KtXLxw8eBDt2rVDSkoKfvnlFwwaNEiuH11dXRw/fhzjx4/HJ598Al1dXfTq1atUCa5Q3LkXhiFjfWVfL1mzCQDQvbM7Fs2cpKqwPnpX1x+Fho4WOvoNgZahLp5cD8WBb5Yg/7UqlnE1S+iYGsi+1jUzROcVo6BnaYyc55lIuPcI+79Zgof/ftzJiLKerDsMsa42av40CuqGeki7eg/BXy+Qqxhq17CGxmuf9bO/LkLDzAjVpvaFhoUxMu5GIvjrhchNLKxk6TVwkK3IbnL5Z7n3C/hklKzCGD52NWrMGwynX7+HtECKtMt3C9877+O+yXPykQtQNzOC7eR+0LAwQWZwJMK+mYe8F5+fVhULoODVXP2MgPuIHLMcVaZ6o4pvf0ginyJi2A/Ivv9qSkfsz4cg1tVGjR9HQ81QD+nXQhDaf77sOkpzcmHavTVsffpCrKUOSXQ84jYfQdzmP2V9mH35OdR0tWEz9kvYjP1Stv/5pTu43/vVwk16fz4+Phg4cCCaNm2KZs2aYeXKlcjIyMDgwYMBAAMGDECVKlWKzKPcunWr3OLgl9LT0zFv3jz06tUL1tbWiIiIwNSpU+Ho6AgPD49yOw+RVNnVKESlkJv4QNUh0AurGs9+dyOqMC1zs97diCqEptrHnbR+SJo+Plzu75H1v9Vl0o9O53EKH7N27Vr89NNPiI2NhYuLC1avXi0bZWzbti1q1Kghd6u8+/fvo27dujhx4kSRKXpZWVnw8vJCYGAgUlJSYGtri44dO2LBggXlOpLIBJIqBBPIyoMJZOXCBLLyYAJZeVRIAvn3yjLpR6frhDLp50NTbnMgiYiIiOjjxKWJREREJDzv8RhCeoUJJBEREQlPGT2JRqiYQBIREZHwsAKpFM6BJCIiIiKFsAJJREREwsMhbKUwgSQiIiLh4RC2UjiETUREREQKYQWSiIiIhIdD2EphAklERETCwwRSKRzCJiIiIiKFsAJJREREwiOVqjqCDxoTSCIiIhIeDmErhUPYRERERKQQViCJiIhIeFiBVAoTSCIiIhIe3khcKUwgiYiISHhYgVQK50ASERERkUJYgSQiIiLh4W18lMIEkoiIiISHQ9hK4RA2ERERESmEFUgiIiISHlYglcIEkoiIiISHt/FRCoewiYiIiEghrEASERGR4EgLuApbGUwgiYiISHg4B1IpHMImIiIiIoWwAklERETCw0U0SmECSURERMLDOZBKYQJJREREwsM5kErhHEgiIiIiUggrkERERCQ8rEAqhQkkERERCY+UcyCVwSFsIiIiIlIIK5BEREQkPBzCVgorkERERCQ8BdKy2d7DunXrUKNGDWhra6N58+a4evVqiW23b98OkUgkt2lra8u1kUqlmD17NmxsbKCjowN3d3eEhYW9V2ylxQSSiIiIqILs27cPPj4+mDNnDm7cuIFGjRrBw8MD8fHxJR5jaGiImJgY2fbw4UO515csWYLVq1djw4YNuHLlCvT09ODh4YHs7OxyOw8mkERERCQ80oKy2RS0fPlyDB8+HIMHD4azszM2bNgAXV1dbNu2rcRjRCIRrK2tZZuVldWr05BKsXLlSsycORPdu3dHw4YNsXPnTjx9+hSHDx9+n0+mVJhAEhERkfCoYAg7JycHAQEBcHd3l+0Ti8Vwd3fHpUuXSjwuPT0d1atXh52dHbp37467d+/KXouMjERsbKxcn0ZGRmjevPlb+1QWE0giIiKi9ySRSJCWlia3SSSSYtsmJiYiPz9froIIAFZWVoiNjS32mDp16mDbtm34888/sWvXLhQUFKBly5Z4/PgxAMiOU6TPssBV2FQhVjWereoQ6IXxN+arOgR6zeX6U1UdAr2QX8CaipBIy2gVtp+fH+bNmye3b86cOZg7d26Z9O/m5gY3NzfZ1y1btoSTkxM2btyIBQsWlMl7vA8mkERERCQ877mC+k3Tp0+Hj4+P3D4tLa1i25qbm0NNTQ1xcXFy++Pi4mBtbV2q99PQ0ICrqyvCw8MBQHZcXFwcbGxs5Pp0cXEp7WkojH9uERERkfCU0SIaLS0tGBoaym0lJZCamppo0qQJTp8+LdtXUFCA06dPy1UZ3yY/Px+3b9+WJYv29vawtraW6zMtLQ1XrlwpdZ/vgxVIIiIiogri4+ODgQMHomnTpmjWrBlWrlyJjIwMDB48GAAwYMAAVKlSBX5+fgCA+fPno0WLFnB0dERKSgp++uknPHz4EMOGDQNQuEJ7woQJWLhwIWrVqgV7e3vMmjULtra28PLyKrfzYAJJREREwlNGQ9iK6tOnDxISEjB79mzExsbCxcUF/v7+skUw0dHREItfDRAnJydj+PDhiI2NhYmJCZo0aYKLFy/C2dlZ1mbq1KnIyMjAiBEjkJKSgtatW8Pf37/IDcfLkkgq5dPEqfwtrdZf1SHQC1xEU7lwEU3loSnmo+0qi+ZPD5b7e2TM7Vcm/ejN/a1M+vnQcA4kERERESmEQ9hEREQkPCoawv5YMIEkIiIi4XmPxxDSKxzCJiIiIiKFsAJJREREwsMhbKUwgSQiIiLBKatHGQoVh7CJiIiISCGsQBIREZHwcAhbKUwgiYiISHiYQCqFCSQREREJD2/joxTOgSQiIiIihbACSURERMLDIWylMIEkIiIiwZEygVQKh7CJiIiISCGsQBIREZHwsAKpFCaQREREJDx8Eo1SOIRNRERERAphBZKIiIiEh0PYSmECSURERMLDBFIpHMImIiIiIoWwAklERESCI5WyAqkMJpBEREQkPBzCVgoTSCIiIhIeJpBK4RxIIiIiIlIIK5BEREQkOHwWtnKYQBIREZHwMIFUCoewiYiIiEghrEASERGR8PBR2EphAklERESCwzmQyuEQNhEREREphBVIIiIiEh5WIJXCBJKIiIiEh3MglcIhbCIiIiJSCBPIChIVFQWRSISgoKAPPoa2bdtiwoQJZRITERGRKkgLpGWyCdUHN4Q9aNAgpKSk4PDhw6oOpdxERUXB3t5e9rWpqSmaNGmCH3/8Ea6uru/dr52dHWJiYmBubl4WYX60Wvn0QoOv20HLUBdPr4fi5Pe/ICUqrsT2jfq3h8s37WFY1QIA8Cz0MS6tOoTIc7cqKmTBuB50G7/s2Y/ge+FIeJaEVX6z0P7TlqoO66NQfWofWHu7Q81QF2nX7iPcdxOyI2PfeozN4E6oOvoLaFoYIz34ISJmbEV6YLjsdcclI2D8aUNoWpmgIDMbaddCEbnwV2SFP5W1MW7dANV9+0LXqRoKMrMR9/t5RPntAfI5vvhSlSl9Yfl1B6gb6uL59XuInLYJksiYtx5jNagTbL71goaFMTKDoxA1cwsygl5dGwvvDjDv0QZ6DRygZqCL63X7Iz8ts7xPpXLhf2JKYQVSQbm5uRX2XqdOnUJMTAyOHz+O9PR0dO7cGSkpKe/VV05ODtTU1GBtbQ119Q/u74YK0+zbbnAd3BEnp2/D7i/mIDdTgi93+UJNS6PEY57HJuGfH/bh164zsavbLERfDIbXFh+Y1a5SgZELQ1ZWNuo4OmDGpNGqDuWjUnWMF2yHdkHY1E0I6vI9CjIlqL93FkRv+e/evHtLOMwdiOhlfyCw41Rk3I1C/d9mQsPcUNYm/dYDhE5Yh4BPJ+B234WACKi/dxYgLvzVo+dcHfV2f4/ks4EIdJ+CeyNXwKxjU9jP7F/u5/yhsPmuB6yHdEXUtA24020aCjIlqLvn7dfG9ItWqDZnMB4v/x13PCYjMzgKdffMhrqZkayNmo4WUs4F4smaAxVxGpUSK5DK+aATyLZt22LcuHGYOnUqTE1NYW1tjblz58pel0qlmDt3LqpVqwYtLS3Y2tpi3LhxstdFIlGRSqaxsTG2b98O4NWQ7759+/DZZ59BW1sbu3fvBgBs2bIFTk5O0NbWRt26dfHzzz/L9XP16lW4urpCW1sbTZs2RWBgoMLnZ2ZmBmtrazRt2hRLly5FXFwcrly5AgD477//0KZNG+jo6MDOzg7jxo1DRkaG7NgaNWpgwYIFGDBgAAwNDTFixIhih7DPnz+PZs2aQUtLCzY2Npg2bRry8vJkr2dkZGDAgAHQ19eHjY0Nli1bpvB5fEgaD+2Ey2v+RMTJG0i89wjHJm6AvqUxHDs2KfGYB6cCEXn2JlKi4pAcGYv/fvoDOZnZsHF1rMDIhaGN2ycYN2Ig3D9rpepQPipVhndF9MoDSDp+DZkhD3F/7BpoWZnAvFOzko8Z6YnY3acQt/csMkMfI3zqJhRkSWDV93NZm9hdp5B2OQSSRwnIuB2JqB/2QruqBbTtCqv15t1bISPkIaKX70d2VCxSLwUjcsGvsBnkATU97XI/7w+B9bBueLJqP5KPX0NWyENEjFsNTStTmLzl2tiM8ET8npNI3HcGWWGPEem7EQVZElj0e+3abDmKmLWHkB4QWhGnQW9Yt24datSoAW1tbTRv3hxXr14tse3mzZvRpk0bmJiYwMTEBO7u7kXaDxo0CCKRSG7r1KlTuZ7DB51AAsCOHTugp6eHK1euYMmSJZg/fz5OnjwJADhw4ABWrFiBjRs3IiwsDIcPH0aDBg0Ufo9p06Zh/PjxCAkJgYeHB3bv3o3Zs2dj0aJFCAkJweLFizFr1izs2LEDAJCeno5u3brB2dkZAQEBmDt3LiZPnqzUeero6AAorCRGRESgU6dO6NWrF27duoV9+/bhv//+w5gxY+SOWbp0KRo1aoTAwEDMmjWrSJ9PnjxBly5d8Mknn+DmzZtYv349tm7dioULF8raTJkyBefPn8eff/6JEydO4Ny5c7hx44ZS51JZGVWzgL6lMR7+d0e2L+d5FmKCImDbpFap+hCJRajj2QIaOlqIuRFWXqESlRntapbQtDJByj+vplzkP8/E88AwGDStXewxIg11GDR0kDsGUilS/r0Nw6Z1ij1GrKsF677tkPUwDpKnzwr3aamjIFt+VKcgOwdqOlrQb1RTyTP78GlVs4KmlQnS/r0p25f/PBPpgWEwaFL85yzSUIdew5pI+1f+2qT+e6vEYwSroIw2Be3btw8+Pj6YM2cObty4gUaNGsHDwwPx8fHFtj937hz69euHs2fP4tKlS7Czs0PHjh3x5MkTuXadOnVCTEyMbPvtt98UD04BH/xYZsOGDTFnzhwAQK1atbB27VqcPn0aHTp0QHR0NKytreHu7g4NDQ1Uq1YNzZqV/FdbSSZMmICePXvKvp4zZw6WLVsm22dvb4/g4GBs3LgRAwcOxJ49e1BQUICtW7dCW1sb9erVw+PHj/Htt9++1zmmpKRgwYIF0NfXR7NmzTBjxgx4e3vLFrLUqlULq1evxmeffYb169dDW7vwL/fPP/8ckyZNkvUTFRUl1+/PP/8MOzs7rF27FiKRCHXr1sXTp0/h6+uL2bNnIzMzE1u3bsWuXbvQvn17AIUJe9WqVd/rPCo7PQtjAEBmYprc/szENOhZGBVzxCvmdari68Nzoa6lgZyMbPw5YiWehT196zFElYGGpQkAICchRW5/TkIqNC2Niz/G1AAidTXkJKS+cUwKdBzlp27YDPKA/az+UNPTQWbYE9z5aj6kuYWjHMlnb6LK8K6w8GqFhL8uQdPSGNV8egNAie8tJBovPoPcNz7n3IQU2XV7k/qLa5P7xvXMTSx6bYROqqI5kMuXL8fw4cMxePBgAMCGDRvw999/Y9u2bZg2bVqR9i9HPl/asmULDhw4gNOnT2PAgAGy/VpaWrC2ti7f4F/zUSSQr7OxsZFl8b1798bKlSvh4OCATp06oUuXLvD09FR4DmDTpk1l/87IyEBERASGDh2K4cOHy/bn5eXByKgwyQgJCUHDhg1liRwAuLm5KXxuLVu2hFgsRkZGBhwcHLBv3z5YWVnh5s2buHXrltx/VFKpFAUFBYiMjISTk1ORuIsTEhICNzc3iEQi2b5WrVohPT0djx8/RnJyMnJyctC8eXPZ66ampqhT5+1/xUokEkgkErl9edJ8qIvUSn3uFcHJqyU6+A2RfX1w0NL37ivpQQx2dpoBLUMd1O7SDJ2Xj8S+rxYyiaRKx6JnG9T6aYTs67v9/cr1/eIP/Ivk8zehaWWCqt9+gbqbfHDzi5mQSnKRcv4mIuf/CsclI1Bn7TgU5OQievl+GLk5QyoV3twysx6fwn7JSNnX979ZpMJoqLSK+52npaUFLS2tIm1zcnIQEBCA6dOny/aJxWK4u7vj0qVLpXq/zMxM5ObmwtTUVG7/uXPnYGlpCRMTE3z++edYuHAhzMzM3uOMSueDTyA1NOQnEotEIhQUFP5ZYWdnh/v37+PUqVM4efIkRo8ejZ9++gnnz5+HhoYGRCJRkR9SxS2S0dPTk/07PT0dQOGchNcTKwBQUyvbBGnfvn1wdnaGmZkZjI2N5WIYOXKk3HzOl6pVq1Zs3BXJz88P8+bNk9vXwbABOho1LOEI1Qg/eQMxgRGyr9W0Cr8ddM0NkRGfItuva26I+ODot/ZVkJuPlIeFK7XjbkfBupEDGg/phJPTt5V94ERKSDp+DTdem14hfvHfvaaFMXJf++9e08II6Xeiiu0jN+k5pHn50HyjMv9mH0DhkGv+80xkR8YiJCAMbve3w7xzMyQcvgAAeLLxKJ5sPApNKxPkpWZAy84C9jP7I/thyXc++Fgln7iK9MBXcxLFmoW/3zQsjJAbnyzbr2FhjMy7kcX2kffi2mi8GFGRHWNuXKQqKXhlVIEs7nfenDlz5NZkvJSYmIj8/HxYWVnJ7beyssK9e/dK9X6+vr6wtbWFu7u7bF+nTp3Qs2dP2NvbIyIiAt9//z06d+6MS5culXlu8tIHn0C+i46ODjw9PeHp6YnvvvsOdevWxe3bt9G4cWNYWFggJubVrRDCwsKQmfn22xhYWVnB1tYWDx48gLe3d7FtnJyc8OuvvyI7O1tWhbx8+bLCsdvZ2aFmzaLzgBo3bozg4GA4Oiq3SMPJyQkHDhyAVCqVVSEvXLgAAwMDVK1aFaamptDQ0MCVK1dkiWlycjJCQ0Px2Wefldjv9OnT4ePjI7fv53ojS2itOrkZ2UjJyJbblx6fguqt6iHhRcKoqa8DG5eaCPr1tEJ9i0QiqGl+9N9e9AHKz8hGfob87Xly4pJh3KYBMu5GAQDU9HVg4FoLMdtPFNuHNDcPz289gHGbBnjmf61wp0gE49YN8HTb/0p+c1Hh/xS3gjgnrjBBsuzRGtmPE5B+q/gE6WNWkJENSTHXxrB1Q2S+dm30XWshbqd/sX1Ic/OQcSsChq0bItn/xUILkQhGrRsidvux8gz/g1NWQ9jF/c4rrvpYFn744Qfs3bsX586dkxvl7Nu3r+zfDRo0QMOGDVGzZk2cO3dONgWtrH3Uv+G2b9+O/Px8NG/eHLq6uti1axd0dHRQvXp1AIVzBNeuXQs3Nzfk5+fD19e3SEWzOPPmzcO4ceNgZGSETp06QSKR4Pr160hOToaPjw++/vprzJgxA8OHD8f06dMRFRWFpUvff3j0Tb6+vmjRogXGjBmDYcOGQU9PD8HBwTh58iTWrl1b6n5Gjx6NlStXYuzYsRgzZgzu37+POXPmwMfHB2KxGPr6+hg6dCimTJkCMzMzWFpaYsaMGRCL3772qrjSfWUbvi7Jja3+aDHOC8lRcUiNjkeryV8iPT4F4ScCZG16/zYd4f7XEbijcLFWG9+vEHn2JtKePoOmnjacvFrCzs0J+79ZoqrT+GhlZmYh+vGraQFPnsbhXmgEjAwNYGNtqcLIPmxPNv8Nuwm9kPUgBtnR8aju2xeSuGQk+r9a6dngjzlI/N8VxGwrTFyebDyCOqvG4PnNCDwPDEeV4V0h1tVC3N6zAAoX55h3b4WU8zeR+ywNmjZmsBvrhYLsHCSffrUQr8roL5B8JgiQFsCsSwtUHeOFeyOWAwW8SR9QuFq6yvgvkR0ZA0l0HKpO7YecuKRXySGAuvvmItn/CuJ+KUzeYzYdQc2VY5FxMxzpgWGwHu4Jsa4WEvaekR2jYWEMDUtjaNvbAAB061ZHfkYWJE8SkZ+SXrEn+YErabi6OObm5lBTU0NcnHyFPS4u7p3zF5cuXYoffvgBp06dKjJ9700ODg4wNzdHeHg4E8j3YWxsjB9++AE+Pj7Iz89HgwYNcOTIEdmcgGXLlmHw4MFo06YNbG1tsWrVKgQEBLyjV2DYsGHQ1dXFTz/9hClTpkBPTw8NGjSQLWrR19fHkSNHMGrUKLi6usLZ2Rk//vgjevXqVSbn1bBhQ5w/fx4zZsxAmzZtIJVKUbNmTfTp00ehfqpUqYJjx45hypQpaNSoEUxNTTF06FDMnDlT1uann35Ceno6PD09YWBggEmTJiE1NfUtvX7Yrq4/Cg0dLXT0GwItQ108uR6KA98sQb7k1dQG42qW0DE1kH2ta2aIzitGQc/SGDnPM5Fw7xH2f7MED/+9U9xbkBLu3AvDkLG+sq+XrNkEAOje2R2LZk4q6TB6h8drD0NNVwu1lo6EuqEeUq/ew91+CyF97b977RpW0DB9dY/HxD8vQsPMENWn9i28kfjdKNzttwi5iYU/HwokuTBq4YQqI7pC3UgPuQmpSL0cgpueM5D72kI1089dUW18L4g01ZER/BDBg5Yg+Yzitz37WMWsOwSxrhbsl4yCuqEenl8LwX3vBW9cG2uov3Ztkv66AA0zQ1Sd0k823H3PewHyEl/97LYc4IGqk179znA+XDjfMmLCGiT+frYCzqwSUMHfKJqammjSpAlOnz4NLy+vwjAKCnD69Okid1J53ZIlS7Bo0SIcP378nesbAODx48d49uwZbGxsyir0IkRSIc5Upgq3tBpvDFxZjL8xX9Uh0Gsu15+q6hDoBU0xq56VRfOnB8v9PRI6lDwVSxEWJ88r1H7fvn0YOHAgNm7ciGbNmmHlypX4/fffce/ePVhZWWHAgAGoUqUK/PwKF7j9+OOPmD17Nvbs2YNWrV7dA1dfXx/6+vpIT0/HvHnz0KtXL1hbWyMiIgJTp07F8+fPcfv27XIbTv+oK5BERERElUmfPn2QkJCA2bNnIzY2Fi4uLvD395ctrImOjpabKrZ+/Xrk5OTgyy+/lOvn5UIdNTU13Lp1Czt27EBKSgpsbW3RsWNHLFiwoNySR4AJpEqMGjUKu3btKva1/v37Y8OGDRUcERERkbCo6j6QADBmzJgSh6zPnTsn9/Wb93B+k46ODo4fP15GkZUeE0gVmD9/folPpjE0NCx2PxEREZUdVSaQHwMmkCpgaWkJS0uuGCUiIlIZqejdbahEH/yzsImIiIioYrECSURERILDIWzlMIEkIiIiwZEWcAhbGRzCJiIiIiKFsAJJREREgsMhbOUwgSQiIiLBkXIVtlI4hE1ERERECmEFkoiIiASHQ9jKYQJJREREgsNV2MrhEDYRERERKYQVSCIiIhIcqVTVEXzYmEASERGR4HAIWzlMIImIiEhwmEAqh3MgiYiIiEghrEASERGR4HAOpHKYQBIREZHgcAhbORzCJiIiIiKFsAJJREREgsNnYSuHCSQREREJDh9lqBwOYRMRERGRQliBJCIiIsEp4BC2UphAEhERkeBwDqRyOIRNRERERAphBZKIiIgEh/eBVA4TSCIiIhIcPolGOUwgiYiISHBYgVQO50ASERERkUJYgSQiIiLB4W18lMMEkoiIiASHt/FRDoewiYiIiEghrEASERGR4HAVtnKYQBIREZHgcA6kcjiETUREREQKYQJJREREgiOVispkex/r1q1DjRo1oK2tjebNm+Pq1atvbf/HH3+gbt260NbWRoMGDXDs2LE3zkWK2bNnw8bGBjo6OnB3d0dYWNh7xVZaTCCJiIhIcKTSstkUtW/fPvj4+GDOnDm4ceMGGjVqBA8PD8THxxfb/uLFi+jXrx+GDh2KwMBAeHl5wcvLC3fu3JG1WbJkCVavXo0NGzbgypUr0NPTg4eHB7Kzs9/343knkVTKaaRU/pZW66/qEOiF8TfmqzoEes3l+lNVHQK9oCkuUHUI9ELzpwfL/T1u2HUvk34aP/pTofbNmzfHJ598grVr1wIACgoKYGdnh7Fjx2LatGlF2vfp0wcZGRk4evSobF+LFi3g4uKCDRs2QCqVwtbWFpMmTcLkyZMBAKmpqbCyssL27dvRt29fJc6uZKxAEhERkeAUSEVlskkkEqSlpcltEomk2PfMyclBQEAA3N3dZfvEYjHc3d1x6dKlYo+5dOmSXHsA8PDwkLWPjIxEbGysXBsjIyM0b968xD7LAldhU4VomZul6hDoBVa8KpcWd5aoOgR6IdxtjKpDoApUVjcS9/Pzw7x58+T2zZkzB3Pnzi3SNjExEfn5+bCyspLbb2VlhXv37hXbf2xsbLHtY2NjZa+/3FdSm/LABJKIiIgEp6xu4zN9+nT4+PjI7dPS0iqTviszJpBERERE70lLS6vUCaO5uTnU1NQQFxcntz8uLg7W1tbFHmNtbf3W9i//f1xcHGxsbOTauLi4lPY0FMY5kERERCQ40jLaFKGpqYkmTZrg9OnTsn0FBQU4ffo03Nzcij3Gzc1Nrj0AnDx5Utbe3t4e1tbWcm3S0tJw5cqVEvssC6xAEhERkeCo6kk0Pj4+GDhwIJo2bYpmzZph5cqVyMjIwODBgwEAAwYMQJUqVeDn5wcAGD9+PD777DMsW7YMXbt2xd69e3H9+nVs2rQJACASiTBhwgQsXLgQtWrVgr29PWbNmgVbW1t4eXmV23kwgSQiIiKqIH369EFCQgJmz56N2NhYuLi4wN/fX7YIJjo6GmLxqwHili1bYs+ePZg5cya+//571KpVC4cPH0b9+vVlbaZOnYqMjAyMGDECKSkpaN26Nfz9/aGtrV1u58H7QFKFuGjTS9Uh0Av5fP5rpcJV2JUHV2FXHk5hx97dSEkXrL8sk35axe4vk34+NKxAEhERkeDwtvHK4SIaIiIiIlIIK5BEREQkOFJwOo8ymEASERGR4BRwBYhSOIRNRERERAphBZKIiIgEp4BD2EphAklERESCwzmQymECSURERILD2/goh3MgiYiIiEghrEASERGR4HAIWzlMIImIiEhwOIStHA5hExEREZFCWIEkIiIiwWEFUjlMIImIiEhwOAdSORzCJiIiIiKFsAJJREREglPAAqRSmEASERGR4PBRhsrhEDYRERERKYQVSCIiIhIcqaoD+MAxgSQiIiLB4W18lMMEkoiIiASnQMQ5kMrgHEgiIiIiUggrkERERCQ4nAOpHCaQREREJDicA6kcDmETERERkUJYgSQiIiLB4ZNolMMEkoiIiASHT6JRDoewiYiIiEghrEASERGR4HAVtnKYQBIREZHgcA6kcjiETUREREQKYQWSiIiIBIf3gVQOK5AVLCoqCiKRCEFBQSW2OXfuHEQiEVJSUso9nkGDBsHLy6vU7UsTPxERUWUnLaNNqFiBLMagQYOwY8cOAIC6ujqqVq2K3r17Y/78+dDW1laqbzs7O8TExMDc3LwsQi21qKgo2NvbIzAwEC4uLrL9q1atglQqnG8Buyl9YeXtDjVDXTy/dh8Ppm1CdmTMW4+xHtQJtqO7Q9PCGBnBUYicsRXpQeEAAHVjfdhN7gPjzxpBs4o58pLSkPS/q4heshf5zzPl+rH4qh1sR3pCx8EG+elZSDxyEZHfbym3c63sqk/tA+sX1yLt2n2E+25CdmTsW4+xGdwJVUd/AU0LY6QHP0TEjK1IDwyXve64ZASMP20ITSsTFGRmI+1aKCIX/oqs8KeyNsatG6C6b1/oOlVDQWY24n4/jyi/PUA+6xGKuB50G7/s2Y/ge+FIeJaEVX6z0P7TlqoO64Nn4t0NpsN6Qd3CBJJ7kYidvx7Zt0JLbG/QqTUsJnwDjapWyIl6iviftiHj/HW5Npo17WA5ZTB0mzWASE0NkvBoPB6zCHkxCdCoYgnHc9uL7fvx2MV47v9fWZ5epcI5kMphBbIEnTp1QkxMDB48eIAVK1Zg48aNmDNnjtL9qqmpwdraGurqlSN3NzIygrGxsarDqBBVvvOCzdAuiPDdiNtdp6MgMxvOv82CSEujxGPMvmiJGnMH4fGy33HTYwoygh/C+bdZ0DAzBABoWplA09oUUfN3IqjdRISNXwvjdq5wXD5arh+bkZ6oNq0fnqw9iMC2E3D3q3lIORdUnqdbqVUd4wXboV0QNnUTgrp8j4JMCervffu1MO/eEg5zByJ62R8I7DgVGXejUP+3mdAwN5S1Sb/1AKET1iHg0wm43XchIALq750FiAt/1Ok5V0e93d8j+WwgAt2n4N7IFTDr2BT2M/uX+zl/bLKyslHH0QEzJo1+d2MqFYMun8Ly++FIXLsHkV5jkR3yANW2LYCaqVGx7XVcnVBlhS9S9p9AZPexSD91CXY/z4JWreqyNhrVrFH9t5+Q8+AxHvb3xQPP0Uhc9xukkhwAQG5MIkLdvOW2hFW/Ij89E+n/XC/2fYkAJpAl0tLSgrW1Nezs7ODl5QV3d3ecPHkSAFBQUAA/Pz/Y29tDR0cHjRo1wv79+2XHJicnw9vbGxYWFtDR0UGtWrXwyy+/ACh+CPjYsWOoXbs2dHR00K5dO0RFRRWJ57///kObNm2go6MDOzs7jBs3DhkZGbLXa9SogcWLF2PIkCEwMDBAtWrVsGnTJtnr9vb2AABXV1eIRCK0bdsWQNEhbH9/f7Ru3RrGxsYwMzNDt27dEBERoezHWSnYDO+Gxyv3I/n4NWSGPETYuDXQtDKBaadmJR5jO9ITcbtPIX7fWWSFPsaDqRuRnyWBZb/2AIDM+49wf9hPSD55HZKHcUi7cAfRP+yBSYemgFrht5eakR6q+fZD+Lg1SDz0HyQP45AZ8hDJJ4T7w7nK8K6IXnkASS+uxf2xa6BlZQLzt1yLKiM9Ebv7FOL2nkVm6GOET92EgiwJrPp+LmsTu+sU0i6HQPIoARm3IxH1w15oV7WAtp0FAMC8eytkhDxE9PL9yI6KReqlYEQu+BU2gzygpqfc6ILQtHH7BONGDIT7Z61UHcpHw2xID6Ts80fqgZPICX+E2NlrUZAlgfGXHYttbzqwO9L/DUDSlgPIiXiEhJW/Ijs4AibfeMraWEwciIzz1xG/ZBskwQ+QGx2L9DNXkJ+UWtigoAD5iclym0GHlnj+v38hzcyuiNNWmYIy2spLUlISvL29YWhoCGNjYwwdOhTp6elvbT927FjUqVMHOjo6qFatGsaNG4fU1FS5diKRqMi2d+9eheNjAlkKd+7cwcWLF6GpqQkA8PPzw86dO7FhwwbcvXsXEydORP/+/XH+/HkAwKxZsxAcHIz//e9/CAkJwfr160scsn706BF69uwJT09PBAUFYdiwYZg2bZpcm4iICHTq1Am9evXCrVu3sG/fPvz3338YM2aMXLtly5ahadOmCAwMxOjRo/Htt9/i/v37AICrV68CAE6dOoWYmBgcPHiw2HgyMjLg4+OD69ev4/Tp0xCLxejRowcKCj7s4T2talbQtDJByr+3ZPvyn2fieWAYDJrWKfYYkYY69BvWROprx0AqReq/t2DQpHaJ76VmqIv89EzZkKjxp40gEomgaWMKl39WoUnAJtTeOAmatmZlc3IfGO1qloXX4p/irkXxn6tIQx0GDR3kjoFUipR/b8OwhOsn1tWCdd92yHoYB8nTZ4X7tNRRkJ0r164gOwdqOlrQb1RTyTMjUoKGOrTrOSLjYtCrfVIpMi4GQce1brGH6LjWRcbFQLl96f8GQMflRXuRCPptP0FO1BPYbVuAWpf3oMb+FdB3dysxDO16jtB2romUP04oe0aVXmVPIL29vXH37l2cPHkSR48exT///IMRI0aU2P7p06d4+vQpli5dijt37mD79u3w9/fH0KFDi7T95ZdfEBMTI9sUWQvxUuUYR62Ejh49Cn19feTl5UEikUAsFmPt2rWQSCRYvHgxTp06BTe3wm9CBwcH/Pfff9i4cSM+++wzREdHw9XVFU2bNgVQWB0syfr161GzZk0sW7YMAFCnTh3cvn0bP/74o6yNn58fvL29MWHCBABArVq1sHr1anz22WdYv369bF5mly5dMHp04XCSr68vVqxYgbNnz6JOnTqwsCiswJiZmcHa2rrEeHr16iX39bZt22BhYYHg4GDUr19fgU+wctG0NAYA5CakyO3PTUiFpoVxsceomxpApK6GnGKO0XGsUuIxdhN7I27XKdk+7epWgFiEKuN6IWrWNuSlZaCa79dw3jcHNz/3gTQ3731P64OkYWkCAEU+15yEVNl1KnKM7FqkvnFMSpFrYTPIA/az+kNNTweZYU9w56v5ss84+exNVBneFRZerZDw1yVoWhqjmk9vACjxvYkqgrqJIUTqashPTJbbn/8sBVo17Yo/xtwE+Ykp8u0TU6BuUfg9pmZmDDV9XZiN6I2EFTsR/9Mv0G/TBFXXzUD0N9OQefVOkT6Ne3eEJDwaWYEhZXNi9F5CQkLg7++Pa9euyXKJNWvWoEuXLli6dClsbW2LHFO/fn0cOHBA9nXNmjWxaNEi9O/fH3l5eXJT54yNjd+aC5QGK5AlaNeuHYKCgnDlyhUMHDgQgwcPRq9evRAeHo7MzEx06NAB+vr6sm3nzp2yod5vv/0We/fuhYuLC6ZOnYqLFy+W+D4hISFo3ry53L6XielLN2/exPbt2+Xez8PDAwUFBYiMjJS1a9iwoezfIpEI1tbWiI+PV+i8w8LC0K9fPzg4OMDQ0FCW/EZHR5e6D4lEgrS0NLktR5qvUBzKMu/ZBs3Dd8k2kYZaub+nmr4OnH79Hpmhj/Bo6b5XL4hFEGtqIHLmVqScC0L6jTCEjl4BHXtrGLX6cJPy0rLo2QYtI36VbeJyvhbxB/7FDfcpuOk1C1kPnqLuJh/Z3MqU8zcROf9XOC4ZgdbRv6HpxdVIOn0DAAS1mIyEQSQuXCXy/PRlJG0/DEnIAzzb9AfSz16Fcb8uRdtracLQsy1S/jhe0aGqhFRUNltxv/MkEolSsV26dAnGxsay5BEA3N3dIRaLceXKlVL3k5qaCkNDwyLrLr777juYm5ujWbNm2LZt23v9/GMFsgR6enpwdHQEUFiFa9SoEbZu3Sqrwv3999+oUkW+8qGlpQUA6Ny5Mx4+fIhjx47h5MmTaN++Pb777jssXbr0vWJJT0/HyJEjMW7cuCKvVatWTfZvDQ35BQgikUjhoWdPT09Ur14dmzdvhq2tLQoKClC/fn3k5OSUug8/Pz/MmzdPbt8QvboYauCsUCzKSDp+Dek3wmRfizQLPxsNC2PkxqfI9mtYGCHjblSxfeQlPYc0L79IhVLDwkiuDwAQ62nDac9M5Kdn496QJZDmvUqYc+MKKwpZoY9e9f0sDblJz6FZpWJX46tC0vFruPHatRBrFf7Y0XzjWmhaGCH9TlSxfeTKroX8YoI3+wAKh8Pzn2ciOzIWIQFhcLu/HeadmyHh8AUAwJONR/Fk41FoWpkgLzUDWnYWsJ/ZH9kP45Q/WaL3lJecBmlePtTMTeT2q5kZIy8hqfhjEpOhZm4s397cGHkJya/6zM2DJFy+ACCJeATdJvWK9GfQqTXE2lpIPXxaiTP5cJTV8HNxv/PmzJmDuXPnvnefsbGxsLS0lNunrq4OU1NTxMa+/W4VLyUmJmLBggVFhr3nz5+Pzz//HLq6ujhx4gRGjx6N9PT0YnOMt2EFshTEYjG+//57zJw5E87OztDS0kJ0dDQcHR3lNju7V8MMFhYWGDhwIHbt2oWVK1fKLWh5nZOTk2x+4kuXL1+W+7px48YIDg4u8n6Ojo6yeZnv8rJdfn7JlcBnz57h/v37mDlzJtq3bw8nJyckJyeX2L4k06dPR2pqqtz2jX7x89TKS0FGNrKjYmVbVugj5MQlw7h1A1kbNX0dGLjWwvPr94vtQ5qbh/RbETB67RiIRDBq3RDPA17dVkNNXwf19s6GNDcP9wb5QSqRn2OXdu0eAECn5qs/ONSN9aFhagDJ44SyON1KLf+Na5F5/3HhtWhT3LUo/nYl0tw8PL/1QO4YiEQwbt0AaSVcv8I2hf9T3OrunLhkFGTnwLJHa2Q/TkD6rcgibYgqTG4esu+GQ8+t0at9IhH0WrogK/BesYdkBd6DnpuL3D69Vq7ICron6zPrdii07KvKtdGqUQW5T4uOThn37ojnZ64gPylNqVMRmuJ+502fPr3YttOmTSt2Ecvr2717xV9vRaSlpaFr165wdnYuksjOmjULrVq1gqurK3x9fTF16lT89NNPCr8HK5Cl1Lt3b0yZMgUbN27E5MmTMXHiRBQUFKB169ZITU3FhQsXYGhoiIEDB2L27Nlo0qQJ6tWrB4lEgqNHj8LJyanYfkeNGoVly5ZhypQpGDZsGAICArB9+3a5Nr6+vmjRogXGjBmDYcOGQU9PD8HBwTh58iTWrl1bqvgtLS2ho6MDf39/VK1aFdra2jAykq/mmJiYwMzMDJs2bYKNjQ2io6OLLOgpDS0tLVk19iVNUfkPIb9LzOajqDrhS2RFxkASHQ87337IiUtGkv+rBN759zlI+t9VxP7yPwDA041HUGvVWKTfjEB6UBhshneDmq4W4veeAVCY+DjvnQ2xjhZCx6yCmr4u1PR1AQC5z9KAggJkP4jBM/+rsF8wBBFTNiD/eSaqfd8fWeFPkXah6BwkIXiy+W/YTeiFrAcxyI6OR3XfvpDEJSPxtWvR4I85SPzfFcRs8y88ZuMR1Fk1Bs9vRuB5YDiqDO8Ksa4W4vaeBVC4OMe8eyuknL+J3Gdp0LQxg91YLxRk5yD5xTA1AFQZ/QWSzwQB0gKYdWmBqmO8cG/EcuADXyhW0TIzsxD9+NX9NZ88jcO90AgYGRrAxtryLUdSSZ5tOwTbJT7IvhOGrFuhMB3UHWIdLaQcKLwDiM2SSciLe4aEZdsBAEk7/kT13T/CdEgPpJ+7BsOun0Gnfi3Ezlwj6zNpywFUWTkNmdduI+PyLeh/2gT6nzfHw/6+cu+tUc0Gup/Ux6Nhyt+u7kNRVt/xxf3OK8mkSZMwaNCgt7ZxcHAodgpaXl4ekpKS3jl38fnz5+jUqRMMDAxw6NChIqOTb2revDkWLFgAiURS6vMAmECWmrq6OsaMGYMlS5YgMjISFhYW8PPzw4MHD2BsbIzGjRvj+++/B1BY7Zs+fTqioqKgo6ODNm3alLhEvlq1ajhw4AAmTpyINWvWoFmzZrLb8bzUsGFDnD9/HjNmzECbNm0glUpRs2ZN9OnTR6H4V69ejfnz52P27Nlo06YNzp07J9dGLBZj7969GDduHOrXr486depg9erVslv+fOierDsMsa42av40CuqGeki7eg/BXy+Qqxhq17CGhqmB7Otnf12EhpkRqk3tCw0LY2TcjUTw1wuRm1i4mEOvgYNsRXaTyz/LvV/AJ6NkFcbwsatRY95gOP36PaQFUqRdvlv43nkVOze0sni89jDUdLVQa+lIqBvqIfXqPdztt/CNa2EFDdNX93hM/PMiNMwMUX1q38Ibid+Nwt1+i2TXokCSC6MWTqgyoivUjfSQm5CK1MshuOk5A7mJryoqpp+7otr4XhBpqiMj+CGCBy1B8hn5laz0bnfuhWHI2FdJyJI1haMs3Tu7Y9HMSaoK64P2/Ng/iDc1hMX4b6BmYQJJyANED52N/GcpAAANWwtA+irtyQoMwROfJbCYOAAWkwYhJ+oJHo1eAEnYw1d9nryEmDlrYT7yK1jNGoWcyMd4PGYRsgKC5d7b+MuOyItNRMZ/NyAUqpj1bGFhIVvU+jZubm5ISUlBQEAAmjRpAgA4c+YMCgoKiqybeF1aWho8PDygpaWFv/76q1QPPwkKCoKJiYlCySMAiKScOU4V4KJNr3c3ogqRL+XjFyqTFneWqDoEeiHcbcy7G1GFcAo7Vu7vscaubB4gMPbRrjLp502dO3dGXFwcNmzYgNzcXAwePBhNmzbFnj17AABPnjxB+/btsXPnTjRr1gxpaWno2LEjMjMzcejQIejp6cn6srCwgJqaGo4cOYK4uDi0aNEC2traOHnyJCZPnozJkycXmcf5LqxAEhERkeBU9kcZ7t69G2PGjEH79u0hFovRq1cvrF69WvZ6bm4u7t+/j8zMwsfm3rhxQ7ZC++Ui4JciIyNRo0YNaGhoYN26dZg4cSKkUikcHR2xfPlyDB8+XOH4mEASERGR4FT2Wc+mpqayamNxatSoIXf7nbZt277zdjydOnVCp06dyiQ+JpBEREQkOJU9gazseBsfIiIiIlIIK5BEREQkOFxBrBwmkERERCQ4lX0RTWXHIWwiIiIiUggrkERERCQ4XESjHCaQREREJDicA6kcDmETERERkUJYgSQiIiLBKWANUilMIImIiEhwOAdSORzCJiIiIiKFsAJJREREgsMBbOUwgSQiIiLB4RC2cphAEhERkeDwSTTK4RxIIiIiIlIIK5BEREQkOLyNj3KYQBIREZHgMH1UDoewiYiIiEghrEASERGR4HAVtnKYQBIREZHgcA6kcjiETUREREQKYQWSiIiIBIf1R+UwgSQiIiLB4RxI5TCBJCIiIsHhHEjlcA4kERERESmEFUgiIiISHNYflcMEkoiIiASHcyCVwyFsIiIiIlIIK5BEREQkOFIOYiuFCSQREREJDoewlcMhbCIiIiJSCCuQREREJDi8D6RymEASERGR4DB9VA6HsImIiIhIIaxAEhERkeBwCFs5rEASERGR4BSU0VZekpKS4O3tDUNDQxgbG2Po0KFIT09/6zFt27aFSCSS20aNGiXXJjo6Gl27doWuri4sLS0xZcoU5OXlKRwfK5BEREQkOJX9PpDe3t6IiYnByZMnkZubi8GDB2PEiBHYs2fPW48bPnw45s+fL/taV1dX9u/8/Hx07doV1tbWuHjxImJiYjBgwABoaGhg8eLFCsXHBJKIiIioEgkJCYG/vz+uXbuGpk2bAgDWrFmDLl26YOnSpbC1tS3xWF1dXVhbWxf72okTJxAcHIxTp07BysoKLi4uWLBgAXx9fTF37lxoamqWOkYOYRMREZHglNUQtkQiQVpamtwmkUiUiu3SpUswNjaWJY8A4O7uDrFYjCtXrrz12N27d8Pc3Bz169fH9OnTkZmZKddvgwYNYGVlJdvn4eGBtLQ03L17V6EYWYGkCqGplq/qEOiF/AL+3ViZhLuNUXUI9ILjpbWqDoEqUFkNYfv5+WHevHly++bMmYO5c+e+d5+xsbGwtLSU26eurg5TU1PExsaWeNzXX3+N6tWrw9bWFrdu3YKvry/u37+PgwcPyvp9PXkEIPv6bf0WhwkkERER0XuaPn06fHx85PZpaWkV23batGn48ccf39pfSEjIe8cyYsQI2b8bNGgAGxsbtG/fHhEREahZs+Z791scJpBEREQkOGW1glpLS6vEhPFNkyZNwqBBg97axsHBAdbW1oiPj5fbn5eXh6SkpBLnNxanefPmAIDw8HDUrFkT1tbWuHr1qlybuLg4AFCoX4AJJBEREQlQgbTiV2FbWFjAwsLine3c3NyQkpKCgIAANGnSBABw5swZFBQUyJLC0ggKCgIA2NjYyPpdtGgR4uPjZUPkJ0+ehKGhIZydnRU6F06GIiIiIqpEnJyc0KlTJwwfPhxXr17FhQsXMGbMGPTt21e2AvvJkyeoW7eurKIYERGBBQsWICAgAFFRUfjrr78wYMAAfPrpp2jYsCEAoGPHjnB2dsY333yDmzdv4vjx45g5cya+++67UldRX2ICSURERIIjLaOtvOzevRt169ZF+/bt0aVLF7Ru3RqbNm2SvZ6bm4v79+/LVllramri1KlT6NixI+rWrYtJkyahV69eOHLkiOwYNTU1HD16FGpqanBzc0P//v0xYMAAuftGlpZIKlVBDZcE53pVL1WHQC9wFXblYqiXreoQ6AWuwq48NMwdyv09vq7eo0z62fPwUJn086HhbxIiIiIiUggX0RAREZHgVPZHGVZ2TCCJiIhIcMrqNj5CxQSSiIiIBKeAFUilcA4kERERESmEFUgiIiISHM6BVA4TSCIiIhIczoFUDoewiYiIiEghrEASERGR4PA5KsphAklERESCw1XYyuEQNhEREREphBVIIiIiEhwuolEOE0giIiISHN7GRzkcwiYiIiIihbACSURERILDRTTKYQJJREREgsPb+CiHCSQREREJDhfRKIdzIImIiIhIIaxAEhERkeBwFbZymEASERGR4HARjXI4hE1ERERECmEFkoiIiASHq7CVwwSSiIiIBIdD2MrhEDYRERERKYQVSCIiIhIcrsJWDhNIIiIiEpwCzoFUCoewiYiIiEghrEASERGR4LD+qBwmkERERCQ4XIWtHCaQREREJDhMIJXDOZBEREREpBBWIImIiEhw+CQa5TCBJCIiIsHhELZyVDKEvX37dhgbG8u+njt3LlxcXN56zKBBg+Dl5SX7um3btpgwYUK5xPc+oqKiIBKJEBQUpOpQ3urNz56IiIhIUQpVIAcNGoQdO3bIvjY1NcUnn3yCJUuWoGHDhqXup0+fPujSpYsib13EwYMHoaGhoVQfpdG2bVucP38eAKCpqQlzc3M0btwYgwcPRs+ePWXt7OzsEBMTA3Nz83KPSRll8dl/KCwGdob1qB7QsDBGZkgUHs3ajIygsBLbm3RtCdspX0OrqiWyo2LwZPFOpJ4JkGtjO7kfzPt1gLqRHtKv3cPD7zdAEhkje73BpU3QsrOUO+ax307ErjsIADBwqw/LYZ7Qc6kFNQNdSCJjELvhEJIO/VOGZ/7hqjKlLyy/7gB1Q108v34PkdM2yX2+xbEa1Ak233oVXufgKETN3IKMoHDZ6xbeHWDeow30GjhAzUAX1+v2R35aZnmfygfDxLsbTIf1grqFCST3IhE7fz2yb4WW2N6gU2tYTPgGGlWtkBP1FPE/bUPG+etybTRr2sFyymDoNmsAkZoaJOHReDxmEfJiEqBRxRKO57YX2/fjsYvx3P+/sjw9wbgedBu/7NmP4HvhSHiWhFV+s9D+05aqDqtS45NolKNwBbJTp06IiYlBTEwMTp8+DXV1dXTr1k2hPnR0dGBpafnuhm9hamoKAwMDpfooreHDhyMmJgYRERE4cOAAnJ2d0bdvX4wYMULWRk1NDdbW1lBXr9yzAsris/8QmHi2gt3sIXi6Yi+CO/sgKzgKtXbNgbqZUbHt9ZrUgcO6SUjcewrBnXyQ4n8FNbdMg3adarI21qN7wHJwN0RP34AQz6nIz8xG7V1zINKS/0PmyU97EOQ6SLbFb/v7tfepi6yQh4gY8SOCO0xA4u+nYb9yPIzaNy2fD+IDYvNdD1gP6YqoaRtwp9s0FGRKUHfPrCKf7+tMv2iFanMG4/Hy33HHYzIyg6NQd89sueuspqOFlHOBeLLmQEWcxgfFoMunsPx+OBLX7kGk11hkhzxAtW0LoGZa/PeJjqsTqqzwRcr+E4jsPhbppy7B7udZ0KpVXdZGo5o1qv/2E3IePMbD/r544Dkaiet+g1SSAwDIjUlEqJu33Jaw6lfkp2ci/Z/rxb4vvVtWVjbqODpgxqTRqg7lgyGVSstkKy9JSUnw9vaGoaEhjI2NMXToUKSnp5fY/uVIaHHbH3/8IWtX3Ot79+5VOD6FE0gtLS1YW1vD2toaLi4umDZtGh49eoSEhAQAwLlz5yASiZCSkiI7JigoCCKRCFFRUQDePYyan58PHx8fGBsbw8zMDFOnTi1ykd4cwq5RowYWL16MIUOGwMDAANWqVcOmTZvkjrl48SJcXFygra2Npk2b4vDhw6UadtbV1YW1tTWqVq2KFi1a4Mcff8TGjRuxefNmnDp1CkDRIeyXn8Px48fh6uoKHR0dfP7554iPj8f//vc/ODk5wdDQEF9//TUyM19VQwoKCuDn5wd7e3vo6OigUaNG2L9/v+z1l/2ePn0aTZs2ha6uLlq2bIn79+/L2ty8eRPt2rWDgYEBDA0N0aRJE1y/fr3Ez379+vWoWbMmNDU1UadOHfz6669yr4tEImzZsgU9evSArq4uatWqhb/++uutn5mqWY3ojsTfTuDZ72eQHfYYD6etR0G2BOZ92xfffqgnUs/dQNyGw8gOf4ynS/cg884DWA56Va21HOqJmNW/I+XEVWSFPETUhFXQsDKFsUdzub7y07OQl5Ai2wqyJLLXYtfux9Ole5ARcB+Sh7GI33oUqecCYdK5Rfl8EB8Q62Hd8GTVfiQfv1aYZI9bDU0rU5h0albiMTYjPBG/5yQS951BVthjRPpuREGWBBb9Ppe1id1yFDFrDyE9oOSqmlCZDemBlH3+SD1wEjnhjxA7ey0KsiQw/rJjse1NB3ZH+r8BSNpyADkRj5Cw8ldkB0fA5BtPWRuLiQORcf464pdsgyT4AXKjY5F+5gryk1ILGxQUID8xWW4z6NASz//3L6SZ2RVx2h+lNm6fYNyIgXD/rJWqQ6Ey4u3tjbt37+LkyZM4evQo/vnnH7nC1ZtejoS+vs2bNw/6+vro3LmzXNtffvlFrt3rUwRLS6k5kOnp6di1axccHR1hZmamTFdyli1bhu3bt2Pbtm3477//kJSUhEOHDpXquKZNmyIwMBCjR4/Gt99+K0us0tLS4OnpiQYNGuDGjRtYsGABfH193zvGgQMHwsTEBAcPHnxru7lz52Lt2rW4ePEiHj16hK+++gorV67Enj178Pfff+PEiRNYs2aNrL2fnx927tyJDRs24O7du5g4cSL69+8vG0Z/acaMGVi2bBmuX78OdXV1DBkyRPaat7c3qlatimvXriEgIADTpk0rcbj/0KFDGD9+PCZNmoQ7d+5g5MiRGDx4MM6ePSvXbt68efjqq69w69YtdOnSBd7e3khKSlL0Y6sQIg116DWoibR/b73aKZUi7d+b0Gtcp9hj9JrUkW8PIO18IPSbFLbXrGYFTStTuTb5zzORERQqa/OSzXc94XJ7J5z9l8NqlBeg9vZvMzUDXeSllPxXpRBoVbOCppUJ0v69KduX/zwT6YFhMGhS/DUTaahDr2HR65z6760Sj6HXaKhDu54jMi4GvdonlSLjYhB0XOsWe4iOa11kXAyU25f+bwB0XF60F4mg3/YT5EQ9gd22Bah1eQ9q7F8BfXe3EsPQrucIbeeaSPnjhLJnRKSQAkjLZCsPISEh8Pf3x5YtW9C8eXO0bt0aa9aswd69e/H06dNij3k5Evr6dujQIXz11VfQ19eXa2tsbCzXTltbW+EYFU4gjx49Cn19fejr68PAwAB//fUX9u3bB7G47NbjrFy5EtOnT0fPnj3h5OSEDRs2wMio+CGV13Xp0gWjR4+Go6MjfH19YW5uLkuE9uzZA5FIhM2bN8PZ2RmdO3fGlClT3jtGsViM2rVry6qqJVm4cCFatWoFV1dXDB06FOfPn8f69evh6uqKNm3a4Msvv5TFKJFIsHjxYmzbtg0eHh5wcHDAoEGD0L9/f2zcuFGu30WLFuGzzz6Ds7Mzpk2bhosXLyI7u/Cv9+joaLi7u6Nu3bqoVasWevfujUaNGhUb39KlSzFo0CCMHj0atWvXho+PD3r27ImlS5fKtRs0aBD69esHR0dHLF68GOnp6bh69ep7fnrlS93UACJ1NeQmpMjtz0tMhYalSbHHaFgYIy9Rvn1uQio0LExkrxf2UXIbAIjfdhQPvluG+1/NQsKu47AZ8yWqzhhYYqwm3VpBr1EtJP5+unQn95HSsDQGUPh5vi43IaXEa1bSdc5NTJFdLyqZuokhROpqyE9Mltuf/ywF6hamxR9jboL8N74H8hNToP7ie0DNzBhq+rowG9EbGf8EIHrwTDw/cRFV182AbrP6xfZp3LsjJOHRyAoMUf6kiBRQmYewL126BGNjYzRt+mp6k7u7O8RiMa5cuVKqPgICAhAUFIShQ4cWee27776Dubk5mjVrhm3btr3XeSg8Ya9du3ZYv349ACA5ORk///wzOnfujKtXr6J69ervOPrdUlNTERMTg+bNXw0Lqquro2nTpu88wdcX8ohEIlhbWyM+Ph4AcP/+fTRs2FAuy27WrOShsdKQSqUQiUSljsnKygq6urpwcHCQ2/cyEQsPD0dmZiY6dOgg10dOTg5cXV1L7NfGxgYAEB8fj2rVqsHHxwfDhg3Dr7/+Cnd3d/Tu3Rs1a9YsNr6QkJAiJfFWrVph1apVJb6fnp4eDA0NZZ/tmyQSCSQSidy+HGk+NEVqxbb/mMRtfjW0nxXyEAW5eaj+w7d48sOvkObkybU1aFkfNZaPxcOp65Ad+qiiQ1Upsx6fwn7JSNnX979ZpMJoqKyIxIU/D5+fvoyk7YcBAJKQB9Bp7ATjfl2QefWOfHstTRh6tkXiut8qOlSiMlPc7zwtLS1oaWm9d5+xsbFF1iuoq6vD1NQUsbGxpepj69atcHJyQsuW8oup5s+fj88//xy6uro4ceIERo8ejfT0dIwbN06hGBUuG+rp6cHR0RGOjo745JNPsGXLFmRkZGDz5s2FHb6oRL6e7OXm5ir6Nu/lzWFakUiEgoKCcnmv/Px8hIWFwd7evtQxiUSit8b4cnLs33//jaCgINkWHBwsNw+yuH4ByPqZO3cu7t69i65du+LMmTNwdnYu1RSA0p7Hm3G/yc/PD0ZGRnLb9uclr34ua3lJzyHNyy9ShVI3N0JufHKxx+QmpEDdXL69hoURchOSZa8X9lFym+JkBIZCrKEOzaryPwj0W9SD4y8z8GjeNjw7cO6d5/SxST5xFbc7TJJteUnPARR+nq/TsDAu8ZqVdJ01zI2LVCWpqLzkNEjz8qFmLl/hVTMzRl5C8dNT8hKTofbG94CauTHyXnwP5CWnQZqbB0l4tFwbScQjaNgUXbxn0Kk1xNpaSD0s7Ao8qUZZDWEX9zvPz8+v2PecNm1aiQtdXm737t1T+tyysrKwZ8+eYquPs2bNko2M+vr6YurUqfjpp58Ufg+lx51FIhHEYjGysrIAABYWFgCAmJhXt95Q5N6IRkZGsLGxkSvR5uXlISAg4C1HvVudOnVw+/Ztub8Srl279t797dixA8nJyejVq5dScb3O2dkZWlpaiI6OliXpLzc7OzuF+qpduzYmTpyIEydOoGfPnvjll1+Kbefk5IQLFy7I7btw4QKcnZ3f+zymT5+O1NRUuW2QQa337k9R0tw8ZNyOgEHr124tJRLBsHVDZNy4X+wxGQH3Ydha/lZUhm1ckB5Q2D4nOg45cUlybcT6OtBzqS1rUxzdevaQ5ucj79mroVkDt/qotWMmHi/eicTdwpz3VZCRDUlUrGzLCn2EnLhkuc9XTV8H+q618LyEz1eam4eMWxHy100kglHrhiUeQ6/JzUP23XDoub02vUUkgl5LF2QFFv8LLCvwHvTcXOT26bVyRVbQPVmfWbdDoWVfVa6NVo0qyH1adMTCuHdHPD9zBflJaUqdCtH7kJbR/xX3O2/69OnFvuekSZMQEhLy1s3BwUFuBPWlvLw8JCUlwdra+p3ntn//fmRmZmLAgAHvbNu8eXM8fvy4SBX1XRQewpZIJLLyaXJyMtauXYv09HR4ehauwnuZ7MydOxeLFi1CaGgoli1bptB7jB8/Hj/88ANq1aqFunXrYvny5XKrut/H119/jRkzZmDEiBGYNm0aoqOjZfP83jUMnZmZidjYWOTl5eHx48c4dOgQVqxYgW+//Rbt2rVTKq7XGRgYYPLkyZg4cSIKCgrQunVrpKam4sKFCzA0NMTAgSXPpXspKysLU6ZMwZdffgl7e3s8fvwY165dKzHRnTJlCr766iu4urrC3d0dR44cwcGDB2Wry99HcaX7ih6+jtv0J+xXjEfmzXBkBIXBapgnxDraSNxXWOmosXI8cmOf4ckPuwrbbz2COvsXwWpEd6Sevg7T7m2g27Amonx/lvUZv/UIbMb1RnbkU+Q8ioft5K+RG5eElOOFf+zoNa4DPdfaeH7xNvIzsqDfpA7s5gzBs4PnkZ+aAaBw2Npx+0zEbz2K5GOXoP6ieibNzUO+wBfSxG45iirjv0R2ZAwk0XGoOrUfcuKSkOz/aq5t3X1zkex/BXG//A8AELPpCGquHIuMm+FIDwyD9XBPiHW1kLD3jOwYDQtjaFgaQ9u+cKqHbt3qyM/IguRJouA/82fbDsF2iQ+y74Qh61YoTAd1h1hHCykHTgIAbJZMQl7cMyQs2w4ASNrxJ6rv/hGmQ3og/dw1GHb9DDr1ayF25quFgElbDqDKymnIvHYbGZdvQf/TJtD/vDke9pdftKhRzQa6n9THo2FzKux8P2aZmVmIfvxqccWTp3G4FxoBI0MD2Fh//Lduex8FZTR/UZHhagsLC1mh7W3c3NyQkpKCgIAANGnSBABw5swZFBQUyE3xK8nWrVvxxRdflOq9goKCYGJiovCQu8IJpL+/v2zOnYGBAerWrYs//vgDbdu2BVA41Pnbb7/h22+/RcOGDfHJJ59g4cKF6N27d6nfY9KkSYiJicHAgQMhFosxZMgQ9OjRA6mpqe8+uASGhoY4cuQIvv32W7i4uKBBgwaYPXs2vv7663euPtq8eTM2b94MTU1NmJmZoUmTJti3bx969Ojx3vGUZMGCBbCwsICfnx8ePHgAY2NjNG7cGN9//32pjldTU8OzZ88wYMAAxMXFwdzcHD179sS8efOKbe/l5YVVq1Zh6dKlGD9+POzt7fHLL7/IrueHKvnIBaibGcF2cj9oWJggMzgSYd/MQ15i4X9DWlUsgIJXPzwyAu4jcsxyVJnqjSq+/SGJfIqIYT8g+/6robjYnw9BrKuNGj+OhpqhHtKvhSC0/3xIJYVTNKQ5uTDt3hq2Pn0h1lKHJDoecZuPIG7zn7I+zL78HGq62rAZ+yVsxn4p2//80h3c7z2zvD+WSi1m3SGIdbVgv2QU1A318PxaCO57L5B9vgCgXcMa6qaGsq+T/roADTNDVJ3Sr/BG4ncjcc97gew6A4DlAA9UndRH9rXz4cL5lhET1iDxd/m7DQjN82P/IN7UEBbjv4GahQkkIQ8QPXQ28p+lAAA0bC0A6aupKlmBIXjiswQWEwfAYtIg5EQ9waPRCyAJe/iqz5OXEDNnLcxHfgWrWaOQE/kYj8csQlZAsNx7G3/ZEXmxicj470aFnOvH7s69MAwZ+ypJX7Km8DZ23Tu7Y9HMSaoKi96Tk5MTOnXqhOHDh2PDhg3Izc3FmDFj0LdvX9ja2gIAnjx5gvbt22Pnzp1yazrCw8Pxzz//4NixY0X6PXLkCOLi4tCiRQtoa2vj5MmTWLx4MSZPnqxwjCKpgJ8mvnv3bgwePBipqanQ0dFRdTgftetVvVQdAr2QX6CSJ5hSCQz1eO/DysLx0lpVh0AvaJg7vLuRkupZvbuSVxp340q3KlpRSUlJGDNmDI4cOQKxWIxevXph9erVslvyREVFwd7eHmfPnpUr+nz//ffYtWsXoqKiitwhx9/fH9OnT0d4eDikUikcHR3x7bffYvjw4QrfTUdQCeTOnTvh4OCAKlWq4ObNmxgzZgzatm2LXbt2qTq0jx4TyMqDCWTlwgSy8mACWXlURALpZKncnVheComvnLe0K2+V+7l7ZSw2NhazZ89GbGwsbGxs0Lt3byxaxNuHEBERESlCUBVIUh1WICsPViArF1YgKw9WICuPiqhA1rX8pEz6uRf//nd0+ZAJqgJJREREBJTdKmyhYimCiIiIiBTCCiQREREJjhSsQCqDCSQREREJDoewlcMhbCIiIiJSCCuQREREJDgcwlYOE0giIiISHOlrj+kkxTGBJCIiIsEpYAVSKZwDSUREREQKYQWSiIiIBIcP4lMOE0giIiISHA5hK4dD2ERERESkEFYgiYiISHA4hK0cJpBEREQkOHwSjXI4hE1ERERECmEFkoiIiASHT6JRDhNIIiIiEhzOgVQOh7CJiIiISCGsQBIREZHg8D6QymECSURERILDIWzlMIEkIiIiweFtfJTDOZBEREREpBBWIImIiEhwOIStHCaQREREJDhcRKMcDmETERERkUJYgSQiIiLB4RC2cphAEhERkeBwFbZyOIRNRERERAphBZKIiIgER8pFNEphAklERESCwyFs5XAIm4iIiIgUwgokERERCQ5XYSuHCSQREREJDudAKocJJBEREQkOK5DK4RxIIiIiokpm0aJFaNmyJXR1dWFsbFyqY6RSKWbPng0bGxvo6OjA3d0dYWFhcm2SkpLg7e0NQ0NDGBsbY+jQoUhPT1c4PiaQREREJDhSqbRMtvKSk5OD3r1749tvvy31MUuWLMHq1auxYcMGXLlyBXp6evDw8EB2drasjbe3N+7evYuTJ0/i6NGj+OeffzBixAiF4xNJWcOlCnC9qpeqQ6AX8gv4d2NlYqiX/e5GVCEcL61VdQj0goa5Q7m/h7pmlTLpJy/nSZn0U5Lt27djwoQJSElJeWs7qVQKW1tbTJo0CZMnTwYApKamwsrKCtu3b0ffvn0REhICZ2dnXLt2DU2bNgUA+Pv7o0uXLnj8+DFsbW1LHRd/kxARERG9J4lEgrS0NLlNIpFUeByRkZGIjY2Fu7u7bJ+RkRGaN2+OS5cuAQAuXboEY2NjWfIIAO7u7hCLxbhy5YpC78dFNFQhmj4+rOoQlCaRSODn54fp06dDS0tL1eEIGq9F5cFrUXnwWiimrCqHc+fOxbx58+T2zZkzB3Pnzi2T/ksrNjYWAGBlZSW338rKSvZabGwsLC0t5V5XV1eHqamprE1psQJJVEoSiQTz5s1TyV+WJI/XovLgtag8eC1UY/r06UhNTZXbpk+fXmzbadOmQSQSvXW7d+9eBZ/B+2EFkoiIiOg9aWlplbriO2nSJAwaNOitbRwc3m/+p7W1NQAgLi4ONjY2sv1xcXFwcXGRtYmPj5c7Li8vD0lJSbLjS4sJJBEREVEFsLCwgIWFRbn0bW9vD2tra5w+fVqWMKalpeHKlSuyldxubm5ISUlBQEAAmjRpAgA4c+YMCgoK0Lx5c4Xej0PYRERERJVMdHQ0goKCEB0djfz8fAQFBSEoKEjuno1169bFoUOHAAAikQgTJkzAwoUL8ddff+H27dsYMGAAbG1t4eXlBQBwcnJCp06dMHz4cFy9ehUXLlzAmDFj0LdvX4VWYAOsQBKVmpaWFubMmcPJ6ZUAr0XlwWtRefBafFxmz56NHTt2yL52dXUFAJw9exZt27YFANy/fx+pqamyNlOnTkVGRgZGjBiBlJQUtG7dGv7+/tDW1pa12b17N8aMGYP27dtDLBajV69eWL16tcLx8T6QRERERKQQDmETERERkUKYQBIRERGRQphAEhEREZFCmEASERERkUK4CpuIiOgDlZKSgqtXryI+Ph4FBQVyrw0YMEBFUZEQcBU2UQny8/Oxfft2nD59utgfzmfOnFFRZEREwJEjR+Dt7Y309HQYGhpCJBLJXhOJREhKSlJhdPSxYwJJVIIxY8Zg+/bt6Nq1K2xsbOR+OAPAihUrVBQZkWrcunWr1G0bNmxYjpEQANSuXRtdunTB4sWLoaurq+pwSGCYQBKVwNzcHDt37kSXLl1UHYpg+fj4lLrt8uXLyzESAgCxWAyRSISSfm28fE0kEiE/P7+CoxMePT093L59+72fnUykDM6BJCqBpqYmHB0dVR2GoAUGBsp9fePGDeTl5aFOnToAgNDQUKipqcme6UrlKzIyUtUh0Gs8PDxw/fp1JpCkEkwgiUowadIkrFq1CmvXri0yfE0V4+zZs7J/L1++HAYGBtixYwdMTEwAAMnJyRg8eDDatGmjqhAFpXr16qoOgV7TtWtXTJkyBcHBwWjQoAE0NDTkXv/iiy9UFBkJAYewiUrQo0cPnD17FqampqhXr16RH84HDx5UUWTCVKVKFZw4cQL16tWT23/nzh107NgRT58+VVFkwhYcHIzo6Gjk5OTI7WfyUv7E4pLvxMdpBFTeWIEkKoGxsTF69Oih6jDohbS0NCQkJBTZn5CQgOfPn6sgImF78OABevTogdu3b8vNi3xZrWfyUv7evDMEUUViBZKIPggDBgzAv//+i2XLlqFZs2YAgCtXrmDKlClo06YNduzYoeIIhcXT0xNqamrYsmUL7O3tcfXqVTx79gyTJk3C0qVLOa2A6CPHBJLoHRISEnD//n0AQJ06dWBhYaHiiIQpMzMTkydPxrZt25CbmwsAUFdXx9ChQ/HTTz9BT09PxREKi7m5Oc6cOYOGDRvCyMgIV69eRZ06dXDmzBlMmjSpyAIoKh/nz5/H0qVLERISAgBwdnaW/VFFVJ74KEOiEmRkZGDIkCGwsbHBp59+ik8//RS2trYYOnQoMjMzVR2e4Ojq6uLnn3/Gs2fPEBgYiMDAQCQlJeHnn39m8qgC+fn5MDAwAFCYTL6cg1q9enXZH1xUvnbt2gV3d3fo6upi3LhxGDduHHR0dNC+fXvs2bNH1eHRR44VSKISjBw5EqdOncLatWvRqlUrAMB///2HcePGoUOHDli/fr2KIxSux48fAwCqVq2q4kiEq02bNpg0aRK8vLzw9ddfIzk5GTNnzsSmTZsQEBCAO3fuqDrEj56TkxNGjBiBiRMnyu1fvnw5Nm/eLKtKEpUHJpBEJTA3N8f+/fvRtm1buf1nz57FV199VeyCDio/BQUFWLhwIZYtW4b09HQAgIGBASZNmoQZM2a8dUUqlb3jx48jIyMDPXv2RHh4OLp164bQ0FCYmZlh3759+Pzzz1Ud4kdPS0sLd+/eLXK/2vDwcNSvXx/Z2dkqioyEgKuwiUqQmZkJKyurIvstLS05hK0CM2bMwNatW/HDDz/IVYTnzp2L7OxsLFq0SMURCouHh4fs346Ojrh37x6SkpJgYmLC+6ZWEDs7O5w+fbpIAnnq1CnY2dmpKCoSClYgiUrQvn17mJmZYefOndDW1gYAZGVlYeDAgUhKSsKpU6dUHKGw2NraYsOGDUXuL/jnn39i9OjRePLkiYoiI1KN9evXY8KECRgyZAhatmwJALhw4QK2b9+OVatWYeTIkSqOkD5mTCCJSnDnzh14eHhAIpGgUaNGAICbN29CW1sbx48fL3JDaypf2trauHXrFmrXri23//79+3BxcUFWVpaKIhOOnj17Yvv27TA0NETPnj3f2pY32q8Yhw4dwrJly2TzHZ2cnDBlyhR0795dxZHRx45D2EQlqF+/PsLCwrB7927cu3cPANCvXz94e3tDR0dHxdEJT6NGjbB27VqsXr1abv/atWtlCT6VLyMjI9nwtKGhIYeqK4EePXrwgQekEqxAEtEH4fz58+jatSuqVasGNzc3AMClS5fw6NEjHDt2jPe9IyKqQEwgiV7z119/oXPnztDQ0MBff/311rZ81m/Fe/r0KdatWyerCDs5OWH06NGwtbVVcWTC8/nnn+PgwYMwNjaW25+WlgYvLy+cOXNGNYF95ExNTREaGgpzc/N3LlhKSkqqwMhIaJhAEr1GLBYjNjYWlpaWb70tjEgk4rN+SdBe/155XXx8PKpUqSJ7WhCVrR07dqBv377Q0tLC9u3b35pADhw4sAIjI6FhAklEldatW7dQv359iMVi3Lp1661tGzZsWEFRCdvL6+Di4oIzZ87A1NRU9lp+fj78/f2xceNGREVFqShCIqoITCCJSrBz50706dMHWlpacvtzcnKwd+9eDBgwQEWRCcebFWGRSITifmSxIlxxXl4HAMVeCx0dHaxZswZDhgyp6NAE58aNG9DQ0ECDBg0AFN7S6pdffoGzszPmzp0LTU1NFUdIHzMmkEQlUFNTQ0xMTJEhumfPnsHS0pIJSwV4+PAhqlWrBpFIhIcPH761bfXq1SsoKmF7+PAhpFIpHBwccPXqVVhYWMhe09TUhKWlJdTU1FQYoXB88sknmDZtGnr16oUHDx7A2dkZPXv2xLVr19C1a1esXLlS1SHSR4wJJFEJxGIx4uLi5H5BAoX3gmzXrh0nqFeg3NxcjBw5ErNmzYK9vb2qwyGqFIyMjHDjxg3UrFkTP/74I86cOYPjx4/jwoUL6Nu3Lx49eqTqEOkjxvtAEr3B1dUVIpEIIpEI7du3h7r6q2+T/Px8REZGolOnTiqMUHg0NDRw4MABzJo1S9Wh0GvCwsJw9uxZxMfHo6CgQO612bNnqygq4ZBKpbLP/dSpU+jWrRuAwkccJiYmqjI0EgAmkERv8PLyAgAEBQXBw8MD+vr6stc0NTVRo0YN9OrVS0XRCZeXlxcOHz6MiRMnqjoUArB582Z8++23MDc3h7W1tdxqYJFIxASyAjRt2hQLFy6Eu7s7zp8/j/Xr1wMAIiMjYWVlpeLo6GPHIWyiEuzYsQN9+vSRPQebVGvhwoVYtmwZ2rdvjyZNmkBPT0/u9XHjxqkoMmGqXr06Ro8eDV9fX1WHIli3bt2Ct7c3oqOj4ePjgzlz5gAAxo4di2fPnmHPnj0qjpA+ZkwgieiD8La5jyKRCA8ePKjAaMjQ0BBBQUFwcHBQdSj0huzsbKipqUFDQ0PVodBHjAkkUQny8/OxYsUK/P7774iOjkZOTo7c61xEQ0I2dOhQfPLJJxg1apSqQxG8nJycYuehVqtWTUURkRBwDiRRCebNm4ctW7Zg0qRJmDlzJmbMmIGoqCgcPnyY87tUKCcnB5GRkahZs6bcAieqWI6Ojpg1axYuX76MBg0aFKl2cUpB+QsNDcXQoUNx8eJFuf1SqZT3RqVyxwokUQlq1qyJ1atXo2vXrjAwMEBQUJBs3+XLlzm/qIJlZmZi7Nix2LFjB4DCX54ODg4YO3YsqlSpgmnTpqk4QmHhlALVa9WqFdTV1TFt2jTY2NgUeaxho0aNVBQZCQH/fCcqQWxsrOwJD/r6+khNTQUAdOvWjbeTUYHp06fj5s2bOHfunNxtlNzd3TF37lwmkBUsMjJS1SEIXlBQEAICAlC3bl1Vh0ICJFZ1AESVVdWqVRETEwOgsBp54sQJAMC1a9eKPN6Qyt/hw4exdu1atG7dWq7SUq9ePURERKgwMmHLycnB/fv3kZeXp+pQBMfZ2Zn3eySVYQJJVIIePXrg9OnTAApvizFr1izUqlULAwYM4HN+VSAhIaHIYyUBICMjo8jQHZW/zMxMDB06FLq6uqhXrx6io6MBFH6v/PDDDyqOThh+/PFHTJ06FefOncOzZ8+QlpYmtxGVJ86BJCqly5cv4+LFi6hVqxY8PT1VHY7gfPrpp+jduzfGjh0LAwMD3Lp1C/b29hg7dizCwsLg7++v6hAFZfz48bhw4QJWrlyJTp064datW3BwcMCff/6JuXPnIjAwUNUhfvTE4sIa0Jt/QHERDVUEzoEkKkF2drbcTcRbtGiBFi1aqDAiYVu8eDE6d+6M4OBg5OXlYdWqVQgODsbFixdx/vx5VYcnOIcPH8a+ffvQokULTilQkbNnz6o6BBIwJpBEJbC0tESPHj3Qv39/tG/fXvbXPqlG69atcfPmTfj5+aFBgwY4ceIEGjdujEuXLskWO1HF4ZQC1fvss89UHQIJGH8jEpVgx44dyMzMRPfu3VGlShVMmDAB169fV3VYgpSWloaTJ08iJCQEixcvxtWrVxEcHIxdu3YxeVSRpk2b4u+//5Z9/TJp3LJlC9zc3FQVluD8+++/6N+/P1q2bIknT54AAH799Vf8999/Ko6MPnasQBKVoEePHujRoweeP3+O/fv347fffkOLFi3g4OCA/v3782biFSQoKAhdunRBXFwcpFIpDAwM8Pvvv8PDw0PVoQkapxSo3oEDB/DNN9/A29sbN27cgEQiAQCkpqZi8eLFOHbsmIojpI8ZF9EQKSA4OBje3t64desWJ6hXEA8PD6Snp2Pp0qXQ1tbGggULcPv2bYSFhak6NMGLiIjADz/8gJs3byI9PR2NGzeGr68vq8IVxNXVFRMnTsSAAQNgYGCAmzdvwsHBAYGBgejcuTNiY2NVHSJ9xJhAEr1DdnY2/vrrL+zZswf+/v6wsrJCv379eKuSCmJubi6b7wgAKSkpMDU1RUpKCgwNDVUcHZHq6OrqIjg4GDVq1JBLIB88eABnZ2dkZ2erOkT6iHEIm6gEx48fx549e3D48GGoq6vjyy+/xIkTJ/Dpp5+qOjRBSUpKQtWqVWVfGxsbQ09PD8+ePWMCWcEUubcgr035s7a2Rnh4OGrUqCG3/7///oODg4NqgiLBYAJJVIIePXrA09MTO3fuRJcuXaChoaHqkAQrODhYbjhOKpUiJCQEz58/l+1r2LChKkITFGNj41KvsOYUj/I3fPhwjB8/Htu2bYNIJMLTp09x6dIlTJ48mY9bpXLHIWyiYuTl5WH9+vXo3bs3rK2tVR2OoInFYohEIhT3o+rlft40uWK8vjgmKioK06ZNw6BBg2Srri9duoQdO3bAz88PAwcOVFWYgiGVSrF48WL4+fkhMzMTAKClpYXJkydjwYIFKo6OPnZMIIlKoKuri5CQEFSvXl3VoQjaw4cPS9WO16litW/fHsOGDUO/fv3k9u/ZswebNm3CuXPnVBOYAOXk5CA8PBzp6elwdnaGvr6+qkMiAWACSVSCtm3bYsKECfDy8lJ1KKSA0aNHY/78+TA3N1d1KB81XV1d3Lx5E7Vq1ZLbHxoaChcXF1lFjIg+TpwDSVSC0aNHY9KkSXj8+DGaNGkCPT09udc5565y2rVrFyZPnswEspzZ2dlh8+bNWLJkidz+LVu2wM7OTkVRCUt2djbWrFmDs2fPIj4+HgUFBXKv37hxQ0WRkRCwAklUguIeXcg5d5Xf67czofJz7Ngx9OrVC46OjmjevDkA4OrVqwgLC8OBAwfQpUsXFUf48fP29saJEyfw5ZdfwsrKqsgCpzlz5qgoMhICJpBEJXjX3DvOuaucmEBWnEePHmH9+vW4d+8eAMDJyQmjRo1iBbKCGBkZ4dixY2jVqpWqQyEB4hA2UQmYIBK9nZ2dHRYvXqzqMASrSpUqMDAwUHUYJFBFx+iISObXX39Fq1atYGtrK6tIrly5En/++aeKIyOqeLdu3ZLNs7t169ZbNyp/y5Ytg6+vb6nvVEBUlliBJCrB+vXrMXv2bEyYMAGLFi2SzXk0NjbGypUr0b17dxVHSFSxXFxcEBsbC0tLS7i4uLz1/pycI1z+mjZtiuzsbDg4OEBXV7fIww6SkpJUFBkJARNIohKsWbMGmzdvhpeXl9xzr5s2bYrJkyerMDJ6m/79+/MxeuUkMjISFhYWsn+TavXr1w9PnjzB4sWLi11EQ1SeuIiGqAQ6Ojq4d+8eqlevLrcwIywsDA0bNkRWVpaqQ/zoKTIUytsqkdDo6uri0qVLaNSokapDIQFiBZKoBPb29ggKCiqymMbf3x9OTk4qikpYXh8mfVd1hUOmFWvHjh0wNzdH165dAQBTp07Fpk2b4OzsjN9++42L0CpA3bp1+YcsqQwX0RCVwMfHB9999x327dsHqVSKq1evYtGiRZg+fTqmTp2q6vAEITIyEg8ePEBkZCQOHDgAe3t7/PzzzwgMDERgYCB+/vln1KxZEwcOHFB1qIKzePFi6OjoACh8BvbatWuxZMkSmJubY+LEiSqOThh++OEHTJo0CefOncOzZ8+QlpYmtxGVJw5hE73F7t27MXfuXERERAAAbG1tMW/ePAwdOlTFkQlPs2bNMHfu3CI3qD527BhmzZqFgIAAFUUmTLq6urh37x6qVasGX19fxMTEYOfOnbh79y7atm2LhIQEVYf40Xv5sIM3q/N82AFVBA5hE72Ft7c3vL29kZmZifT0dFhaWqo6JMG6ffs27O3ti+y3t7dHcHCwCiISNn19fTx79gzVqlXDiRMn4OPjAwDQ1tbmsGoFOXv2rKpDIAFjAklUgqysLEilUujq6kJXVxcJCQlYuXIlnJ2d0bFjR1WHJzhOTk7w8/PDli1boKmpCQDIycmBn58f56SqQIcOHTBs2DC4uroiNDRUVhm+e/cuatSoodrgBOKzzz5TdQgkYEwgiUrQvXt39OzZE6NGjUJKSgqaNWsGTU1NJCYmYvny5fj2229VHaKgbNiwAZ6enqhatapsxfWtW7cgEolw5MgRFUcnPOvWrcPMmTPx6NEjHDhwAGZmZgCAgIAA9OvXT8XRfbxu3bqF+vXrQywWv/MuBbwzAZUnzoEkKoG5uTnOnz+PevXqYcuWLVizZg0CAwNx4MABzJ49GyEhIaoOUXAyMjKwe/duuWcvf/3119DT01NxZEQVQywWy27mLhaLeTN3UhlWIIlKkJmZKXvO7IkTJ9CzZ0+IxWK0aNGCjw5TET09PYwYMULVYdAL//77LzZu3IgHDx7gjz/+QJUqVfDrr7/C3t4erVu3VnV4HyXezJ0qC97Gh6gEjo6OOHz4MB49eoTjx4/L5j3Gx8fzSScq8uuvv6J169ZyzyZfsWIFn02uAgcOHICHhwd0dHRw48YNSCQSAEBqaioWL16s4ug+XtWrV5etuq5evfpbN6LyxASSqASzZ8/G5MmTUaNGDTRr1gxubm4ACquRrq6uKo5OeNavXw8fHx907twZycnJsuE5ExMTrFy5UrXBCdDChQuxYcMGbN68We4ZzK1atcKNGzdUGJlwnDlzBmPGjEG3bt3g6emJcePG4Z9//lF1WCQQnANJ9BaxsbGIiYlBo0aNZPdcu3r1KgwNDVG3bl0VRycszs7OWLx4Mby8vOQeLXnnzh20bdsWiYmJqg5RUHR1dREcHIwaNWrIXY8HDx7A2dkZ2dnZqg7xozZq1Chs+n97dxoT5fW2AfyaARyEAqKgYECFOFKUJUZFoy11IyqpiNQFG6tSFNHGBXCrlWptSRtwbW0UCE7ViFYFEmttrSw1cQFcEGgHEBQVC6YtaFUWh2XeD0b+zMDQ8qbMsTzX75PznOfDFZOBm7PdCQmwtbXFsGHDoNVqUVpaisePH2PlypX46quvREekHo57IIk64eDgAAcHB1RUVAAAnJ2d4ePjIziVNJWXl3c486tQKFBbWysgkbQ5ODigrKys3ZU9Fy9ehKurq5hQEpGWlgaVSoWDBw9i8eLFrUvaLS0t+Oabb7BixQr4+fkhICBAcFLqybiETWRAU1MToqOjYWNjgyFDhmDIkCGwsbHBli1b0NjYKDqe5LzsTa6PvcnFWLZsGdasWYOcnBzIZDJUVlbi6NGjiIqK4hVX3UylUiEyMhJLlizR6UIjl8vx/vvvY+3atUhKShKYkKSAM5BEBqxatQqpqamIjY1t3f945coVbNu2DdXV1di/f7/ghNLysjd5Q0NDa2/yY8eOtV4uTsa1adMmtLS0YMqUKairq4Ovry8UCgXWr1+PpUuXio7Xo924cQNbtmwxOB4UFIR33nnHiIlIirgHksgAGxsbHD9+HDNmzNB5fvbsWSxYsAB//fWXoGTSxd7krx6NRoOysjI8e/YMw4cPR3x8POLi4vDw4UPR0Xosc3Nz3LlzBwMHDuxw/LfffsPQoUPZUpK6FWcgiQxQKBQdtmRzcXFpbaVHxtHU1ITk5GRMmzaNvckFe/78ObZt24bz58+3zjgGBgZCpVJh9uzZMDExQUREhOiYPZpGo9E5+a7P1NQUGo3GiIlIijgDSWTA9u3bUVxcDJVKBYVCAeDFL8/Q0FAolUps3bpVcEJpsbCwQFFREe+3E2zjxo2Ij4/H1KlTcfnyZfzxxx8ICQlBdnY2Nm/ejLlz58LExER0zB5NLpcjLCwMFhYWHY7X1dUhMTGRnWioW3EGkqiNoKAgnc/p6elwcnKCt7c3ACA/Px8ajQZTpkwREU/SfHx8kJeXxwJSsJMnT+Lw4cMICAjAL7/8Ai8vLzQ1NSE/P1/nQAd1H19fX5SUlPztO0TdiQUkURs2NjY6n/U3ojs7OxszDrWxcuVKREVF4cGDBxg1alS7/tdeXl6CkknLy/9/APDw8IBCoUBERASLRyP6+eefRUcg4hI2Ef03vLzIvS2ZTAatVguZTMblOiMxMTHBw4cPW/sxW1lZoaCgAC4uLoKTEZExcQaSqBMFBQW4desWAMDNzQ2enp6CE0lXeXm56AgEQKvVYsmSJa37ghsaGhAeHt5uRjg1NVVEPCIyEs5AEnUgNzcXoaGhUKvVePkVkclkGDFiBJKSkjBmzBjBCYnECAkJ+UfvqVSqbk5CRCKxgCTSo1arMXbsWLi7uyMiIqK1y4larcbu3btRUlKC7OxsDB8+XHBSaTl8+HCn44sWLTJSEiIiYgFJpGfevHloampCSkpKu4MBWq0WQUFBMDMzw4kTJwQllCZbW1udz42Njairq0OvXr1gYWGBmpoaQcmIiKSHeyCJ9GRlZeGHH37o8FSpTCbD5s2b4e/vLyCZtD169Kjds9LSUqxYsQLr168XkIhIrIKCgg6fy2QymJubY9CgQa17VYn+bZyBJNJjbm6O0tJSg1f2VFRUQKlUoqGhwcjJqCPXrl3DwoULUVxcLDoKkVHJ5fJOr08yMzPD/PnzER8fD3NzcyMmIylofy8GkcQNHjwYubm5BsdzcnJ4mfUrxNTUFJWVlaJjEBldWloalEolEhIScPPmTdy8eRMJCQlwc3NDcnIykpKSkJmZiS1btoiOSj0Ql7CJ9AQHByMyMhJubm7w8PDQGSssLMS6det4YEOA06dP63zWarWoqqrCvn37MGHCBEGpiMSJiYnB3r17MW3atNZnnp6ecHJyQnR0NHJzc2FpaYmoqCjs2LFDYFLqibiETaSnoaEBU6ZMQU5ODvz8/ODu7g6tVouioiKkp6fDx8cHmZmZXBIyMv2LxGUyGezt7TF58mTs3LkTjo6OgpIRidG7d2/k5eXh9ddf13leXFyMkSNHor6+Hnfv3sXw4cNRV1cnKCX1VCwgiTqg0Wiwe/duHDt2rPUi8WHDhiE4OBgRERHcmE5Ewo0cORLe3t5ISEhAr169ALy4nWDZsmXIz89HXl4eLl26hIULF/IifvrXsYAkauPLL79EWFgYzM3Ncf/+fTg5OXXYQo/Ea25uRmFhIQYPHtzuih8iKbh8+TICAgIgl8tbe8EXFhaiubkZZ86cwbhx43DkyBE8fPiQNxXQv44FJFEbLw9k9O/fHyYmJqiqqkL//v1FxyIAa9euhaenJ0JDQ9Hc3AxfX19cuXIFFhYWOHPmDCZOnCg6IpHRPX36FEePHtVpufruu+/CyspKcDLq6XiIhqiNgQMHIiUlBf7+/tBqtXjw4IHB63oGDRpk5HTSdurUKSxcuBAA8N133+Hu3bsoLi7GkSNH8NFHH+HSpUuCExIZn5WVFcLDw0XHIAniDCRRGwkJCVi1ahWampoMvqPVaiGTydDc3GzEZGRubo6ysjI4OTkhLCwMFhYW2LNnD8rLy+Ht7Y0nT56IjkhkdKWlpcjKysLvv/+OlpYWnbGPP/5YUCqSAs5AErURFhaGBQsW4N69e/Dy8kJ6ejr69esnOhYBGDBgANRqNRwdHfHjjz9i//79AIC6ujqYmJgITkdkfImJiVixYgXs7Ozg4OCgc6m4TCZjAUndigUkkR4rKyt4eHhApVJhwoQJPHH9iggJCcG8efPg6OgImUyGqVOnAnhxsbv+NSZEUvDZZ58hJiYGGzduFB2FJIhL2EQG1NfX4/z58zrX+Pj5+aF3796Ck0nXqVOnUFFRgblz58LJyQkAcOjQIfTp0wezZs0SnI7IuKytrXHz5k24urqKjkISxAKSqAOnT5/G0qVL8eeff+o8t7OzQ1JSEmbOnCkoGRHRC6GhoRgzZgwP0ZAQXMIm0nP58mXMmTMHAQEBiIqKgru7OwBArVZj586dmDNnDi5cuIBx48YJTio9GRkZyMjI6PDAwMGDBwWlIhJj6NChiI6ORnZ2Njw9PWFmZqYzvnr1akHJSAo4A0mkx9/fH87OzoiPj+9wfPny5aioqMDZs2eNnEzaPvnkE2zfvh2jR49u3QfZVlpamqBkRGK4uLgYHJPJZLhz544R05DUsIAk0tO3b19cuHABnp6eHY4XFBTgrbfewqNHj4ycTNocHR0RGxuL9957T3QUIiLJY482Ij319fWwtrY2OG5jY2PwcnHqPhqNBuPHjxcdg4iIwD2QRO0olUpkZmYiJCSkw/GMjAwolUojp6KlS5ciOTkZ0dHRoqMQCRMZGYlPP/0UlpaWiIyM7PTdXbt2GSkVSRELSCI9ISEhWLduHQYMGAB/f3+dse+//x4bNmzA5s2bBaWTroaGBiQkJCA9PR1eXl7tDgzwlyVJQV5eHhobG1v/bYj+HmGifxv3QBLpaWlpwfz585GSkgI3Nze4u7tDq9WiqKgIpaWlCAwMxMmTJyGXcweIMU2aNKnT8aysLCMlISIiFpBEBpw4cQLJyck6F4kHBwcjODhYcDIiIiKxuIRN1IGmpiY0NDTgwIEDcHBwEB1H0oKCgv72HZlMhpSUFCOkIRLrn3wfXkpNTe3GJCR1LCCJOmBqaorw8HAUFRWJjiJ5NjY2oiMQvTLafh+0Wi3S0tJgY2OD0aNHAwCuX7+Ox48fd6nQJPr/4BI2kQETJ07E2rVrERgYKDoKEVE7GzduRE1NDQ4cOAATExMAQHNzM1auXAlra2vExcUJTkg9GQtIIgNOnDiBDz/8EBERERg1ahQsLS11xr28vAQlIyIC7O3tcfHiRbi5uek8Lykpwfjx41FdXS0oGUkBl7CJDHh5WKZtP1mZTAatVguZTIbm5mZR0YiI0NTUhOLi4nYFZHFxcbte8UT/NhaQRAaUl5eLjkBEZFBISAhCQ0Nx+/Zt+Pj4AABycnLwxRdfGGyEQPRv4RI2ERHRf1BLSwt27NiBvXv3oqqqCsCLnvFr1qxBVFRU675Iou7AApLob6jVaty/fx8ajUbneUBAgKBERES6njx5AgCwtrYWnISkggUkkQF37tzB7NmzUVhY2Lr3EfhfizDugSQiIqniHkgiA9asWQMXFxdkZGTAxcUFubm5qK6uRlRUFHbs2CE6HhFJ0MiRI/9xn+sbN250cxqSMhaQRAZcuXIFmZmZsLOzg1wuh1wuxxtvvIHPP/8cq1evRl5enuiIRCQxvJeWXhUsIIkMaG5uhpWVFQDAzs4OlZWVcHNzw+DBg1FSUiI4HRFJ0datW0VHIALAApLIIA8PD+Tn58PFxQVjx45FbGwsevXqhYSEBLi6uoqOR0REJAwP0RAZcO7cOdTW1iIoKAhlZWV4++23cevWLfTr1w/ffvstJk+eLDoiEUlM3759cevWLdjZ2cHW1rbT/ZA1NTVGTEZSwwKSqAtqamr+9oc2EVF3OXToEIKDg6FQKHDo0KFO3128eLGRUpEUsYAk+htlZWW4ffs2fH190bt379ZWhkRERFLFPZBEBlRXV2PevHnIysqCTCZDaWkpXF1dERoaCltbW+zcuVN0RCKSsPv373c6PmjQICMlISliAUlkQEREBMzMzHD//n24u7u3Pp8/fz4iIyNZQBKRUEOGDOl0NYTNDqg7sYAkMuCnn37CuXPn4OTkpPNcqVTi3r17glIREb2gfxdtY2Mj8vLysGvXLsTExAhKRVLBApLIgNraWlhYWLR7XlNTA4VCISAREdH/eHt7t3s2evRoDBw4EHFxcQgKChKQiqRCLjoA0aumsrISAPDmm2/i8OHDrc9lMhlaWloQGxuLSZMmiYpHRNQpNzc3XL16VXQM6uE4A0mkZ8SIEfj6668RFxeHyZMn49q1a9BoNNiwYQN+/fVX1NTU4NKlS6JjEpHEPXnyROezVqtFVVUVtm3bBqVSKSgVSQULSCI9MTExWL58OaZPnw61Wo0DBw7AysoKz549Q1BQED744AM4OjqKjklEEtenT592h2i0Wi2cnZ1x/PhxQalIKngPJFEHysvLERoaCrVajYSEBAQEBIiORESk48KFCzqf5XI57O3tMXToUJiacn6IuhcLSKJO7Nu3DxEREXB3d2/3A/nGjRuCUhEREYnFP1GIDLh37x5SU1Nha2uLWbNm8S96InqlVFdXo1+/fgCAiooKJCYmor6+HjNnzoSvr6/gdNTTcQaSqAOJiYmIiorC1KlTER8fD3t7e9GRiIgAAIWFhZg5cyYqKiqgVCpx/PhxTJ8+HbW1tZDL5aitrcWpU6cQGBgoOir1YCwgifRMnz4dubm52LNnDxYtWiQ6DhGRjhkzZsDU1BSbNm3CkSNHcObMGUybNg2JiYkAgFWrVuH69evIzs4WnJR6MhaQRHr8/PygUqnadaAhInoV2NnZITMzE15eXnj27Bmsra1x9epVjBo1CgBQXFyMcePG4fHjx2KDUo/GTV1Ees6fPy86AhGRQTU1NXBwcAAAvPbaa7C0tIStrW3ruK2tLZ4+fSoqHkkEO9EQERH9x+jf/6j/mai7cQaSiIjoP2bJkiVQKBQAgIaGBoSHh8PS0hIA8Pz5c5HRSCK4B5KIiOg/JCQk5B+9p1KpujkJSRkLSCIiIiLqEu6BJCIiIqIuYQFJRERERF3CApKIiIiIuoQFJBERERF1CQtIIiIiIuoSFpBERERE1CUsIImIiIioS1hAEhEREVGXsIAkIiIioi75P5OTf+jDybomAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def display_corr_matrix(df):  \n",
    "  df_corr = df.corr(numeric_only=True)\n",
    "  sns.heatmap(df_corr, annot=True, vmax=1, vmin=-1)\n",
    "display_corr_matrix(df_insurance)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can notice that the **YearOfObservation** attribute is highly correlated with the **Insured_Period** attribute. So his can be used in PCA later"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dealing with Nan Values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Garden - Most Frequent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "YearOfObservation      0\n",
       "Insured_Period         0\n",
       "Residential            0\n",
       "Building_Painted       0\n",
       "Building_Fenced        0\n",
       "Garden                 0\n",
       "Settlement             0\n",
       "Building Dimension    77\n",
       "Building_Type          0\n",
       "NumberOfWindows        0\n",
       "Geo_Code              73\n",
       "Claim                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def handle_nan_values(df:pd.DataFrame, column:str, strategy:str):\n",
    "  mf_imputer = SimpleImputer(missing_values=np.nan, strategy=strategy)\n",
    "  garden_arr = mf_imputer.fit_transform(df_insurance.loc[:, [column]])\n",
    "  df_insurance[column] = pd.DataFrame(data=garden_arr, columns=[column])\n",
    "  \n",
    "handle_nan_values(df_insurance, 'Garden', 'most_frequent')\n",
    "df_insurance.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Building Dimension - Nan Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "YearOfObservation      0\n",
       "Insured_Period         0\n",
       "Residential            0\n",
       "Building_Painted       0\n",
       "Building_Fenced        0\n",
       "Garden                 0\n",
       "Settlement             0\n",
       "Building Dimension     0\n",
       "Building_Type          0\n",
       "NumberOfWindows        0\n",
       "Geo_Code              73\n",
       "Claim                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkQAAAGwCAYAAABIC3rIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAABSr0lEQVR4nO3deVxU9cI/8M+ZFQYYVmFAFjFXzCU1lbZHiysudVu8laVlZVpe9EntWvlUanXvz7K0umV6b92inrLtudnNLM0wNRU3klwjNQwXBgSEYWeW7++PYQ6MIrKfgfm8X695MXPOd858D8eYT9/tSEIIASIiIiIvplK6AkRERERKYyAiIiIir8dARERERF6PgYiIiIi8HgMREREReT0GIiIiIvJ6DERERETk9TRKV6AzcDgcOHv2LAICAiBJktLVISIioiYQQqC0tBRRUVFQqRpvA2IgaoKzZ88iJiZG6WoQERFRC5w6dQrR0dGNlmEgaoKAgAAAzl+o0WhUuDZERETUFBaLBTExMfL3eGMYiJrA1U1mNBoZiIiIiDqZpgx34aBqIiIi8noMREREROT1GIiIiIjI6zEQERERkddjICIiIiKvx0BEREREXo+BiIiIiLweAxERERF5PQYiIiIi8noMREREROT1GIiIiIjI6zEQERERkddjICIiIiKvx0DUCTkcAntPFsFmdyhdFSIioi6BgagTendHNu5cnY7FXx1WuipERERdgqKBaNWqVRg0aBCMRiOMRiMSExPx7bffyvurqqqQkpKC0NBQ+Pv7Y9KkScjLy3M7Rk5ODiZOnAiDwYDw8HAsWLAANpvNrcyWLVswdOhQ6PV69OrVC6mpqR1xeu3m832nAQAf78nB8fxShWtDRETU+SkaiKKjo/Hiiy8iIyMD+/btw4033ohbb70Vhw87Wz7mzZuHdevW4fPPP8fWrVtx9uxZ3HHHHfL77XY7Jk6ciJqaGuzcuRPvv/8+UlNTsWjRIrlMdnY2Jk6ciDFjxiAzMxNz587Fww8/jI0bN3b4+baFY3mlyMpzhiCHAJZtyFK4RkRERJ2fJIQQSleivpCQELz88sv405/+hG7dumHNmjX405/+BAD45Zdf0L9/f6Snp2PUqFH49ttvcfPNN+Ps2bOIiIgAAKxevRpPPvkkzp07B51OhyeffBLr16/HoUOH5M+YPHkyiouLsWHDhibVyWKxIDAwECUlJTAajW1/0s2wYtOv+HvaMfSPNCLLbIFDAP+elYhhcSGK1ouIiMjTNOf722PGENntdnzyyScoLy9HYmIiMjIyYLVakZSUJJfp168fYmNjkZ6eDgBIT0/HwIED5TAEAMnJybBYLHIrU3p6utsxXGVcx2hIdXU1LBaL28MTCCHw9YGzAICZN8TjruExAIBXNx1TslpERESdnuKB6ODBg/D394der8ejjz6KtWvXIiEhAWazGTqdDkFBQW7lIyIiYDabAQBms9ktDLn2u/Y1VsZisaCysrLBOi1duhSBgYHyIyYmpi1OtdWO5pbit3Pl0GlUSOofgenXxQMA9v1eBIfDoxr6iIiIOhXFA1Hfvn2RmZmJ3bt3Y9asWZg2bRqOHDmiaJ0WLlyIkpIS+XHq1ClF6+Piah0a07cbAny06NnNHzqNClVWB06dr1C4dkRERJ2XRukK6HQ69OrVCwAwbNgw7N27F6+//jruvvtu1NTUoLi42K2VKC8vDyaTCQBgMpmwZ88et+O5ZqHVL3PhzLS8vDwYjUb4+vo2WCe9Xg+9Xt8m59eWdpwoBACMvzISAKBWSegd7o/DZy3IMpciLtRPyeoRERF1Woq3EF3I4XCguroaw4YNg1arRVpamrwvKysLOTk5SExMBAAkJibi4MGDyM/Pl8ts2rQJRqMRCQkJcpn6x3CVcR2jMzld5GwF6h3hL2/rGxEAAPg1j9PviYiIWkrRFqKFCxdi/PjxiI2NRWlpKdasWYMtW7Zg48aNCAwMxPTp0zF//nyEhITAaDRizpw5SExMxKhRowAAY8eORUJCAu677z4sW7YMZrMZzzzzDFJSUuQWnkcffRRvvvkmnnjiCTz00EPYvHkzPvvsM6xfv17JU2+2ihobCstrAADRwQZ5ex+TMxBl5ZUpUi8iIqKuQNFAlJ+fj/vvvx+5ubkIDAzEoEGDsHHjRvzhD38AALz66qtQqVSYNGkSqqurkZycjLfeekt+v1qtxtdff41Zs2YhMTERfn5+mDZtGp5//nm5THx8PNavX4958+bh9ddfR3R0NN555x0kJyd3+Pm2xtli5wDwAB8NAn218na5hcjMFiIiIqKW8rh1iDyRJ6xD9ENWPh58by/6mQKwYe4N8vYzxZW49sXN0KgkHHl+HHQaj+sFJSIiUkRzvr8VH1RNztuPFBQUNFom/Xg5ACBAVYOffvpJ3i6EgEErocIqsP7HvYgL1F7qEBcJCwtDbGxsyypNRETUhTAQKSwnJwf9+vdHZUXj0+aD/msaAkfdibSvPsf/PfZPt30RU5bBJzoBU1OeQMXRbU3+bF+DAb8cPcpQREREXo+BSGEFBQWorKjAlCdfRkTsFZcst7tAjdMVwHVJE9D7jnFu+34qUiO7DLh+6uMYEDS3SZ+bl3MCH720AAUFBQxERETk9RiIPERE7BWI7j3gkvt3FJ8CUIWYmBhEh/u77Ss4VYzsX8+hRmdEdO+odq4pERFR18MRuJ2EpcoKADD6XJxhQ/10ACBPyyciIqLmYSDqBGx2Bypq7AAAo+/Fg6ZD/Z2BqKTSCqvd0aF1IyIi6goYiDqB0iobAECrlqBvYFq9QaeBT+32kkprh9aNiIioK2Ag6gTqusu0kCSpwTIBPs6WI1d4IiIioqZjIOoELLUhp6HuMhf/2rFFpVVsISIiImouBqJOwFJ56QHVLgG1+8qq2UJERETUXAxEnYCrG8zVLdaQALmFiIGIiIiouRiIOoHGpty7BOg5hoiIiKilGIg6ATkQNTKGKIBjiIiIiFqMgcjD2R0C5dXONYgCmjiGyCFEh9SNiIioq2Ag8nBVVmcYkiTAV6u+ZDk/nQaSBDgE5EUciYiIqGkYiDxcZW0g8tGoL7kGEQCoVBL89ew2IyIiagkGIg9XWdva46u7dOuQiysQlXFgNRERUbMwEHk4V5dZY91lLpx6T0RE1DIMRB5O7jLTXv5S8fYdRERELcNA5OEqW9JCVM0xRERERM3BQOThqmocAACfpgQiPbvMiIiIWoKByMPJLURNGFTNLjMiIqKWYSDycC3pMqu02mGzO9q1XkRERF0JA5GHq5IHVV8+EOk1KmjVzrWKSnnXeyIioiZjIPJwzWkhkiSJN3klIiJqAQYiD1fVjDFEAG/ySkRE1BIMRB7MZnfAanfeqLUp6xABgF/tTLNy3s+MiIioyRiIPJiru0wlATp1UwORsyWpgmOIiIiImoyByINVWevWIGrsxq71+elqW4iq2UJERETUVAxEHqw5A6pdDLVjjcpr2EJERETUVAxEHky+030zApE8hohdZkRERE3GQOTB5DWImjjDDHAfVC2EaJd6ERERdTUMRB6sJV1mfrXhye4QqLFxtWoiIqKmYCDyYC0JRBq1CjqN87Jy6j0REVHTMBB5sKoa1207mneZXK1EHEdERETUNAxEHqzS1rxVql3qxhExEBERETUFA5EHq6pxjgFqTpcZwLWIiIiImouByINVNuNO9/W5VqtmCxEREVHTMBB5KCFEiwZVA/VbiBiIiIiImoKByEPZHAJ2h+vGrs0LRAb5fmbsMiMiImoKBiIP5VqlWq2SoFU37T5mLnILEbvMiIiImoSByEPV7y5r6o1dXepu38EWIiIioqZgIPJQ8m07mrkGEVA3qLrG7oDVztWqiYiILoeByENVWZ1BxkfTvPFDAKBTq6BROVuVOLCaiIjo8hiIPJTrPmSu23A0hyRJbjd5JSIiosYxEHmoarszyOhb0GUGAAada6YZW4iIiIguh4HIQ1XXdpnp1c3vMgPAFiIiIqJmUDQQLV26FFdffTUCAgIQHh6O2267DVlZWW5lRo8eDUmS3B6PPvqoW5mcnBxMnDgRBoMB4eHhWLBgAWw295aRLVu2YOjQodDr9ejVqxdSU1Pb+/RapTVdZgBv8EpERNQcigairVu3IiUlBbt27cKmTZtgtVoxduxYlJeXu5WbMWMGcnNz5ceyZcvkfXa7HRMnTkRNTQ127tyJ999/H6mpqVi0aJFcJjs7GxMnTsSYMWOQmZmJuXPn4uGHH8bGjRs77Fybq7o2ELW0y4w3eCUiImo6jZIfvmHDBrfXqampCA8PR0ZGBm644QZ5u8FggMlkavAY3333HY4cOYLvv/8eERERGDJkCF544QU8+eSTWLJkCXQ6HVavXo34+HgsX74cANC/f39s374dr776KpKTk9vvBFuhuvZO9y1vIXJeWq5WTUREdHkeNYaopKQEABASEuK2/aOPPkJYWBiuvPJKLFy4EBUVFfK+9PR0DBw4EBEREfK25ORkWCwWHD58WC6TlJTkdszk5GSkp6c3WI/q6mpYLBa3R0dzdZnpWxiIDLzBKxERUZMp2kJUn8PhwNy5c3HttdfiyiuvlLffe++9iIuLQ1RUFA4cOIAnn3wSWVlZ+OKLLwAAZrPZLQwBkF+bzeZGy1gsFlRWVsLX19dt39KlS/Hcc8+1+Tk2h9xl1oJ1iIB6LUQcVE1ERHRZHhOIUlJScOjQIWzfvt1t+8yZM+XnAwcORGRkJG666SacOHECV1xxRbvUZeHChZg/f7782mKxICYmpl0+61KqWzmo2jXtvrLGDocQUDXz9h9ERETexCO6zGbPno2vv/4aP/zwA6KjoxstO3LkSADA8ePHAQAmkwl5eXluZVyvXeOOLlXGaDRe1DoEAHq9Hkaj0e3R0VrbZeardQYigbobxRIREVHDFA1EQgjMnj0ba9euxebNmxEfH3/Z92RmZgIAIiMjAQCJiYk4ePAg8vPz5TKbNm2C0WhEQkKCXCYtLc3tOJs2bUJiYmIbnUnbcgiBGnvrApFKJcmhiN1mREREjVM0EKWkpODDDz/EmjVrEBAQALPZDLPZjMrKSgDAiRMn8MILLyAjIwMnT57EV199hfvvvx833HADBg0aBAAYO3YsEhIScN999+Hnn3/Gxo0b8cwzzyAlJQV6vR4A8Oijj+K3337DE088gV9++QVvvfUWPvvsM8ybN0+xc2+M1VZ3Q9aWdpkBdTd5reDAaiIiokYpGohWrVqFkpISjB49GpGRkfLj008/BQDodDp8//33GDt2LPr164fHH38ckyZNwrp16+RjqNVqfP3111Cr1UhMTMTUqVNx//334/nnn5fLxMfHY/369di0aRMGDx6M5cuX45133vHgKffOQKRWSdCoWn6JDBxYTURE1CSKDqoWQjS6PyYmBlu3br3sceLi4vDNN980Wmb06NHYv39/s+qnlOpWjh9ycQ2s5tR7IiKixnnEoGpy19oB1S5cnJGIiKhpGIg8UGtXqXbh4oxERERNw0DkgVq7KKOLq8uMY4iIiIgax0DkgdpuDBEHVRMRETUFA5EHqmnlKtUufq4Womp2mRERETWGgcgDucYQtbqFSO9sIaqyOWB3ND6jj4iIyJsxEHmgthpD5KNRQVV7CzMuzkhERHRpDEQeqK2m3UuSxHFERERETcBA5IFae6f7+rg4IxER0eUxEHmgthpDBNSbes/FGYmIiC6JgcgD1bTRGCKAU++JiIiagoHIA7VllxnveE9ERHR5DEQeqK0WZgTqWojK2UJERER0SQxEHsbmqFszqC0CERdnJCIiujwGIg/jGj8EANo2bCHiGCIiIqJLYyDyMPL4IbUKKklq9fEMet7glYiI6HIYiDxMWw6oBuqm3dfYHbDaHZcpTURE5J0YiDxMW61S7aJTq6CpvX8HW4mIiIgaxkDkYdpyUUbAdfsOTr0nIiJqDAORh2nrLjOg3tR7rlZNRETUIAYiDyN3mWlbv0q1CxdnJCIiahwDkYeRF2VUt0MLEccQERERNYiByMPUtEuXGVuIiIiIGsNA5GHaNRBxDBEREVGDGIg8TI29bmHGtuKn52rVREREjWEg8jDt2UJUzi4zIiKiBjEQeRjXatLtMe2+osYOIUSbHZeIiKirYCDyMK4WIm2bzjJzthDZHULukiMiIqI6DEQepqYdWoi0apU8JokDq4mIiC7GQORhamxtP6gaqLvrPccRERERXYyByIMIIdpllhlQfy0ithARERFdiIHIg9gdAq4xz23ZZQYAfjpOvSciIroUBiIPUn/As1Yttemx5an31ewyIyIiuhADkQepm2EmQZLaOBBxcUYiIqJLYiDyIO0xw8yF9zMjIiK6NAYiD2K1OQcQtfWAaoBjiIiIiBrDQORBqu3OsNKWizK68PYdREREl8ZA5EHkFqJ26DLz4+07iIiILomByIO01xpEAOBb20IkBFBl5e07iIiI6mMg8iDtcad7F7VKgo/WeVx2mxEREbljIPIg7dlCBNR1m3EtIiIiIncMRB7E6lqHqB1aiADAr3YtonLONCMiInLDQORB2nMdIgDw03O1aiIiooYwEHmQ9rrTvQu7zIiIiBrGQORB2n0MkavLrJpdZkRERPUxEHmQ9pxlBtTrMuMsMyIiIjcMRB7Eaq+7uWt7YJcZERFRwxQNREuXLsXVV1+NgIAAhIeH47bbbkNWVpZbmaqqKqSkpCA0NBT+/v6YNGkS8vLy3Mrk5ORg4sSJMBgMCA8Px4IFC2CzuX/pb9myBUOHDoVer0evXr2Qmpra3qfXbO3dQuRfr8uMi1UTERHVUTQQbd26FSkpKdi1axc2bdoEq9WKsWPHory8XC4zb948rFu3Dp9//jm2bt2Ks2fP4o477pD32+12TJw4ETU1Ndi5cyfef/99pKamYtGiRXKZ7OxsTJw4EWPGjEFmZibmzp2Lhx9+GBs3buzQ872c9h5D5LqfmV0IcLFqIiKiOholP3zDhg1ur1NTUxEeHo6MjAzccMMNKCkpwb/+9S+sWbMGN954IwDgvffeQ//+/bFr1y6MGjUK3333HY4cOYLvv/8eERERGDJkCF544QU8+eSTWLJkCXQ6HVavXo34+HgsX74cANC/f39s374dr776KpKTkzv8vC+lPe9lBgAatQo+GhWqbA5U2tunW46IiKgz8qgxRCUlJQCAkJAQAEBGRgasViuSkpLkMv369UNsbCzS09MBAOnp6Rg4cCAiIiLkMsnJybBYLDh8+LBcpv4xXGVcx7hQdXU1LBaL26O92YWz5QZovxYioG6mWRUnmhEREck8JhA5HA7MnTsX1157La688koAgNlshk6nQ1BQkFvZiIgImM1muUz9MOTa79rXWBmLxYLKysqL6rJ06VIEBgbKj5iYmDY5x8bY6nVhtddK1UD9QMQWIiIiIhePCUQpKSk4dOgQPvnkE6WrgoULF6KkpER+nDp1qt0/s7a3DBqVBJXUfmHFNfWeLURERER1FB1D5DJ79mx8/fXX2LZtG6Kjo+XtJpMJNTU1KC4udmslysvLg8lkksvs2bPH7XiuWWj1y1w4My0vLw9GoxG+vr4X1Uev10Ov17fJuTWV1eEMQe01fsjFNfWeY4iIiIjqKNpCJITA7NmzsXbtWmzevBnx8fFu+4cNGwatVou0tDR5W1ZWFnJycpCYmAgASExMxMGDB5Gfny+X2bRpE4xGIxISEuQy9Y/hKuM6hidwdZm15/ghgF1mREREDVG0hSglJQVr1qzBf/7zHwQEBMhjfgIDA+Hr64vAwEBMnz4d8+fPR0hICIxGI+bMmYPExESMGjUKADB27FgkJCTgvvvuw7Jly2A2m/HMM88gJSVFbuV59NFH8eabb+KJJ57AQw89hM2bN+Ozzz7D+vXrFTv3C7m6zNq/hYhdZkRERBdStIVo1apVKCkpwejRoxEZGSk/Pv30U7nMq6++iptvvhmTJk3CDTfcAJPJhC+++ELer1ar8fXXX0OtViMxMRFTp07F/fffj+eff14uEx8fj/Xr12PTpk0YPHgwli9fjnfeecezptzXdplpO6iFiF1mREREdRRtIRJNWC7Zx8cHK1euxMqVKy9ZJi4uDt98802jxxk9ejT279/f7Dp2lA5rIeK0eyIioot4zCwzb9dhY4hqu8wckKDy8W/XzyIiIuosGIg8hLWDWog0ahX0tZ+h9g9p188iIiLqLBiIPITNNe2+nVuIgLqbvDIQEREROTEQeQjXGCKtpv0HOxtqF2dU+zEQERERAQxEHsPaQWOIAMC/dnFGdUBou38WERFRZ8BA5CFsHbRSNQD4+zgDkYaBiIiICAADkceQp9135BiigLB2/ywiIqLOgIHIQ8hdZh3YQsRARERE5MRA5CFsomNWqgaAAL0WALvMiIiIXBiIPIStI1uIXF1mfsGw2i+/WjgREVFXx0DkIawdOIbIR6uCSnJ+YGEl7+FBRETEQOQJVGo4RMfNMpMkCb7OpYgYiIiIiMBA5BFUOl/5eUeMIQIAg7q2hajC0SGfR0RE5MkYiDyAVBuI1CoJalX7r1QNAL5qdpkRERG5MBB5AJXOAKBjxg+5+DrHVaOwgoGIiIiIgcgDqPTOFqKOGD/k4mohKmALEREREQORJ5CUaCHiGCIiIiIZA5EHcA2q7og73bvIXWZsISIiImIg8gRKtBC5ZpmVVDlQY2MrEREReTcGIg/gaiHqyDFEOhUgbFYIAPmlVR32uURERJ6IgcgDyIGoA1uIJAmwlRYAAMwlDEREROTdGIg8gKSv7TLrwBYiALDXBqKzDEREROTlWvQN3LNnTxQWFl60vbi4GD179mx1pbyNPKi6A1uIgPotRJUd+rlERESepkXfwCdPnoTdfvHspOrqapw5c6bVlfI2kgJjiIC6FqJcthAREZGX0zSn8FdffSU/37hxIwIDA+XXdrsdaWlp6NGjR5tVzlsosVI1ANhLna18ucUMRERE5N2aFYhuu+02AM67pU+bNs1tn1arRY8ePbB8+fI2q5y3UGKWGQDYLPkAgLPsMiMiIi/XrEDkcDjXq4mPj8fevXsRFhbWLpXyNpJSY4hKnIHo9HkGIiIi8m7NCkQu2dnZbV0Pr6ZSaJaZKxAVldegvNoGP32L/jkQERF1ei3+BkxLS0NaWhry8/PlliOXd999t9UV8yaSAusQAYCoqYC/TkJZjcCZ4kr0iQjo0M8nIiLyFC36Bn7uuecwduxYpKWloaCgAOfPn3d7UPMoNYYIALoZ1ACA0+crOvyziYiIPEWLWohWr16N1NRU3HfffW1dH69jdwhFVqp2CfdTI7vYxnFERETk1Vr0DVxTU4NrrrmmrevilapsQn7ekXe7d+lmcGZiBiIiIvJmLQpEDz/8MNasWdPWdfFKlbWBSIKARqVMCxHALjMiIvJuLeoyq6qqwj//+U98//33GDRoELRardv+FStWtEnlvEGl1TkgXavQXeXqAhFbiIiIyHu1KBAdOHAAQ4YMAQAcOnTIbZ8kdXy3T2fmaiFSoLcMANCtNhCdYSAiIiIv1qJA9MMPP7R1PbxWhbU2EKnEZUq2j/DaWWaF5TWoqLHBoONaRERE5H0U6qghlyqbsl1mfjoVjD7OEMRWIiIi8lYtag4YM2ZMo11jmzdvbnGFvI3SXWYAEB1swJFcC06fr0RvLs5IREReqEWByDV+yMVqtSIzMxOHDh266Kav1DhXl5lWoS4zAIgO9q0NRJxpRkRE3qlFgejVV19tcPuSJUtQVlbWqgp5m0qr8i1E3YOdC0NyphkREXmrNh25MnXqVN7HrJlcY4gUuGuHLDrYeXNZBiIiIvJWbfo1nJ6eDh8fn7Y8ZJcnd5kpOobI1ULELjMiIvJOLeoyu+OOO9xeCyGQm5uLffv24dlnn22TinkLeVC1wmOIAOAUW4iIiMhLtSgQBQYGur1WqVTo27cvnn/+eYwdO7ZNKuYtKl1dZgq2EMWEOLvMisprUFplRYCP9jLvICIi6lpaFIjee++9tq6H16qUZ5kpVwejjxahfjoUltfg98IKXNk98PJvIiIi6kJatSxxRkYGjh49CgAYMGAArrrqqjaplDfxhC4zAIgLNaCwvAYnC8sZiIiIyOu0KBDl5+dj8uTJ2LJlC4KCggAAxcXFGDNmDD755BN069atLevYpblu7qpklxkA9Ajzw085xfi9kAOriYjI+7Soo2bOnDkoLS3F4cOHUVRUhKKiIhw6dAgWiwX//d//3eTjbNu2DbfccguioqIgSRK+/PJLt/0PPPAAJElye4wbN86tTFFREaZMmQKj0YigoCBMnz79orWQDhw4gOuvvx4+Pj6IiYnBsmXLWnLa7cLVQqRklxkA9Aj1AwBkF5QrWxEiIiIFtOhreMOGDXjrrbfQv39/eVtCQgJWrlyJb7/9tsnHKS8vx+DBg7Fy5cpLlhk3bhxyc3Plx8cff+y2f8qUKTh8+DA2bdqEr7/+Gtu2bcPMmTPl/RaLBWPHjkVcXBwyMjLw8ssvY8mSJfjnP//ZjDNuP3ULMyrbZdYjzBmIfi9kICIiIu/Toi4zh8MBrfbimUharRYOh6PJxxk/fjzGjx/faBm9Xg+TydTgvqNHj2LDhg3Yu3cvhg8fDgB44403MGHCBLzyyiuIiorCRx99hJqaGrz77rvQ6XQYMGAAMjMzsWLFCrfgpASHQ6BKHkOkaFUQL7cQscuMiIi8T4u+hm+88UY89thjOHv2rLztzJkzmDdvHm666aY2qxwAbNmyBeHh4ejbty9mzZqFwsJCeV96ejqCgoLkMAQASUlJUKlU2L17t1zmhhtugE6nk8skJycjKysL58+fb/Azq6urYbFY3B7tocJqh6tdSMmFGQEgLsw59b6grBqlVVZlK0NERNTBWhSI3nzzTVgsFvTo0QNXXHEFrrjiCsTHx8NiseCNN95os8qNGzcOH3zwAdLS0vDSSy9h69atGD9+POx2OwDAbDYjPDzc7T0ajQYhISEwm81ymYiICLcyrteuMhdaunQpAgMD5UdMTEybnVN95dU2AIBw2KFSOBC5pt4D4MBqIiLyOi3qMouJicFPP/2E77//Hr/88gsAoH///khKSmrTyk2ePFl+PnDgQAwaNAhXXHEFtmzZ0uYtUfUtXLgQ8+fPl19bLJZ2CUUGnRqPDDNi6UuvQHrogTY/fnNx6j0REXmrZrUQbd68GQkJCbBYLJAkCX/4wx8wZ84czJkzB1dffTUGDBiAH3/8sb3qip49eyIsLAzHjx8HAJhMJuTn57uVsdlsKCoqkscdmUwm5OXluZVxvb7U2CS9Xg+j0ej2aA8BPlokX+GH0oyv2uX4zVU3sJotRERE5F2aFYhee+01zJgxo8GAEBgYiEceeQQrVqxos8pd6PTp0ygsLERkZCQAIDExEcXFxcjIyJDLbN68GQ6HAyNHjpTLbNu2DVZr3biYTZs2oW/fvggODm63unZGnHpPRETeqlmB6Oeff75oHaD6xo4d6xZOLqesrAyZmZnIzMwEAGRnZyMzMxM5OTkoKyvDggULsGvXLpw8eRJpaWm49dZb0atXLyQnJwNwdtONGzcOM2bMwJ49e7Bjxw7Mnj0bkydPRlRUFADg3nvvhU6nw/Tp03H48GF8+umneP311926xMiJU++JiMhbNSsQ5eXlNTjd3kWj0eDcuXNNPt6+fftw1VVXybf8mD9/Pq666iosWrQIarUaBw4cwB//+Ef06dMH06dPx7Bhw/Djjz9Cr9fLx/joo4/Qr18/3HTTTZgwYQKuu+46tzWGAgMD8d133yE7OxvDhg3D448/jkWLFik+5d4Tceo9ERF5q2YNqu7evTsOHTqEXr16Nbj/wIEDcndWU4wePRpCXHpBwo0bN172GCEhIVizZk2jZQYNGtSuY5u6ivpT78uqbfDXt+pWd0RERJ1Gs1qIJkyYgGeffRZVVVUX7ausrMTixYtx8803t1nlqGPVn3p/kuOIiIjIizSrCeCZZ57BF198gT59+mD27Nno27cvAOCXX37BypUrYbfb8fTTT7dLRalj9Ozmh8LyGpw4V8ap90RE5DWaFYgiIiKwc+dOzJo1CwsXLpS7uyRJQnJyMlauXHnRIojUufQKD8Dek+dxLK/s8oWJiIi6iGYPEomLi8M333yD8+fP4/jx4xBCoHfv3pzC3kX0DvcHABzLL1W4JkRERB2nxaNmg4ODcfXVV7dlXcgD9I5wBSK2EBERkfdQ+B7r5Gl6hwcAcK5WXW2zK1wbIiKijsFARG4ijHoE+Ghgdwic5HpERETkJRiIyI0kSRxHREREXoeBiC7i6jbjTDMiIvIWDER0kbqB1WwhIiIi78BARBfp5eoyYwsRERF5CQYiukjvCGeXWXZBOax2h8K1ISIian8MRHSRqEAf+OnUsDkEfi/kPc2IiKjrYyCii0iShF4RHFhNRETeg4GIGuSaep+Vx4HVRETU9TEQUYP6mZwtREdzLQrXhIiIqP0xEFGDEqKMAIAjDEREROQFGIioQQMiAwEAp4oqUVJpVbg2RERE7YuBiBoUaNCie5AvAHabERFR18dARJckd5udZSAiIqKujYGILmlAbSA6zEBERERdHAMRXVJCJAdWExGRd2AgoktydZkdyytFtc2ucG2IiIjaDwMRXVL3IF8E+mphcwiuWE1ERF0aAxFdkiRJ7DYjIiKvwEBEjRrAmWZEROQFGIioUa5xRIfOlChcEyIiovbDQESNGhTtXLH60NkSWO0OhWtDRETUPhiIqFE9w/wR4KNBldWBLDPvfE9ERF0TAxE1SqWSMDg6CACQeapY0boQERG1FwYiuqwhMUEAGIiIiKjrYiCiy2IgIiKiro6BiC5rSGwQAODEuTJYqqzKVoaIiKgdMBDRZYX56xEd7AshgAOnOP2eiIi6HgYiapK6brPzylaEiIioHTAQUZNcFRsMgOOIiIioa2IgoiapP7BaCKFsZYiIiNoYAxE1yYAoI3RqFQrKapBTVKF0dYiIiNoUAxE1iY9WjYG1t/HYk12kcG2IiIjaFgMRNdmI+BAADERERNT1MBBRk43o4QxEe08yEBERUdfCQERNNqxHMCQJOFlYgXxLldLVISIiajMMRNRkRh8t+puMAIA9bCUiIqIuhIGImoXjiIiIqCtiIKJmYSAiIqKuiIGImuXq2oHVWXmlKK6oUbg2REREbYOBiJqlW4AePcP8IASw9yTva0ZERF2DRskP37ZtG15++WVkZGQgNzcXa9euxW233SbvF0Jg8eLFePvtt1FcXIxrr70Wq1atQu/eveUyRUVFmDNnDtatWweVSoVJkybh9ddfh7+/v1zmwIEDSElJwd69e9GtWzfMmTMHTzzxREeeqsc6evRos9/TyyjwWwHwZfoRhFadafb7w8LCEBsb2+z3ERERtRdFA1F5eTkGDx6Mhx56CHfcccdF+5ctW4a///3veP/99xEfH49nn30WycnJOHLkCHx8fAAAU6ZMQW5uLjZt2gSr1YoHH3wQM2fOxJo1awAAFosFY8eORVJSElavXo2DBw/ioYceQlBQEGbOnNmh5+tJLEXnAABTp05t9nsNfa9Ft9sWYu3Oo3jr4ZRmv9/XYMAvR48yFBERkcdQNBCNHz8e48ePb3CfEAKvvfYannnmGdx6660AgA8++AARERH48ssvMXnyZBw9ehQbNmzA3r17MXz4cADAG2+8gQkTJuCVV15BVFQUPvroI9TU1ODdd9+FTqfDgAEDkJmZiRUrVnh1IKosswAAJj7yNPoOGtas91bbga/PCOi6xeHPf/8CPuqmvzcv5wQ+emkBCgoKGIiIiMhjKBqIGpOdnQ2z2YykpCR5W2BgIEaOHIn09HRMnjwZ6enpCAoKksMQACQlJUGlUmH37t24/fbbkZ6ejhtuuAE6nU4uk5ycjJdeegnnz59HcHDwRZ9dXV2N6upq+bXFYmmns1ReaFQconsPaPb7ullycK60GragGETXrk1ERETUWXnsoGqz2QwAiIiIcNseEREh7zObzQgPD3fbr9FoEBIS4lamoWPU/4wLLV26FIGBgfIjJiam9SfUxcQGGwAAp4oqFa4JERFR63lsIFLSwoULUVJSIj9OnTqldJU8TkyILwAgp6gCQgiFa0NERNQ6HhuITCYTACAvL89te15enrzPZDIhPz/fbb/NZkNRUZFbmYaOUf8zLqTX62E0Gt0e5C4qyBdqSUJZtQ3FlValq0NERNQqHhuI4uPjYTKZkJaWJm+zWCzYvXs3EhMTAQCJiYkoLi5GRkaGXGbz5s1wOBwYOXKkXGbbtm2wWuu+tDdt2oS+ffs2OH6ImkarViEy0DnT71RRhcK1ISIiah1FA1FZWRkyMzORmZkJwDmQOjMzEzk5OZAkCXPnzsVf//pXfPXVVzh48CDuv/9+REVFyWsV9e/fH+PGjcOMGTOwZ88e7NixA7Nnz8bkyZMRFRUFALj33nuh0+kwffp0HD58GJ9++ilef/11zJ8/X6Gz7jpiQpzjiHIYiIiIqJNTdJbZvn37MGbMGPm1K6RMmzYNqampeOKJJ1BeXo6ZM2eiuLgY1113HTZs2CCvQQQAH330EWbPno2bbrpJXpjx73//u7w/MDAQ3333HVJSUjBs2DCEhYVh0aJFXj3lvq3EhRqQ/lshcooqYHcIqFWS0lUiIiJqEUUD0ejRoxsdkCtJEp5//nk8//zzlywTEhIiL8J4KYMGDcKPP/7Y4npSw8ID9DDo1KioseNscaXcYkRERNTZeOwYIvJ8kiQhLtQZgk4WlitcGyIiopZjIKJW6RHqBwA4WchxRERE1HkxEFGrxIYYIAEoKq+BpYrT74mIqHNiIKJW8dGqYaqdfv97AVuJiIioc2Igolar6zbjOCIiIuqcGIio1eLDnIEop6gCNrtD4doQERE1HwMRtVqYvw4BPhrYHIKLNBIRUafEQEStJkkSeta2Ev1WwG4zIiLqfBiIqE307OYPAPjtXDkcjSy2SURE5IkYiKhNdA/yhU6jQqXVDnNJldLVISIiahYGImoTapWEHrWrVrPbjIiIOhsGImozV8jdZmUK14SIiKh5GIiozcSFGqCSgPMVVhSV1yhdHSIioiZjIKI2o9eoEVt7x/tjeaUK14aIiKjpGIioTfWOCAAAHMtntxkREXUeDETUpq4I84NKAgrLa1BYVq10dYiIiJqEgYjalF6rRlztvc3YSkRERJ0FAxG1ud7hztlmDERERNRZMBBRm+sZ5ge1JKGI3WZERNRJMBBRm9Nr1YitXaQxi7PNiIioE2AgonbRz+ScbfaLuRSC9zYjIiIPx0BE7aJnmB90ahVKq2w4U1ypdHWIiIgaxUBE7UKjVqF3hHNw9S9mdpsREZFnYyCiduPqNjuWVwab3aFwbYiIiC6NgYjaTfcgXwT4aFBjd+C3gnKlq0NERHRJDETUbiRJkluJDp+1KFwbIiKiS2MgonaVEGkEAOQUVaCk0qpwbYiIiBrGQETtKsigQ2yIc02iQ2dKFK4NERFRwxiIqN0N7B4IwNlt5uCSRERE5IEYiKjdxYf5waBTo9Jqx9kK/pMjIiLPw28nandqlYQBUc6xRL+V8Z8cERF5Hn47UYe4MioQEoBz1Spou8UrXR0iIiI3DETUIYy+WvQKd65cbRxxu8K1ISIicsdARB1maFwwAMCv/w0oqLArXBsiIqI6DETUYUxGH4TpHZDUGnxzjCtXExGR52Agog7VO8DZMvTdbxUoreJCjURE5BkYiKhDRfoKWAtPocIq8O72k0pXh4iICAADEXUwSQKKt68BALyz/TeUVLCViIiIlMdARB2u4pftiAvUoLTKhrd//E3p6hARETEQkRIEJl8ZAAB4d0c2CsuqFa4PERF5OwYiUsSIKD0Gdg9ERY0dr31/TOnqEBGRl2MgIkVIkoT/mdAfAPDR7t9x5KxF4RoREZE3YyAixSReEYqJgyLhEMCSrw5DCKF0lYiIyEsxEJGinp7QH75aNfacLMJ/Ms8qXR0iIvJSDESkqKggX8y+sRcA4Ll1h5FfWqVwjYiIyBsxEJHiZlzfEwmRRpyvsOJ/vjjIrjMiIupwDESkOJ1GhRV3D4ZOrcL3R/PxecZppatERERexqMD0ZIlSyBJktujX79+8v6qqiqkpKQgNDQU/v7+mDRpEvLy8tyOkZOTg4kTJ8JgMCA8PBwLFiyAzWbr6FOhy+hnMmLeH/oAcA6wPpZXqnCNiIjIm3h0IAKAAQMGIDc3V35s375d3jdv3jysW7cOn3/+ObZu3YqzZ8/ijjvukPfb7XZMnDgRNTU12LlzJ95//32kpqZi0aJFSpwKXcbMG3rimitCUVFjx8z/zYCFN38lIqIO4vGBSKPRwGQyyY+wsDAAQElJCf71r39hxYoVuPHGGzFs2DC899572LlzJ3bt2gUA+O6773DkyBF8+OGHGDJkCMaPH48XXngBK1euRE1NjZKnRQ1QqyS8cc9V6B7ki+yCcsz/NBMOB8cTERFR+/P4QHTs2DFERUWhZ8+emDJlCnJycgAAGRkZsFqtSEpKksv269cPsbGxSE9PBwCkp6dj4MCBiIiIkMskJyfDYrHg8OHDl/zM6upqWCwWtwd1jFB/PVZPHQadxjme6Ll1XJ+IiIjan0cHopEjRyI1NRUbNmzAqlWrkJ2djeuvvx6lpaUwm83Q6XQICgpye09ERATMZjMAwGw2u4Uh137XvktZunQpAgMD5UdMTEzbnhg1amB0IF65czAA4P303/Hm5uMK14iIiLo6jdIVaMz48ePl54MGDcLIkSMRFxeHzz77DL6+vu32uQsXLsT8+fPl1xaLhaGog/1xcBQKy6rx3LojWL7pVwT4aPDAtfFKV4uIiLooj24hulBQUBD69OmD48ePw2QyoaamBsXFxW5l8vLyYDKZAAAmk+miWWeu164yDdHr9TAajW4P6ngPXhuPObWLNi5ZdwTv/PibwjUiIqKuyqNbiC5UVlaGEydO4L777sOwYcOg1WqRlpaGSZMmAQCysrKQk5ODxMREAEBiYiL+9re/IT8/H+Hh4QCATZs2wWg0IiEhQbHzIODo0aNNKjc6VOBsfz/8+2g5/rr+KE78fgp/6u8PSZKa/ZlhYWGIjY1t9vuIiKjr8+hA9Je//AW33HIL4uLicPbsWSxevBhqtRr33HMPAgMDMX36dMyfPx8hISEwGo2YM2cOEhMTMWrUKADA2LFjkZCQgPvuuw/Lli2D2WzGM888g5SUFOj1eoXPzjtZis4BAKZOndqs9wVeew+CrpuCjw+V4Z//+xmKNq0GhKNZx/A1GPDL0aMMRUREdBGPDkSnT5/GPffcg8LCQnTr1g3XXXcddu3ahW7dugEAXn31VahUKkyaNAnV1dVITk7GW2+9Jb9frVbj66+/xqxZs5CYmAg/Pz9MmzYNzz//vFKn5PUqy5wz9iY+8jT6DhrWrPceL7Xh5/NqBFw1AX2uGYcRoTZomtjpm5dzAh+9tAAFBQUMREREdBGPDkSffPJJo/t9fHywcuVKrFy58pJl4uLi8M0337R11aiVQqPiEN17QLPeEw2ge34pNh7OQ26lCrstRtwyOBIGnUf/MyYiok6gUw2qJuodHoA7ruoOvUYFs6UKn+07jeIKLrJJREStw0BEnU5UkC/uGh6DAB8NSiqt+GzfaeSWVCpdLSIi6sQYiKhTCvHT4e7hMQgP0KPSase/fzqDY/m8ISwREbUMAxF1Wn56DSYNjUZ8mB/sDoFvDprxU8553uqDiIiajYGIOjWdRoWbB0ZiUPdAAMCPxwqw7VgBQxERETULAxF1eiqVhNF9u+H6XmEAgMxTxfjuSB7sDoYiIiJqGgYi6hIkScLQuGAkD4iAJAG/mEux/mAubPbmLd5IRETeiYGIupR+JiNuHhQJtUpCdkE5vsw8i2qbXelqERGRh2Mgoi6nZ5g/bh/SHTq1CmeKK/Hvn86gipmIiIgawUBEXVL3YF9MGtYdvlo1zpVWY1ueFmr/EKWrRUREHoqBiLqs8AAf3Dk8Gv56DUptEiLufRH55Talq0VERB6IgYi6tGCDDncOi4ZBLaANjsIzPxQiu6Bc6WoREZGHYSCiLs/oq8XoCCushadQUOHAXf9Ix695XNWaiIjqMBCRV/DVAOY1CxEXqMG50mrc/Y90HDpTonS1iIjIQzAQkddwVBTj+dGhGBQdiPMVVtzz9i78lHNe6WoREZEHYCAirxKgV+HDh0dieFwwSqtsuO+d3dj1W6HS1SIiIoUxEJHXMfpo8cH0Ebi2VyjKa+yY9u4ebP31nNLVIiIiBTEQkVcy6DT417SrcWO/cFTbHJjx/j6s+/ms0tUiIiKFMBCR1/LRqrF66jBMGGhCjd2BOR/vx8ofjkMI3hSWiMjbMBCRV9NpVHjjnqF46Np4AMDLG7Pw1L8PwsqbwhIReRUGIvJ6apWERbck4Lk/DoBKAj7ddwoPpe6FpcqqdNWIiKiDMBAR1Zp2TQ+8ff9wGHRq/HisAHe8tRPHuIAjEZFXYCAiquem/hH47JFERBj1OJ5fhj++uQNf/HRa6WoREVE7YyAiusCV3QOx/r+vx3W9wlBptWP+Zz/jqX8fQJXVrnTViIionWiUrgBRRzp69GiTyz42RIPuen98drgMn+w9hfRfczF/VBBiArXN+sywsDDExsY2t6pERNSBGIjIK1iKnAsvTp06tdnv9YkbjLBb/oLfEYz/Xp+L4u0fwrJnLSCaNhPN12DAL0ePMhQREXkwBiLyCpVlFgDAxEeeRt9Bw5r/fhvwU5ED5iotgkc/iJ5jp2F4qA3GyzQW5eWcwEcvLUBBQQEDERGRB2MgIq8SGhWH6N4DWvTeXkLgqLkUW389h/M1wGazHiN6hmBobBA0Kg7HIyLqzPhXnKiJJElCQqQR942MQ49QA+xCIP1EIT7clYPfzpVxhWsiok6MgYiomfx9NPjj4CgkD4iAn06Nkkor1h3IxX9+Poui8hqlq0dERC3ALjOiFpAkCf1MRvQM88eek0XIzCnG74UV+KjodwyICsTVPYIR4NO82WhERKQcBiKiVtBpVLiuVxiujDJi27ECZBeU4+CZEhw5a0FClBGRvCUaEVGnwC4zojYQZNDhj4OjMGlod3QP8oVdCBw8U4LvcrXoNmkR9pypgo03jCUi8lhsISJqQ9HBBkwa6oszxZX4KacY2QXlMPQagRd3nMe/DmxG8oAIjE0wYVTPUOg0/P8RIiJPwUBE1MYkSUJ0sAHRwQYcPXIYn3+5DjE33IlzpdX4cFcOPtyVgwAfDW7sF44xfcMxqmcoTIE+SlebiMirMRARtaMALVC85T1sWJaCSmMMNh7Ow6YjeSgoq8Z/Ms/iP5lnAQBxoQaMig/FyJ4hGNUzFFFBvgrXnIjIuzAQEXUArVrCyL7hGN03HH+97UrszzmPTUfysPNEIQ6fLcHvhRX4vbACn+47BQCICfHF1XEhuCouGMNig9HXFAC1SlL4LIiIui4GIqIOcOFNZVUAkk1AssmA8hofHC2owZFzNTh0rga/nbfiVFElThWdwRf7zwAAfDQSeodo0TdUh75hWvQJ0SFA3/gYJN5Uloio6RiIiNpRS24qK+l8oe/eH/ru/Zw/I/uiCgYczK/Bwfy6hR+thadQfeYXVJ85iuqzv8BacApA3WrZvKksEVHTMRARtaPW3lQWAIQALFYrCqslFNVIKKxWocwmQRsaA21oDPwH/QEAoJUEQvQCIXoHVCVn8e2rf+FNZYmImoiBiKgDtOamsg2prLEj11IJc0kVcourYLZUweoA8qok5FWpAMQiZu4nmLvxHK49eQBDY4MxNC4YPcP8IEkci0REdCEGIqJOyFenRs8wf/QM8wcAOBwCBeXVyC2pQm5JFU4XWFBuUyGnxIacPafw8R7nYO0gg9YZjmKDMDQ2GP0jjQj20yl5KkREHoGBiKgLUKkkhAf4IDzAB4OjgdPHCvHagofw5ifrUawJxv7fi/Hz6WIUV1ix+Zd8bP4lX35veIAefSIC0CciAD27+cFk9IEp0AcRRh+E+umg4uw2IvICDEREXZSjohhGSzZG9vdBcoQe1mHhOFlixa8FVvxSWINjRVbkl9uRX1qN/NJqbD9ecNExNCrAX6dCQO3DXychQO96XvtTL7m97hHVDb179uj4EyYiagUGIqIuqKmz2ySdr3NwdlgcdN3ioAkyQe0fCnVAKNR+QbA5VCiucqC4qjn3YcuHXnMEwQY9ggxaBBt0CDJoEVT7M8SgQ7CfDqF+7j/9dGqObyIixTAQEXVBbTG7zSFsqLIDNQ4JNQ44H3bX8wt+1m6vdgCAhGqbgNniHOzdVFoVEKBXIbC2BcqgU8FfK8Ffp4JB62yd8tOp4Kd1/vTXquCnk+CnVUGjArp163bZGXVCCNgdAjaHQI3dAZtdwGZ3wOoQsNsF1GoJWpUErVoFrUYFH40KGjXvOUfkDRiIiLqwtp7ddjmHd23Bu3+dB7VvAFS+AVD5OH+qfY1Q+fhD5Wt0vjYEQuVrhNpghMo3ECqtHlYHUFTpQFFlc1qj6gjb7zDoD0KvdQYYhwDsDgG7ABxCwO4A7OIyB2mAr6Y2gOkkOYTJ3YU6FaJCjegZbUKwQYtgP2crmEGnYZgi6mS8KhCtXLkSL7/8MsxmMwYPHow33ngDI0aMULpaRF1GVbkFoqYCyQ/Oa1bLlM1Rg+ralqbqei1PVgdgdUiw1m6z1rZUubZZRV0Xm6TRo9IOVNqbF6iEcAB2G4TDDkmlBtQaSFJdkKm0CVTa7CiouNQRSgGcaXCPWgJ0aqneA9CoJKgkQK2SoJYAtSRBpUKDz9UqQCXVlnOVr93m7+eLkKBA+GrV8NNr4O96+Dh/BvhoYPR1dllqGcyILstrAtGnn36K+fPnY/Xq1Rg5ciRee+01JCcnIysrC+Hh4UpXj6hL6aiWKYcQsNoc+GnbBnyxainGTPlvxPW5EgAgSQISAEly3irF7bkEqOo9d1LX/rRBCGcLk03UBTD3n86AVlhYiNM5J50tYHKrmL8cqOzCFaha0DR1WeUALh4I3xB/vQbBfq7xXDpna5ZrTJffxduCDToYGhnTlZOTg4KCiz9bCIEaO1BudaDCKup+1tR/3dC2up/W2ma8+r8xZzemBB+tGn4+OujUKui1Kug1Kug1audPrQo+GjX0WjV8tCr4aNXw0dR7XvtT77at9rlGfUEZFcezeSGvCUQrVqzAjBkz8OCDDwIAVq9ejfXr1+Pdd9/FU089pXDtiKglVJIEvVYNHeywW84hyhSBAQMSOuzzM9K+wv5/P187VqsnAEAIG+wCbg+HkOTnAs6wJeBchdz5WpJfOy746b7dWa7Mch6H92zDuPET4WcMQoVNoMoqUGFzoNLqDGAVVgfKa4SzfLUNZdU2nCqqbPK5ucZ0BehU0KolOVDarFb8kpUFqDSQNDpIWj0kjb7uudT2rVE2h/OcLNUOoMza5sdviE7tbN3Tqupa6DSuFjtVXQueprZVTyM5l7+o35ontwa6tfDV7btc5rLbbNBoLv817fw3JOCoDfIOcbnXdc9FvdcCzv9RgHBArVbLdZfgPG+V5PqfCefzukftPri/Vkl1n29v4PPtDve6hBoNWD5lVOsvXgt5RSCqqalBRkYGFi5cKG9TqVRISkpCenr6ReWrq6tRXV0tvy4pKQEAWCyWNq9bWVkZAOD0scOorrxkm3yby8s5AQAwn/wVJ/wM/Fx+Lj+3FZ9rralu8L9fNeranRokXfCziU4WHETxD+/ikx/evUxJCZLeDypff6h8Apxju1zjunz8ofINgORT91zlEwC1TwAkjRbVAKorG26D0oTEXLxRCIiaKmfLjhBQw+58CIf8XFPvuVrU/dTAAVVtWRUu7vIUkJB/NgcHdqRB0mid4UuthaTWAmotVFqdc5tGB0ld+7N2G9T19tU+VBpd7XH0zp/q2jCnqrtaVbUP6jj28uN4aEgAYmIa+PfVQq7vbSGa0EorvMCZM2cEALFz50637QsWLBAjRoy4qPzixYsFav/HjA8++OCDDz746NyPU6dOXTYreEULUXMtXLgQ8+fPl187HA4UFRUhNDS0zfuVLRYLYmJicOrUKRiNxjY9NrU9Xq/Ohderc+H16nw8/ZoJIVBaWoqoqKjLlvWKQBQWFga1Wo28vDy37Xl5eTCZTBeV1+v10Ov1btuCgoLas4owGo0e+Y+JGsbr1bnwenUuvF6djydfs8DAwCaV84q5mDqdDsOGDUNaWpq8zeFwIC0tDYmJiQrWjIiIiDyBV7QQAcD8+fMxbdo0DB8+HCNGjMBrr72G8vJyedYZEREReS+vCUR33303zp07h0WLFsFsNmPIkCHYsGEDIiIiFK2XXq/H4sWLL+qiI8/E69W58Hp1LrxenU9XumaSEE2Zi0ZERETUdXnFGCIiIiKixjAQERERkddjICIiIiKvx0BEREREXo+BSEErV65Ejx494OPjg5EjR2LPnj1KV8krLFmyBJIkuT369esn76+qqkJKSgpCQ0Ph7++PSZMmXbSoZ05ODiZOnAiDwYDw8HAsWLAANpvNrcyWLVswdOhQ6PV69OrVC6mpqR1xep3etm3bcMsttyAqKgqSJOHLL7902y+EwKJFixAZGQlfX18kJSXh2LFjbmWKioowZcoUGI1GBAUFYfr06fJ9A10OHDiA66+/Hj4+PoiJicGyZcsuqsvnn3+Ofv36wcfHBwMHDsQ333zT5ufb2V3uej3wwAMX/fc2btw4tzK8Xh1n6dKluPrqqxEQEIDw8HDcdtttyMrKcivTkX8DPep7sE1uFkbN9sknnwidTifeffddcfjwYTFjxgwRFBQk8vLylK5al7d48WIxYMAAkZubKz/OnTsn73/00UdFTEyMSEtLE/v27ROjRo0S11xzjbzfZrOJK6+8UiQlJYn9+/eLb775RoSFhYmFCxfKZX777TdhMBjE/PnzxZEjR8Qbb7wh1Gq12LBhQ4eea2f0zTffiKefflp88cUXAoBYu3at2/4XX3xRBAYGii+//FL8/PPP4o9//KOIj48XlZWVcplx48aJwYMHi127dokff/xR9OrVS9xzzz3y/pKSEhERESGmTJkiDh06JD7++GPh6+sr/vGPf8hlduzYIdRqtVi2bJk4cuSIeOaZZ4RWqxUHDx5s999BZ3K56zVt2jQxbtw4t//eioqK3MrwenWc5ORk8d5774lDhw6JzMxMMWHCBBEbGyvKysrkMh31N9DTvgcZiBQyYsQIkZKSIr+22+0iKipKLF26VMFaeYfFixeLwYMHN7ivuLhYaLVa8fnnn8vbjh49KgCI9PR0IYTzC0ClUgmz2SyXWbVqlTAajaK6uloIIcQTTzwhBgwY4Hbsu+++WyQnJ7fx2XRtF37BOhwOYTKZxMsvvyxvKy4uFnq9Xnz88cdCCCGOHDkiAIi9e/fKZb799lshSZI4c+aMEEKIt956SwQHB8vXSwghnnzySdG3b1/59V133SUmTpzoVp+RI0eKRx55pE3PsSu5VCC69dZbL/keXi9l5efnCwBi69atQoiO/Rvoad+D7DJTQE1NDTIyMpCUlCRvU6lUSEpKQnp6uoI18x7Hjh1DVFQUevbsiSlTpiAnJwcAkJGRAavV6nZt+vXrh9jYWPnapKenY+DAgW6LeiYnJ8NiseDw4cNymfrHcJXh9W2d7OxsmM1mt99tYGAgRo4c6XZ9goKCMHz4cLlMUlISVCoVdu/eLZe54YYboNPp5DLJycnIysrC+fPn5TK8hm1jy5YtCA8PR9++fTFr1iwUFhbK+3i9lFVSUgIACAkJAdBxfwM98XuQgUgBBQUFsNvtF62SHRERAbPZrFCtvMfIkSORmpqKDRs2YNWqVcjOzsb111+P0tJSmM1m6HS6i27mW//amM3mBq+da19jZSwWCyorK9vpzLo+1++3sf92zGYzwsPD3fZrNBqEhIS0yTXkf6PNM27cOHzwwQdIS0vDSy+9hK1bt2L8+PGw2+0AeL2U5HA4MHfuXFx77bW48sorAaDD/gZ64veg19y6g8hl/Pjx8vNBgwZh5MiRiIuLw2effQZfX18Fa0bU9UyePFl+PnDgQAwaNAhXXHEFtmzZgptuuknBmlFKSgoOHTqE7du3K10Vj8AWIgWEhYVBrVZfNGo/Ly8PJpNJoVp5r6CgIPTp0wfHjx+HyWRCTU0NiouL3crUvzYmk6nBa+fa11gZo9HI0NUKrt9vY//tmEwm5Ofnu+232WwoKipqk2vI/0Zbp2fPnggLC8Px48cB8HopZfbs2fj666/xww8/IDo6Wt7eUX8DPfF7kIFIATqdDsOGDUNaWpq8zeFwIC0tDYmJiQrWzDuVlZXhxIkTiIyMxLBhw6DVat2uTVZWFnJycuRrk5iYiIMHD7r9Ed+0aROMRiMSEhLkMvWP4SrD69s68fHxMJlMbr9bi8WC3bt3u12f4uJiZGRkyGU2b94Mh8OBkSNHymW2bdsGq9Uql9m0aRP69u2L4OBguQyvYds7ffo0CgsLERkZCYDXq6MJITB79mysXbsWmzdvRnx8vNv+jvob6JHfg4oM5SbxySefCL1eL1JTU8WRI0fEzJkzRVBQkNuofWofjz/+uNiyZYvIzs4WO3bsEElJSSIsLEzk5+cLIZxTTmNjY8XmzZvFvn37RGJiokhMTJTf75pyOnbsWJGZmSk2bNggunXr1uCU0wULFoijR4+KlStXctp9E5WWlor9+/eL/fv3CwBixYoVYv/+/eL3338XQjin3QcFBYn//Oc/4sCBA+LWW29tcNr9VVddJXbv3i22b98uevfu7TaNu7i4WERERIj77rtPHDp0SHzyySfCYDBcNI1bo9GIV155RRw9elQsXryY07gb0Nj1Ki0tFX/5y19Eenq6yM7OFt9//70YOnSo6N27t6iqqpKPwevVcWbNmiUCAwPFli1b3JZCqKiokMt01N9AT/seZCBS0BtvvCFiY2OFTqcTI0aMELt27VK6Sl7h7rvvFpGRkUKn04nu3buLu+++Wxw/flzeX1lZKf785z+L4OBgYTAYxO233y5yc3PdjnHy5Ekxfvx44evrK8LCwsTjjz8urFarW5kffvhBDBkyROh0OtGzZ0/x3nvvdcTpdXo//PCDAHDRY9q0aUII59T7Z599VkRERAi9Xi9uuukmkZWV5XaMwsJCcc899wh/f39hNBrFgw8+KEpLS93K/Pzzz+K6664Ter1edO/eXbz44osX1eWzzz4Tffr0ETqdTgwYMECsX7++3c67s2rselVUVIixY8eKbt26Ca1WK+Li4sSMGTMu+sLj9eo4DV0rAG5/nzryb6AnfQ9KQgjR0a1SRERERJ6EY4iIiIjI6zEQERERkddjICIiIiKvx0BEREREXo+BiIiIiLweAxERERF5PQYiIiIi8noMREREROT1GIiIqNVSU1MRFBQkv16yZAmGDBnS6HseeOAB3HbbbfLr0aNHY+7cue1Sv5Y4efIkJElCZmam0lVp1IW/eyJqGQYiIi/2wAMPQJIk+REaGopx48bhwIEDzTrO3XffjV9//bVVdfniiy/wwgsvtOoYTTF69Gj5fPV6Pbp3745bbrkFX3zxhVu5mJgY5Obm4sorr2z3OrVGW/zuiYiBiMjrjRs3Drm5ucjNzUVaWho0Gg1uvvnmZh3D19cX4eHhrapHSEgIAgICWnWMppoxYwZyc3Nx4sQJ/Pvf/0ZCQgImT56MmTNnymXUajVMJhM0Gk2H1Kml2uJ3T0QMREReT6/Xw2QywWQyYciQIXjqqadw6tQpnDt3DgCwZcsWSJKE4uJi+T2ZmZmQJAknT54EcPluG7vdjvnz5yMoKAihoaF44okncOFtFC/sMuvRowf+3//7f3jooYcQEBCA2NhY/POf/3R7z86dOzFkyBD4+Phg+PDh+PLLL5vUzWUwGGAymRAdHY1Ro0bhpZdewj/+8Q+8/fbb+P777wFc3GXm+j1s3LgRV111FXx9fXHjjTciPz8f3377Lfr37w+j0Yh7770XFRUV8mc5HA4sXboU8fHx8PX1xeDBg/F///d/8n7XcdPS0jB8+HAYDAZcc801yMrKksv8/PPPGDNmDAICAmA0GjFs2DDs27fvkr/7VatW4YorroBOp0Pfvn3xv//7v277JUnCO++8g9tvvx0GgwG9e/fGV1991ejvjKirYyAiIllZWRk+/PBD9OrVC6GhoW123OXLlyM1NRXvvvsutm/fjqKiIqxdu7ZJ7xs+fDj279+PP//5z5g1a5YcFCwWC2655RYMHDgQP/30E1544QU8+eSTLa7jtGnTEBwcfFHX2YWWLFmCN998Ezt37sSpU6dw11134bXXXsOaNWuwfv16fPfdd3jjjTfk8kuXLsUHH3yA1atX4/Dhw5g3bx6mTp2KrVu3uh336aefxvLly7Fv3z5oNBo89NBD8r4pU6YgOjoae/fuRUZGBp566ilotdoG67d27Vo89thjePzxx3Ho0CE88sgjePDBB/HDDz+4lXvuuedw11134cCBA5gwYQKmTJmCoqKi5v7aiLoOQURea9q0aUKtVgs/Pz/h5+cnAIjIyEiRkZEhl/nhhx8EAHH+/Hl52/79+wUAkZ2dLYQQ4r333hOBgYHy/sWLF4vBgwfLryMjI8WyZcvk11arVURHR4tbb71V3vZf//Vf4rHHHpNfx8XFialTp8qvHQ6HCA8PF6tWrRJCCLFq1SoRGhoqKisr5TJvv/22ACD2799/yXO+8HPqGzlypBg/frwQQojs7Gy3Y7l+D99//71cfunSpQKAOHHihLztkUceEcnJyUIIIaqqqoTBYBA7d+50+5zp06eLe+6555LHXb9+vQAgn1tAQIBITU1tsM4X/u6vueYaMWPGDLcyd955p5gwYYL8GoB45pln5NdlZWUCgPj2228b/Awib8AWIiIvN2bMGGRmZiIzMxN79uxBcnIyxo8fj99//71Njl9SUoLc3FyMHDlS3qbRaDB8+PDLvnfQoEHyc0mSYDKZkJ+fDwDIysrCoEGD4OPjI5cZMWJEq+oqhIAkSU2uU0REBAwGA3r27Om2zVXH48ePo6KiAn/4wx/g7+8vPz744AOcOHHikseNjIwEAPk48+fPx8MPP4ykpCS8+OKLF723vqNHj+Laa69123bttdfi6NGjl/w8Pz8/GI1G+fOIvBEDEZGX8/PzQ69evdCrVy9cffXVeOedd1BeXo63334bAKBSOf9MiHpjfqxWa4fU7cJuIUmS4HA42uWz7HY7jh07hvj4+CbXSZKkRutYVlYGAFi/fr0cOjMzM3HkyBG3cUQNHReAfJwlS5bg8OHDmDhxIjZv3oyEhIQmdTk29TwurDeRN2IgIiI3kiRBpVKhsrISANCtWzcAQG5urlymOWvzBAYGIjIyErt375a32Ww2ZGRktKqeffv2xcGDB1FdXS1v27t3b4uP9/777+P8+fOYNGlSq+pVX0JCAvR6PXJycuTQ6XrExMQ061h9+vTBvHnz8N133+GOO+7Ae++912C5/v37Y8eOHW7bduzYgYSEhBafB5E38Oz5pETU7qqrq2E2mwEA58+fx5tvvomysjLccsstACB/eS9ZsgR/+9vf8Ouvv2L58uXN+ozHHnsML774Inr37o1+/fphxYoVbrPWWuLee+/F008/jZkzZ+Kpp55CTk4OXnnlFQC4bLdXRUUFzGYzbDYbTp8+jbVr1+LVV1/FrFmzMGbMmFbVq76AgAD85S9/wbx58+BwOHDdddehpKQEO3bsgNFoxLRp0y57jMrKSixYsAB/+tOfEB8fj9OnT2Pv3r2XDG4LFizAXXfdhauuugpJSUlYt24dvvjiC3n2HBE1jIGIyMtt2LBBHrMSEBCAfv364fPPP8fo0aMBOLtWPv74Y8yaNQuDBg3C1Vdfjb/+9a+48847m/wZjz/+OHJzczFt2jSoVCo89NBDuP3221FSUtLiehuNRqxbtw6zZs3CkCFDMHDgQCxatAj33nuv27iihrz99tt4++23odPpEBoaimHDhuHTTz/F7bff3uL6XMoLL7yAbt26YenSpfjtt98QFBSEoUOH4n/+53+a9H61Wo3CwkLcf//9yMvLQ1hYGO644w4899xzDZa/7bbb8Prrr+OVV17BY489hvj4eLz33nvy9SSihklCXLAYCBFRJ/XRRx/hwQcfRElJCXx9fZWuDhF1ImwhIqJO64MPPkDPnj3RvXt3/Pzzz3jyySdx1113MQwRUbMxEBFRp2U2m7Fo0SKYzWZERkbizjvvxN/+9jelq0VEnRC7zIiIiMjrcdo9EREReT0GIiIiIvJ6DERERETk9RiIiIiIyOsxEBEREZHXYyAiIiIir8dARERERF6PgYiIiIi83v8HK/91pxDR9qkAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.histplot(x='Building Dimension', data=df_insurance, bins=15, kde=True)\n",
    "# we'll change it with the median\n",
    "handle_nan_values(df_insurance, 'Building Dimension', 'median')\n",
    "df_insurance.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Building Dimension - Transformation/Normalization first\n",
    "This feature contains a lots of outliers, so we'll use robust scaller"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize_with_robust_scaler(df:pd.DataFrame, column:str):\n",
    "  standard_sclaer = RobustScaler()\n",
    "  bd_norm_arr = standard_sclaer.fit_transform(df.loc[:, [column]])\n",
    "  df[column] = bd_norm_arr\n",
    "\n",
    "normalize_with_robust_scaler(df_insurance, 'Building Dimension')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Building Dimension - Outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.histplot(x='Building Dimension', data=df_insurance, bins=15, kde=True)\n",
    "plt.show()\n",
    "sns.boxplot(data=df_insurance, x='Building Dimension')\n",
    "def handle_outliers_with_box_plot(df:pd.DataFrame, column:str):\n",
    "  Q1, Q3 = np.percentile(df[column], [25, 75])\n",
    "  IQR = Q3 - Q1\n",
    "  upper_limit = Q3+1.5*IQR\n",
    "  lower_limit = Q1-1.5*IQR\n",
    "  df[column] = np.where(df[column] > upper_limit, upper_limit,\n",
    "                                                np.where(df[column] < lower_limit, lower_limit, df[column]))\n",
    "handle_outliers_with_box_plot(df_insurance, 'Building Dimension')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Building Dimension'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=df_insurance, x='Building Dimension')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Geo_Code - Nan values - Most Frequent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "YearOfObservation     0\n",
       "Insured_Period        0\n",
       "Residential           0\n",
       "Building_Painted      0\n",
       "Building_Fenced       0\n",
       "Garden                0\n",
       "Settlement            0\n",
       "Building Dimension    0\n",
       "Building_Type         0\n",
       "NumberOfWindows       0\n",
       "Geo_Code              0\n",
       "Claim                 0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#df_insurance['Geo_Code']\n",
    "handle_nan_values(df_insurance, 'Geo_Code', 'most_frequent')\n",
    "df_insurance.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Geo_Code Remove column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building_Painted</th>\n",
       "      <th>Building_Fenced</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>without</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>without</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>without</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>without</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Building_Painted  \\\n",
       "0                  2012             1.0            1                N   \n",
       "1                  2012             1.0            0                N   \n",
       "2                  2013             1.0            1                V   \n",
       "3                  2016             0.5            0                N   \n",
       "4                  2012             1.0            0                N   \n",
       "...                 ...             ...          ...              ...   \n",
       "5007               2013             1.0            0                N   \n",
       "5008               2012             0.5            0                V   \n",
       "5009               2015             1.0            1                V   \n",
       "5010               2012             0.5            0                V   \n",
       "5011               2013             1.0            1                V   \n",
       "\n",
       "     Building_Fenced Garden Settlement  Building Dimension    Building_Type  \\\n",
       "0                  V      V          U            0.100581      Wood-framed   \n",
       "1                  V      V          U           -0.097093  Non-combustible   \n",
       "2                  N      O          R            2.187791  Non-combustible   \n",
       "3                  V      V          U           -0.271512      Wood-framed   \n",
       "4                  V      V          U           -0.097093  Non-combustible   \n",
       "...              ...    ...        ...                 ...              ...   \n",
       "5007               V      V          U           -0.300581         Ordinary   \n",
       "5008               N      O          R           -0.038953   Fire-resistive   \n",
       "5009               N      O          R           -0.341279         Ordinary   \n",
       "5010               N      O          R           -0.308721   Fire-resistive   \n",
       "5011               V      V          U            0.000000      Wood-framed   \n",
       "\n",
       "     NumberOfWindows Claim  \n",
       "0            without   non  \n",
       "1            without   non  \n",
       "2                  4   oui  \n",
       "3            without   oui  \n",
       "4            without   non  \n",
       "...              ...   ...  \n",
       "5007         without   oui  \n",
       "5008               4   non  \n",
       "5009               3   non  \n",
       "5010               4   non  \n",
       "5011         without   non  \n",
       "\n",
       "[5012 rows x 11 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def remove_column(df:pd.DataFrame, column:str):\n",
    "  df.drop(axis=1, labels=[column], inplace=True)\n",
    "\n",
    "remove_column(df_insurance, 'Geo_Code')\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkQAAAGwCAYAAABIC3rIAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA4z0lEQVR4nO3de1wWdd7/8fcFcvIAisgpEc+A5lkz0hSVW1Jz887d1bLStNx6oKvpmnln6rqlZptaymrtplarW7Z3WlkeEAVPaMaKp9DUW8MD4Bk8IsL8/ujB/LzCI15wXTiv5+Mxj3Xm+71mPnO1ypvvfGfGZhiGIQAAAAtzc3YBAAAAzkYgAgAAlkcgAgAAlkcgAgAAlkcgAgAAlkcgAgAAlkcgAgAAllfJ2QVUBEVFRTp+/LiqVasmm83m7HIAAMAdMAxD58+fV2hoqNzcbj0GRCC6A8ePH1dYWJizywAAAKVw5MgR1a5d+5Z9CER3oFq1apJ++UJ9fX2dXA0AALgTeXl5CgsLM3+O3wqB6A4UXybz9fUlEAEAUMHcyXQXJlUDAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLc2ogmjp1qtq1a6dq1aopMDBQffr00b59++z6xMTEyGaz2S0vvfSSXZ/MzEz16tVLlStXVmBgoMaMGaNr167Z9UlOTlbr1q3l5eWlhg0bauHChWV9egAAoIJwaiBKSUlRfHy8tmzZosTERBUUFKh79+66ePGiXb8XX3xRWVlZ5jJ9+nSzrbCwUL169dLVq1e1efNmffzxx1q4cKEmTJhg9jl06JB69eqlLl26KD09XSNHjtQLL7ygVatWldu5AgAA12UzDMNwdhHFTp48qcDAQKWkpKhTp06SfhkhatmypWbNmnXDz6xYsUKPP/64jh8/rqCgIEnSvHnzNHbsWJ08eVKenp4aO3asvv32W+3evdv8XP/+/XXu3DmtXLmyxD7z8/OVn59vrhe/CyU3N5dXdwAAUEHk5eXJz8/vjn5+u9QcotzcXEmSv7+/3fZFixYpICBADz74oMaNG6dLly6ZbampqWrWrJkZhiQpLi5OeXl52rNnj9knNjbWbp9xcXFKTU29YR1Tp06Vn5+fufCmewAA7m8u83LXoqIijRw5Uh06dNCDDz5obn/66acVHh6u0NBQ7dy5U2PHjtW+ffv05ZdfSpKys7PtwpAkcz07O/uWffLy8nT58mX5+PjYtY0bN06jRo0y14tHiAAAwP3JZQJRfHy8du/erY0bN9ptHzp0qPnnZs2aKSQkRN26ddPBgwfVoEGDMqnFy8tLXl5eZbJvAADgelwiEA0bNkzLly/X+vXrVbt27Vv2bd++vSTpwIEDatCggYKDg/X999/b9cnJyZEkBQcHm/9bvO36Pr6+viVGh5whMzNTp06dcnYZtxQQEKA6deo4uwwAAMqEUwORYRgaPny4li5dquTkZNWrV++2n0lPT5ckhYSESJKio6P11ltv6cSJEwoMDJQkJSYmytfXV02aNDH7fPfdd3b7SUxMVHR0tAPPpnQyMzMVGRmly5cv3b6zE/n4VNbevRmEIgDAfcmpgSg+Pl6LFy/WV199pWrVqplzfvz8/OTj46ODBw9q8eLF6tmzp2rWrKmdO3fqlVdeUadOndS8eXNJUvfu3dWkSRM9++yzmj59urKzszV+/HjFx8ebl71eeuklzZkzR6+++qoGDx6stWvXasmSJfr222+ddu7FTp06pcuXL6n94InyDanr7HJuKC/rsLbO/7NOnTpFIAIA3JecGojmzp0r6Zdb66+3YMECDRo0SJ6enlqzZo1mzZqlixcvKiwsTH379tX48ePNvu7u7lq+fLlefvllRUdHq0qVKho4cKAmT55s9qlXr56+/fZbvfLKK3rvvfdUu3Zt/eMf/1BcXFy5nOed8A2pK/86Ec4uAwAAS3L6JbNbCQsLU0pKym33Ex4eXuKS2K/FxMRo+/btd1UfAACwBpd6DhEAAIAzEIgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlEYgAAIDlOTUQTZ06Ve3atVO1atUUGBioPn36aN++fXZ9rly5ovj4eNWsWVNVq1ZV3759lZOTY9cnMzNTvXr1UuXKlRUYGKgxY8bo2rVrdn2Sk5PVunVreXl5qWHDhlq4cGFZnx4AAKggnBqIUlJSFB8fry1btigxMVEFBQXq3r27Ll68aPZ55ZVX9M033+iLL75QSkqKjh8/rieffNJsLywsVK9evXT16lVt3rxZH3/8sRYuXKgJEyaYfQ4dOqRevXqpS5cuSk9P18iRI/XCCy9o1apV5Xq+AADANVVy5sFXrlxpt75w4UIFBgYqLS1NnTp1Um5urj766CMtXrxYXbt2lSQtWLBAUVFR2rJlix5++GGtXr1aP/74o9asWaOgoCC1bNlSf/nLXzR27FhNmjRJnp6emjdvnurVq6d3331XkhQVFaWNGzdq5syZiouLK1FXfn6+8vPzzfW8vLwy/BYAAICzudQcotzcXEmSv7+/JCktLU0FBQWKjY01+0RGRqpOnTpKTU2VJKWmpqpZs2YKCgoy+8TFxSkvL0979uwx+1y/j+I+xfv4talTp8rPz89cwsLCHHeSAADA5bhMICoqKtLIkSPVoUMHPfjgg5Kk7OxseXp6qnr16nZ9g4KClJ2dbfa5PgwVtxe33apPXl6eLl++XKKWcePGKTc311yOHDnikHMEAACuyamXzK4XHx+v3bt3a+PGjc4uRV5eXvLy8nJ2GQAAoJy4xAjRsGHDtHz5cq1bt061a9c2twcHB+vq1as6d+6cXf+cnBwFBwebfX5911nx+u36+Pr6ysfHx9GnAwAAKhinBiLDMDRs2DAtXbpUa9euVb169eza27RpIw8PDyUlJZnb9u3bp8zMTEVHR0uSoqOjtWvXLp04ccLsk5iYKF9fXzVp0sTsc/0+ivsU7wMAAFibUy+ZxcfHa/Hixfrqq69UrVo1c86Pn5+ffHx85OfnpyFDhmjUqFHy9/eXr6+vhg8frujoaD388MOSpO7du6tJkyZ69tlnNX36dGVnZ2v8+PGKj483L3u99NJLmjNnjl599VUNHjxYa9eu1ZIlS/Ttt9867dwBAIDrcOoI0dy5c5Wbm6uYmBiFhISYy+eff272mTlzph5//HH17dtXnTp1UnBwsL788kuz3d3dXcuXL5e7u7uio6P1zDPP6LnnntPkyZPNPvXq1dO3336rxMREtWjRQu+++67+8Y9/3PCWewAAYD1OHSEyDOO2fby9vZWQkKCEhISb9gkPD9d33313y/3ExMRo+/btd10jAAC4/7nEpGoAAABnIhABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLIxABAADLc2ogWr9+vXr37q3Q0FDZbDYtW7bMrn3QoEGy2Wx2y2OPPWbX58yZMxowYIB8fX1VvXp1DRkyRBcuXLDrs3PnTj366KPy9vZWWFiYpk+fXtanBgAAKhCnBqKLFy+qRYsWSkhIuGmfxx57TFlZWebyr3/9y659wIAB2rNnjxITE7V8+XKtX79eQ4cONdvz8vLUvXt3hYeHKy0tTe+8844mTZqkDz/8sMzOCwAAVCyVnHnwHj16qEePHrfs4+XlpeDg4Bu2ZWRkaOXKldq2bZvatm0rSZo9e7Z69uypv/71rwoNDdWiRYt09epVzZ8/X56enmratKnS09M1Y8YMu+B0vfz8fOXn55vreXl5pTxDAABQEbj8HKLk5GQFBgYqIiJCL7/8sk6fPm22paamqnr16mYYkqTY2Fi5ublp69atZp9OnTrJ09PT7BMXF6d9+/bp7NmzNzzm1KlT5efnZy5hYWFldHYAAMAVuHQgeuyxx/TJJ58oKSlJb7/9tlJSUtSjRw8VFhZKkrKzsxUYGGj3mUqVKsnf31/Z2dlmn6CgILs+xevFfX5t3Lhxys3NNZcjR444+tQAAIALceols9vp37+/+edmzZqpefPmatCggZKTk9WtW7cyO66Xl5e8vLzKbP8AAMC1uPQI0a/Vr19fAQEBOnDggCQpODhYJ06csOtz7do1nTlzxpx3FBwcrJycHLs+xes3m5sEAACspUIFoqNHj+r06dMKCQmRJEVHR+vcuXNKS0sz+6xdu1ZFRUVq37692Wf9+vUqKCgw+yQmJioiIkI1atQo3xMAAAAuyamB6MKFC0pPT1d6erok6dChQ0pPT1dmZqYuXLigMWPGaMuWLTp8+LCSkpL0xBNPqGHDhoqLi5MkRUVF6bHHHtOLL76o77//Xps2bdKwYcPUv39/hYaGSpKefvppeXp6asiQIdqzZ48+//xzvffeexo1apSzThsAALgYpwaiH374Qa1atVKrVq0kSaNGjVKrVq00YcIEubu7a+fOnfrNb36jxo0ba8iQIWrTpo02bNhgN79n0aJFioyMVLdu3dSzZ0917NjR7hlDfn5+Wr16tQ4dOqQ2bdpo9OjRmjBhwk1vuQcAANbj1EnVMTExMgzjpu2rVq267T78/f21ePHiW/Zp3ry5NmzYcNf1AQAAa6hQc4gAAADKAoEIAABYHoEIAABYXqkCUf369e1eoVHs3Llzql+//j0XBQAAUJ5KFYgOHz5svj7jevn5+Tp27Ng9FwUAAFCe7uous6+//tr886pVq+Tn52euFxYWKikpSXXr1nVYcQAAAOXhrgJRnz59JEk2m00DBw60a/Pw8FDdunX17rvvOqw4AACA8nBXgaioqEiSVK9ePW3btk0BAQFlUhQAAEB5KtWDGQ8dOuToOgAAAJym1E+qTkpKUlJSkk6cOGGOHBWbP3/+PRcGAABQXkoViP785z9r8uTJatu2rUJCQmSz2RxdFwAAQLkpVSCaN2+eFi5cqGeffdbR9QAAAJS7Uj2H6OrVq3rkkUccXQsAAIBTlCoQvfDCC7d9wzwAAEBFUapLZleuXNGHH36oNWvWqHnz5vLw8LBrnzFjhkOKAwAAKA+lCkQ7d+5Uy5YtJUm7d++2a2OCNQAAqGhKFYjWrVvn6DoAAACcplRziAAAAO4npRoh6tKlyy0vja1du7bUBQEAAJS3UgWi4vlDxQoKCpSenq7du3eXeOkrAACAqytVIJo5c+YNt0+aNEkXLly4p4IAAADKm0PnED3zzDO8xwwAAFQ4Dg1Eqamp8vb2duQuAQAAylypLpk9+eSTduuGYSgrK0s//PCD3njjDYcUBgAAUF5KFYj8/Pzs1t3c3BQREaHJkyere/fuDikMAACgvJQqEC1YsMDRdQAAADhNqQJRsbS0NGVkZEiSmjZtqlatWjmkKAAAgPJUqkB04sQJ9e/fX8nJyapevbok6dy5c+rSpYs+++wz1apVy5E1AgAAlKlS3WU2fPhwnT9/Xnv27NGZM2d05swZ7d69W3l5efrjH//o6BoBAADKVKlGiFauXKk1a9YoKirK3NakSRMlJCQwqRoAAFQ4pRohKioqkoeHR4ntHh4eKioquueiAAAAylOpAlHXrl01YsQIHT9+3Nx27NgxvfLKK+rWrZvDigMAACgPpQpEc+bMUV5enurWrasGDRqoQYMGqlevnvLy8jR79mxH1wgAAFCmSjWHKCwsTP/5z3+0Zs0a7d27V5IUFRWl2NhYhxYHAABQHu5qhGjt2rVq0qSJ8vLyZLPZ9F//9V8aPny4hg8frnbt2qlp06basGFDWdUKAABQJu4qEM2aNUsvvviifH19S7T5+fnpD3/4g2bMmOGw4gAAAMrDXQWiHTt26LHHHrtpe/fu3ZWWlnbPRQEAAJSnuwpEOTk5N7zdvlilSpV08uTJey4KAACgPN1VIHrggQe0e/fum7bv3LlTISEh91wUAABAebqrQNSzZ0+98cYbunLlSom2y5cva+LEiXr88ccdVhwAAEB5uKvb7sePH68vv/xSjRs31rBhwxQRESFJ2rt3rxISElRYWKjXX3+9TAoFAAAoK3cViIKCgrR582a9/PLLGjdunAzDkCTZbDbFxcUpISFBQUFBZVIoAABAWbnrBzOGh4fru+++09mzZ3XgwAEZhqFGjRqpRo0aZVEfAABAmSvVk6olqUaNGmrXrp0jawEAAHCKUr3LDAAA4H5CIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJZHIAIAAJbn1EC0fv169e7dW6GhobLZbFq2bJldu2EYmjBhgkJCQuTj46PY2Fjt37/frs+ZM2c0YMAA+fr6qnr16hoyZIguXLhg12fnzp169NFH5e3trbCwME2fPr2sTw0AAFQgTg1EFy9eVIsWLZSQkHDD9unTp+v999/XvHnztHXrVlWpUkVxcXG6cuWK2WfAgAHas2ePEhMTtXz5cq1fv15Dhw412/Py8tS9e3eFh4crLS1N77zzjiZNmqQPP/ywzM8PAABUDJWcefAePXqoR48eN2wzDEOzZs3S+PHj9cQTT0iSPvnkEwUFBWnZsmXq37+/MjIytHLlSm3btk1t27aVJM2ePVs9e/bUX//6V4WGhmrRokW6evWq5s+fL09PTzVt2lTp6emaMWOGXXACAADW5bJziA4dOqTs7GzFxsaa2/z8/NS+fXulpqZKklJTU1W9enUzDElSbGys3NzctHXrVrNPp06d5OnpafaJi4vTvn37dPbs2RseOz8/X3l5eXYLAAC4f7lsIMrOzpYkBQUF2W0PCgoy27KzsxUYGGjXXqlSJfn7+9v1udE+rj/Gr02dOlV+fn7mEhYWdu8nBAAAXJbLBiJnGjdunHJzc83lyJEjzi4JAACUIZcNRMHBwZKknJwcu+05OTlmW3BwsE6cOGHXfu3aNZ05c8auz432cf0xfs3Ly0u+vr52CwAAuH+5bCCqV6+egoODlZSUZG7Ly8vT1q1bFR0dLUmKjo7WuXPnlJaWZvZZu3atioqK1L59e7PP+vXrVVBQYPZJTExURESEatSoUU5nAwAAXJlTA9GFCxeUnp6u9PR0Sb9MpE5PT1dmZqZsNptGjhypN998U19//bV27dql5557TqGhoerTp48kKSoqSo899phefPFFff/999q0aZOGDRum/v37KzQ0VJL09NNPy9PTU0OGDNGePXv0+eef67333tOoUaOcdNYAAMDVOPW2+x9++EFdunQx14tDysCBA7Vw4UK9+uqrunjxooYOHapz586pY8eOWrlypby9vc3PLFq0SMOGDVO3bt3k5uamvn376v333zfb/fz8tHr1asXHx6tNmzYKCAjQhAkTuOUeAACYnBqIYmJiZBjGTdttNpsmT56syZMn37SPv7+/Fi9efMvjNG/eXBs2bCh1nQAA4P7msnOIAAAAyguBCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWB6BCAAAWF4lZxeAiiMjI8PZJdxSQECA6tSp4+wyAAAVEIEIt3U597Qkm5555hlnl3JLPj6VtXdvBqEIAHDXCES4rYJL5yUZavn0WNWqF+nscm4oL+uwts7/s06dOkUgAgDcNQIR7ljVwDryrxPh7DIAAHA4JlUDAADLIxABAADLIxABAADLIxABAADLIxABAADL4y4zABVOZmamTp065ewybokHhQIVC4EIQIWSmZmpyMgoXb58ydml3BIPCgUqFgIRgArl1KlTunz5ktoPnijfkLrOLueGeFAoUPEQiABUSL4hdXlQKACHYVI1AACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPJcORJMmTZLNZrNbIiMjzfYrV64oPj5eNWvWVNWqVdW3b1/l5OTY7SMzM1O9evVS5cqVFRgYqDFjxujatWvlfSoAAMCFufyTqps2bao1a9aY65Uq/f+SX3nlFX377bf64osv5Ofnp2HDhunJJ5/Upk2bJEmFhYXq1auXgoODtXnzZmVlZem5556Th4eHpkyZUu7nAgAAXJPLB6JKlSopODi4xPbc3Fx99NFHWrx4sbp27SpJWrBggaKiorRlyxY9/PDDWr16tX788UetWbNGQUFBatmypf7yl79o7NixmjRpkjw9Pcv7dAAAgAty6UtmkrR//36Fhoaqfv36GjBggDIzMyVJaWlpKigoUGxsrNk3MjJSderUUWpqqiQpNTVVzZo1U1BQkNknLi5OeXl52rNnz02PmZ+fr7y8PLsFAADcv1w6ELVv314LFy7UypUrNXfuXB06dEiPPvqozp8/r+zsbHl6eqp69ep2nwkKClJ2drYkKTs72y4MFbcXt93M1KlT5efnZy5hYWGOPTEAAOBSXPqSWY8ePcw/N2/eXO3bt1d4eLiWLFkiHx+fMjvuuHHjNGrUKHM9Ly+PUAQAwH3MpUeIfq169epq3LixDhw4oODgYF29elXnzp2z65OTk2POOQoODi5x11nx+o3mJRXz8vKSr6+v3QIAAO5fFSoQXbhwQQcPHlRISIjatGkjDw8PJSUlme379u1TZmamoqOjJUnR0dHatWuXTpw4YfZJTEyUr6+vmjRpUu71AwAA1+TSl8z+9Kc/qXfv3goPD9fx48c1ceJEubu766mnnpKfn5+GDBmiUaNGyd/fX76+vho+fLiio6P18MMPS5K6d++uJk2a6Nlnn9X06dOVnZ2t8ePHKz4+Xl5eXk4+OwAA4CpcOhAdPXpUTz31lE6fPq1atWqpY8eO2rJli2rVqiVJmjlzptzc3NS3b1/l5+crLi5Of/vb38zPu7u7a/ny5Xr55ZcVHR2tKlWqaODAgZo8ebKzTgkAALgglw5En3322S3bvb29lZCQoISEhJv2CQ8P13fffefo0gAAwH2kQs0hAgAAKAsuPUIEoPxlZmbq1KlTzi7jpjIyMpxdAoD7EIEIgCkzM1ORkVG6fPmSs0u5rYL8q84uAcB9hEAEwHTq1CldvnxJ7QdPlG9IXWeXc0NZu1K1++sPde3aNWeXAuA+QiACUIJvSF3514lwdhk3lJd12NklALgPMakaAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYHoEIAABYXiVnFwA4UkZGhrNLuKWAgADVqVPH2WUAAH6FQIT7wuXc05JseuaZZ5xdyi35+FTW3r0ZhCIAcDEEItwXCi6dl2So5dNjVatepLPLuaG8rMPaOv/POnXqFIEIAFwMgQj3laqBdeRfJ8LZZQBwkMzMTJ06dcrZZdwSl8LvDwQiAIBLyszMVGRklC5fvuTsUm6JS+H3BwIRAMAlnTp1SpcvX1L7wRPlG1LX2eXcEJfC7x8EIgCAS/MNqculcJQ5nkMEAAAsj0AEAAAsj0AEAAAsj0AEAAAsj0AEAAAsj0AEAAAsj0AEAAAsj0AEAAAsjwczAuUsIyPD2SXclCvXBgBliUAElJPLuacl2fTMM884u5TbKsi/6uwSAKBcEYiAclJw6bwkQy2fHqta9SKdXc4NZe1K1e6vP9S1a9ecXQoAlCsCEVDOqgbWcdn3MuVlHXZ2CQDgFEyqBgAAlkcgAgAAlkcgAgAAlkcgAgAAlsekagAoI67+XKeAgADVqVPH2WUALoFABAAOVlGeOeXjU1l792YQigARiADA4SrCM6fysg5r6/w/69SpUwQiQAQiACgzrvzMKQD2CEQAYGGuPM/JlWvD/YdABAAWVFHmOUkV4916rh7emEB/ewQiALCgijDPqSK8W6+iBEsm0N8egQgALMyV5zlVhHfrVYRgWTyBfsOGDYqKinJ2OTfl7FEsAhEAAPfIlYMlo1h3xlKBKCEhQe+8846ys7PVokULzZ49Ww899JCzywIAoMxUpFEsZz4GwjKB6PPPP9eoUaM0b948tW/fXrNmzVJcXJz27dunwMBAZ5cHAECZcuVRLFdgmXeZzZgxQy+++KKef/55NWnSRPPmzVPlypU1f/58Z5cGAACczBIjRFevXlVaWprGjRtnbnNzc1NsbKxSU1NL9M/Pz1d+fr65npubK0nKy8tzeG0XLlyQJJ35eZ+u5V92+P4dIS/rZ0lS7rH98qhkc3I1N0aNjkGNjkGNjkGNjlEhaszOlPTLz0RH/qwt3pdhGLfvbFjAsWPHDEnG5s2b7baPGTPGeOihh0r0nzhxoiGJhYWFhYWF5T5Yjhw5ctusYIkRors1btw4jRo1ylwvKirSmTNnVLNmTdlsrpmui+Xl5SksLExHjhyRr6+vs8u5IWp0DGp0DGp0DGp0DGp0LMMwdP78eYWGht62ryUCUUBAgNzd3ZWTk2O3PScnR8HBwSX6e3l5ycvLy25b9erVy7JEh/P19XX5/6NSo2NQo2NQo2NQo2NQo+P4+fndUT9LTKr29PRUmzZtlJSUZG4rKipSUlKSoqOjnVgZAABwBZYYIZKkUaNGaeDAgWrbtq0eeughzZo1SxcvXtTzzz/v7NIAAICTWSYQ9evXTydPntSECROUnZ2tli1bauXKlQoKCnJ2aQ7l5eWliRMnlrjk50qo0TGo0TGo0TGo0TGo0XlshnEn96IBAADcvywxhwgAAOBWCEQAAMDyCEQAAMDyCEQubOHChXf0/CObzaZly5aVeT0AANyvCEQurF+/fvrpp5/M9UmTJqlly5ZOq8fRwWvatGmy2WwaOXKkw/Z5r6ZOnap27dqpWrVqCgwMVJ8+fbRv3z5nl2Wnbt26stlsJZb4+Hhnl2aaNGlSifoiIyOdXZaduXPnqnnz5ubD5aKjo7VixQpnl2WnsLBQb7zxhurVqycfHx81aNBAf/nLX+7svUxOlpWVpaefflqNGzeWm5vbTf+ef/HFF4qMjJS3t7eaNWum7777rkzqOXbsmJ555hnVrFlTPj4+atasmX744YcyOVZprF+/Xr1791ZoaKhL/5J7/vx5jRw5UuHh4fLx8dEjjzyibdu2ObsshyAQuTAfHx8FBgY6u4wysW3bNn3wwQdq3ry5s0uxk5KSovj4eG3ZskWJiYkqKChQ9+7ddfHiRWeXZtq2bZuysrLMJTExUZL0u9/9zsmV2WvatKldnRs3bnR2SXZq166tadOmKS0tTT/88IO6du2qJ554Qnv27HF2aaa3335bc+fO1Zw5c5SRkaG3335b06dP1+zZsx12jOPHj+vatWsO21+x/Px81apVS+PHj1eLFi1u2Gfz5s166qmnNGTIEG3fvl19+vRRnz59tHv3bofWcvbsWXXo0EEeHh5asWKFfvzxR7377ruqUaOGQ49zLy5evKgWLVooISHB2aXc0gsvvKDExER9+umn2rVrl7p3767Y2FgdO3bM2aXdO4e8PRV37JtvvjH8/PyMa9euGYZhGNu3bzckGWPHjjX7DBkyxBgwYICxYMECw8/PzzAMw1iwYEGJl9UtWLDAMAzDkGT8/e9/N/r06WP4+PgYDRs2NL766iu74yYnJxvt2rUzPD09jeDgYGPs2LFGQUGB2R4eHm7MnDnT7jMtWrQwJk6caLZff+zw8PBSfwfnz583GjVqZCQmJhqdO3c2RowYUep9lbUTJ04YkoyUlBRnl3JTI0aMMBo0aGAUFRU5uxTTxIkTjRYtWji7jLtWo0YN4x//+IezyzD16tXLGDx4sN22J5980hgwYIDDjjFp0iQjKCjIGD16tLFz506H7fd6N/t7/vvf/97o1auX3bb27dsbf/jDHxx6/LFjxxodO3Z06D7LkiRj6dKlzi6jhEuXLhnu7u7G8uXL7ba3bt3aeP31151UleMwQlTOHn30UZ0/f17bt2+X9MuIREBAgJKTk80+KSkpiomJsftcv379NHr0aLvfuvv162e2//nPf9bvf/977dy5Uz179tSAAQN05swZSb8MFffs2VPt2rXTjh07NHfuXH300Ud6880377ju4iHRBQsWKCsr656GSOPj49WrVy/FxsaWeh/lJTc3V5Lk7+/v5Epu7OrVq/rnP/+pwYMHu9yLh/fv36/Q0FDVr19fAwYMUGZmprNLuqnCwkJ99tlnunjxoku9zueRRx5RUlKSeel8x44d2rhxo3r06OGwY4wdO1bvvfeeMjIy1Lp1a7Vu3Vrvv/++Tp48WaJv06ZNVbVq1Zsud1tXampqiX8H4uLilJqaek/n9Gtff/212rZtq9/97ncKDAxUq1at9Pe//92hx7CCa9euqbCwUN7e3nbbfXx8XG4EuFScncisqHXr1sY777xjGIZh9OnTx3jrrbcMT09P4/z588bRo0cNScZPP/1kN0JkGDf/rVuSMX78eHP9woULhiRjxYoVhmEYxv/8z/8YERERdiMICQkJRtWqVY3CwkLDMG4/QlR8nHv9reVf//qX8eCDDxqXL182DOPmvzm6gsLCQqNXr15Ghw4dnF3KTX3++eeGu7u7cezYMWeXYue7774zlixZYuzYscNYuXKlER0dbdSpU8fIy8tzdml2du7caVSpUsVwd3c3/Pz8jG+//dbZJdkpLCw0xo4da9hsNqNSpUqGzWYzpkyZUmbHy8nJMWbOnGm0atXK8PDwMJ544gnjyy+/NEeTDx8+bOzfv/+my9GjR2+435v9Pffw8DAWL15sty0hIcEIDAx06Hl5eXkZXl5exrhx44z//Oc/xgcffGB4e3sbCxcudOhxHMUR/9aWlejoaKNz587GsWPHjGvXrhmffvqp4ebmZjRu3NjZpd0zy7y6w5V07txZycnJGj16tDZs2KCpU6dqyZIl2rhxo86cOaPQ0FA1atRImzZtuuN9Xj8Xp0qVKvL19dWJEyckSRkZGYqOjrYbQejQoYMuXLigo0ePqk6dOo47uVs4cuSIRowYocTExBK/Ybii+Ph47d6926V/8/noo4/Uo0cPhYaGOrsUO9ePFDRv3lzt27dXeHi4lixZoiFDhjixMnsRERFKT09Xbm6u/v3vf2vgwIFKSUlRkyZNnF2aJGnJkiVatGiRFi9erKZNmyo9PV0jR45UaGioBg4c6PDjBQYGauTIkRo5cqRWrFihQYMG6auvvtL27dvVsmVLhYeHO/yY5aGoqEht27bVlClTJEmtWrXS7t27NW/evDL5Hu9nn376qQYPHqwHHnhA7u7uat26tZ566imlpaU5u7R7RiBygpiYGM2fP187duyQh4eHIiMjFRMTo+TkZJ09e1adO3e+6316eHjYrdtsNhUVFd3x593c3ErcuVJQUHDXddxKWlqaTpw4odatW5vbCgsLtX79es2ZM0f5+flyd3d36DFLa9iwYVq+fLnWr1+v2rVrO7ucG/r555+1Zs0affnll84u5baqV6+uxo0b68CBA84uxY6np6caNmwoSWrTpo22bdum9957Tx988IGTK/vFmDFj9Nprr6l///6SpGbNmunnn3/W1KlTy+QH+fnz5/Xvf/9bn376qdavX6/OnTtr4MCBZkBs2rSpfv7555t+/tFHH72rO/WCg4OVk5Njty0nJ0fBwcGlO4GbCAkJKRFyo6Ki9L//+78OPY4VNGjQQCkpKbp48aLy8vIUEhKifv36qX79+s4u7Z4RiJygeB7RzJkzzfATExOjadOm6ezZsxo9evQNP+fp6anCwsK7Pl7xX3zDMMxRok2bNqlatWrmD/tatWopKyvL/ExeXp4OHTpktx8PD49SHb9Yt27dtGvXLrttzz//vCIjIzV27FiXCEOGYWj48OFaunSpkpOTVa9ePWeXdFMLFixQYGCgevXq5exSbuvChQs6ePCgnn32WWeXcktFRUXKz893dhmmS5cuyc3Nfqqnu7v7Xf2yczuFhYVavXq1Pv30Uy1btkxhYWF67rnntHDhwhKjx999990tf1Hy8fG5q2NHR0crKSnJ7pb8xMREh8/j6tChQ4nHZ/z0008VdsTLFVSpUkVVqlTR2bNntWrVKk2fPt3ZJd07J1+ys6yWLVsa7u7uxty5cw3DMIzTp08bHh4ehiRj7969hmEYJeYQLVq0yKhSpYqxfft24+TJk8aVK1cMw7jx9WY/Pz/zLrSjR48alStXNuLj442MjAxj2bJlRkBAgN38oNdee80IDg421q9fb+zcudPo06ePUbVqVbs+jRo1Ml5++WUjKyvLOHPmjEO+B1ebQ/Tyyy8bfn5+RnJyspGVlWUuly5dcnZpdgoLC406derY3Z3oSkaPHm0kJycbhw4dMjZt2mTExsYaAQEBxokTJ5xdmum1114zUlJSjEOHDhk7d+40XnvtNcNmsxmrV692dmmmgQMHGg888ICxfPly49ChQ8aXX35pBAQEGK+++qrDjjF58mTDz8/PGDp0qLFp0yaH7dcwfrmLdvv27UabNm2Mp59+2ti+fbuxZ88es33Tpk1GpUqVjL/+9a9GRkaGMXHiRMPDw8PYtWuXQ+v4/vvvjUqVKhlvvfWWsX//fmPRokVG5cqVjX/+858OPc69OH/+vPl9STJmzJhhbN++3fj555+dXZqdlStXGitWrDD+7//+z1i9erXRokULo3379sbVq1edXdo9IxA5yYgRIwxJRkZGhrmtRYsWRnBwsLn+60B05coVo2/fvkb16tVL3HZ/q0BkGLe/7T43N9fo16+f4evra4SFhRkLFy4sMan666+/Nho2bGhUqlTpnm67v56rBSL96tEG+tUjDlzFqlWrDEnGvn37nF3KDfXr188ICQkxPD09jQceeMDo16+fceDAAWeXZWfw4MFGeHi44enpadSqVcvo1q2bS4UhwzCMvLw8Y8SIEUadOnUMb29vo379+sbrr79u5OfnO+wYhw4dMm9ycLQb/V369b8dS5YsMRo3bmx4enoaTZs2LbOJ7d98843x4IMPGl5eXkZkZKTx4YcflslxSmvdunU3/L4GDhzo7NLsfP7550b9+vXNnyXx8fHGuXPnnF2WQ9gMowI88hQAAKAM8RwiAABgeQQiAABgeQQiAABgeQQiAABgeQQiAABgeQQiAABgeQQiAABgeQQiAABgeQQiAE5z+PBh2Ww2paenO7sU7d27Vw8//LC8vb3VsmXLUu9n0KBB6tOnzz3XY7PZtGzZsnveD4A7QyACLG7QoEGy2WyaNm2a3fZly5aZLwOuyDZv3qyePXuqRo0a8vb2VrNmzTRjxowSLyqeOHGiqlSpon379ikpKUkPP/ywXnrpJbs+8+bNk81m08KFC+22Dxo0SI8++qgk6b333ivRDsD1EYgAyNvbW2+//bbOnj3r7FIc4urVq5KkpUuXqnPnzqpdu7bWrVunvXv3asSIEXrzzTfVv39/Xf/mooMHD6pjx44KDw9XzZo11aVLFyUnJ9vtd926dQoLCyuxPTk5WV27dpUk+fn5qXr16mV5egDKAIEIgGJjYxUcHKypU6fesH3SpEklLiPNmjVLdevWNdeLLxVNmTJFQUFBql69uiZPnqxr165pzJgx8vf3V+3atbVgwYIS+9+7d68eeeQReXt768EHH1RKSopd++7du9WjRw9VrVpVQUFBevbZZ3Xq1CmzPSYmRsOGDdPIkSMVEBCguLg4Xbx4US+++KJ+85vf6MMPP1TLli1Vt25dvfDCC/r444/173//W0uWLJH0y+WptLQ0TZ48WTabTZMmTVKXLl20b98+ZWdnm8dJSUnRa6+9ZheIDh06pJ9//lldunSx+x6ur+2Pf/yjXn31Vfn7+ys4OFiTJk2yO7/9+/erU6dO8vb2VpMmTZSYmFjiO9q1a5e6du0qHx8f1axZU0OHDtWFCxfM78fNzU0nT56UJJ05c0Zubm7q37+/+fk333xTHTt2lCSdPXtWAwYMUK1ateTj46NGjRrd8L8LYCUEIgByd3fXlClTNHv2bB09erTU+1m7dq2OHz+u9evXa8aMGZo4caIef/xx1ahRQ1u3btVLL72kP/zhDyWOMWbMGI0ePVrbt29XdHS0evfurdOnT0uSzp07p65du6pVq1b64YcftHLlSuXk5Oj3v/+93T4+/vhjeXp6atOmTZo3b55Wr16t06dP609/+lOJOnv37q3GjRvrX//6lyQpKytLTZs21ejRo5WVlaU//elP6tChgzw8PLRu3TpJ0o8//qjLly9ryJAhOn36tA4dOiTpl1Ejb29vRUdH3/R7+fjjj1WlShVt3bpV06dP1+TJk83QU1RUpCeffFKenp7aunWr5s2bp7Fjx9p9/uLFi4qLi1ONGjW0bds2ffHFF1qzZo2GDRsmSWratKlq1qxpBskNGzbYrUu/hLmYmBhJ0htvvKEff/xRK1asUEZGhubOnauAgIBb/JcF7n8EIgCSpP/+7/9Wy5YtNXHixFLvw9/fX++//74iIiI0ePBgRURE6NKlS/qf//kfNWrUSOPGjZOnp6c2btxo97lhw4apb9++ioqK0ty5c+Xn56ePPvpIkjRnzhy1atVKU6ZMUWRkpFq1aqX58+dr3bp1+umnn8x9NGrUSNOnT1dERIQiIiLMtqioqBvWGhkZafYJDg5WpUqVVLVqVQUHB6tq1aqqUqWKHnroIXM0KDk5WR07dpSXl5ceeeQRu+3R0dHy8vK66ffSvHlzTZw4UY0aNdJzzz2ntm3bKikpSZK0Zs0a7d27V5988olatGihTp06acqUKXafX7x4sa5cuaJPPvlEDz74oLp27ao5c+bo008/VU5Ojmw2mzp16mRX0/PPP6/8/Hzt3btXBQUF2rx5szp37ixJyszMVKtWrdS2bVvVrVtXsbGx6t279+3+8wL3NQIRANPbb7+tjz/+WBkZGaX6fNOmTeXm9v//WQkKClKzZs3MdXd3d9WsWVMnTpyw+9z1oyuVKlVS27ZtzRp27NihdevWqWrVquYSGRkp6Zd5P8XatGlzw5qunyd0t2JiYuxCRvEIS+fOne22F18uu5nmzZvbrYeEhJjfQUZGhsLCwhQaGmq2/3q0KSMjQy1atFCVKlXMbR06dFBRUZH27dtXoqaUlBR17drVDEnbtm1TQUGBOnToIEl6+eWX9dlnn6lly5Z69dVXtXnz5jv/UoD7FIEIgKlTp06Ki4vTuHHj7La7ubmVCBYFBQUlPu/h4WG3brPZbritqKjojmu6cOGCevfurfT0dLuleN5NsevDgiQ1btxYkm4a7jIyMsw+N9OlSxf99NNPOnbsmJKTk80RluLwcfDgQR05csScUH0z9/od3ImYmBj9+OOP2r9/v3788Ud17NjRDHQpKSlq27atKleuLEnq0aOHfv75Z73yyis6fvy4unXrdsNLi4CVEIgA2Jk2bZq++eYbpaammttq1aql7Oxsu1DkyGcHbdmyxfzztWvXlJaWZl7qat26tfbs2aO6deuqYcOGdsuvQ9D1unfvLn9/f7377rsl2r7++mvt379fTz311C3reuSRR+Tp6am//e1vunLlijkK1a5dO508eVLz5883L62VVlRUlI4cOaKsrCxz2/XfR3GfHTt26OLFi+a2TZs2yc3NTREREZKkZs2aqUaNGnrzzTfVsmVLVa1aVTExMUpJSbEb3SpWq1YtDRw4UP/85z81a9Ysffjhh6U+B+B+QCACYKdZs2YaMGCA3n//fXNbTEyMTp48qenTp+vgwYNKSEjQihUrHHbMhIQELV26VHv37lV8fLzOnj2rwYMHS5Li4+N15swZPfXUU9q2bZsOHjyoVatW6fnnny/xLKHrValSRR988IG++uorDR06VDt37tThw4f10UcfadCgQfrtb39bYmL2r/n4+Ojhhx/W7Nmz1aFDB7m7u0uSPD097bb/egTobsTGxqpx48YaOHCgduzYoQ0bNuj111+36zNgwAB5e3tr4MCB2r17t9atW6fhw4fr2WefVVBQkCSZ84gWLVpkhp/mzZsrPz9fSUlJ5uiWJE2YMEFfffWVDhw4oD179mj58uU3nWsFWAWBCEAJkydPtrukExUVpb/97W9KSEhQixYt9P333zv0Esu0adM0bdo0tWjRQhs3btTXX39t3vUUGhqqTZs2qbCwUN27d1ezZs00cuRIVa9e3W6+0o389re/1bp165SZmalHH31UERERmjlzpl5//XV99tlnd/TgyS5duuj8+fMlRlg6d+6s8+fP33b+0O24ublp6dKlunz5sh566CG98MILeuutt+z6VK5cWatWrdKZM2fUrl07/fa3v1W3bt00Z86cEjUVFhaatbq5ualTp06y2Wzm/CHpl0A3btw4NW/eXJ06dZK7u7s+++yzezoPoKKzGfcy4xAAAOA+wAgRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwPAIRAACwvP8HIOsnqScrU/4AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_insurance.isnull().sum()\n",
    "discrete_attirbutes = get_discrete_attributes(df_insurance)\n",
    "for attribute in discrete_attirbutes:\n",
    "  sns.histplot(data=df_insurance, x=attribute)\n",
    "  plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Displaying the Distribution \n",
    "def display_kde_distribution(attribute, df):\n",
    "  sns.histplot(x=attribute, data=df, kde=True, stat='density')\n",
    "\n",
    "display_kde_distribution(\"Building Dimension\", df_insurance)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "YearOfObservation       int64\n",
       "Insured_Period        float64\n",
       "Residential             int64\n",
       "Building_Painted       object\n",
       "Building_Fenced        object\n",
       "Garden                 object\n",
       "Settlement             object\n",
       "Building Dimension    float64\n",
       "Building_Type          object\n",
       "NumberOfWindows        object\n",
       "Claim                  object\n",
       "dtype: object"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dealing with outliers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. YearOfObservation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgoAAAGwCAYAAADIeE3bAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAfTklEQVR4nO3de5CV9X348Q+bhUVYdrnJAg0CNnLxBhEjQ0kxVpQYRgXRQYoRqVVTsTGa0Y7VipcY46jR6pg0EMVadaAUyaBVKwYwoyL1AohAQBBFRdRK5SYsC/vtH/44P1f2K7CAu0ter5mduOc8z3m+n304e94sZ/M0SSmlAACoRVF9LwAAaLiEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyCqu647V1dWxZs2aaNWqVTRp0mR/rgkAOEBSSrFx48bo3LlzFBXt/ucFdQ6FNWvWRJcuXeq6OwBQj95999345je/udvt6hwKrVq1KhyorKysrg8DAHyNNmzYEF26dCm8ju9OnUNh5z83lJWVCQUAaGT29G0D3swIAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZBXX9wJq8+GHH8b69evrexkAjVp5eXlUVFTU9zJo5BpcKHz44Ydx3g/Pj6ptlfW9FIBGrWmzknj43x4SC+yTBhcK69evj6ptlbHl8BOjunl5fS+HOija8mkcsuoPsaX7oKg+pHV9Lwf+JBVtXR/x1nOxfv16ocA+aXChsFN18/Kobtm+vpfBPqg+pLVzCNDIeTMjAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyGpwoVBZWfn5f1Rvr9+FAEA92bp1ayxfvjy2bt1a30tpeKGwdu3aiIgoqtxUzysBgPqxevXquPjii2P16tX1vZSGFwoAQMMhFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsor3dMPKysqorKwsfL5hw4YDsiAA9p933nmnvpdAHTSk87bHoXDrrbfGjTfeeCDXAsB+dsstt9T3Emjk9jgUrrnmmrjyyisLn2/YsCG6dOlyQBYFwP5x7bXXRteuXet7Geyld955p8FE3h6HQklJSZSUlBzItQCwn3Xt2jV69OhR38ugEfNmRgAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQ1uFDo2LFjRERUl5TW80oAoH4cdthhMWHChDjssMPqeylRXN8L+LKSkpLP/6OowS0NAL4WzZs3jx49etT3MiKiAf5EAQBoOIQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZxfW9gJyirevrewnUUdGWT2v8L/D18z2U/aXBhUJ5eXk0bVYS8dZz9b0U9tEhq/5Q30uAP2lNm5VEeXl5fS+DRq7BhUJFRUU8/G8Pxfr1ahhgX5SXl0dFRUV9L4NGrsGFQsTnseAPNwDUP29mBACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAABkCQUAIEsoAABZQgEAyBIKAECWUAAAsoQCAJAlFACALKEAAGQJBQAgSygAAFlCAQDIEgoAQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKK67pjSikiIjZs2LDfFgMAHFg7X7d3vo7vTp1DYePGjRER0aVLl7o+BABQTzZu3Bjl5eW73a5J2tOk+JLq6upYs2ZNtGrVKpo0aVKXh6jVhg0bokuXLvHuu+9GWVnZfnvchuRgn9F8jd/BPuPBPl/EwT+j+eoupRQbN26Mzp07R1HR7t+BUOefKBQVFcU3v/nNuu6+W2VlZQflyf+ig31G8zV+B/uMB/t8EQf/jOarmz35ScJO3swIAGQJBQAgq8GFQklJSYwfPz5KSkrqeykHzME+o/kav4N9xoN9voiDf0bzfX3q/GZGAODg1+B+ogAANBxCAQDIEgoAQJZQAACy6hwKt956a3znO9+JVq1aRYcOHWLYsGGxbNmyGtts3bo1xo0bF+3atYvS0tIYMWJEfPjhhzW2+fGPfxz9+vWLkpKS6Nu37y7HmTNnTpx55pnRqVOnaNmyZfTt2zceeeSR3a5v9erVMXTo0GjRokV06NAhrrrqqti+fftBNWOTJk12+Zg8eXKDm2/ZsmVx0kknRUVFRTRv3jwOP/zwuO6666Kqquor17ev57Chz7ev5+/rnPGLVqxYEa1atYrWrVvvdn2N5RzWdb7G8hx8++23a13rSy+99JXra0zfR+s6Y2M5hxGf/z8q3nHHHdGjR48oKSmJP/uzP4tbbrnlK9e3bt26GD16dJSVlUXr1q3jwgsvjE2bNu3xfBH7EArPPfdcjBs3Ll566aWYOXNmVFVVxamnnhqbN28ubHPFFVfE448/HlOnTo3nnnsu1qxZE2edddYuj/U3f/M3MXLkyFqP8+KLL8axxx4b06ZNi9dffz3Gjh0b559/fjzxxBPZte3YsSOGDh0a27ZtixdffDH+9V//NR588MG4/vrrD5oZd5o0aVJ88MEHhY9hw4Y1uPmaNm0a559/fjzzzDOxbNmyuPvuu2PixIkxfvz47Nr2xzlsyPPttC/n7+uccaeqqqoYNWpU/OVf/uVu19aYzmFd5tupMTwHd3r22WdrrLVfv37ZbRvb99G6zLhTYzmHl19+efz2t7+NO+64I/74xz/GjBkz4oQTTvjK9Y0ePToWL14cM2fOjCeeeCL+8Ic/xMUXX7zH80VERNpPPvrooxQR6bnnnksppfTpp5+mpk2bpqlTpxa2Wbp0aYqINHfu3F32Hz9+fOrTp88eHesHP/hBGjt2bPb+J598MhUVFaW1a9cWbvv1r3+dysrKUmVl5R5OtKuGNGNKKUVEmj59+h6vf3e+zvmuuOKK9N3vfjd7/4E4hw1pvpT2//lL6cDPePXVV6fzzjsvTZo0KZWXl3/lWhrjOdyb+VJqPM/BVatWpYhI8+fP3+O1NLbvo3WZMaXGcw6XLFmSiouL0x//+Mc9XsuSJUtSRKSXX365cNtTTz2VmjRpkt5///09fpz99h6F9evXR0RE27ZtIyLi1Vdfjaqqqhg8eHBhm169esVhhx0Wc+fO3edj7TxObebOnRvHHHNMVFRUFG4bMmRIbNiwIRYvXrxPx41oGDPuNG7cuGjfvn2ccMIJ8cADD+zxZUNzx4w48POtWLEinn766TjxxBOz2xyIc9iQ5ttpf56/iAM746xZs2Lq1Klx33337dH2je0c7u18OzWm5+AZZ5wRHTp0iO9+97sxY8aMr9y2sX4f3ZsZd2oM5/Dxxx+Pww8/PJ544ono3r17dOvWLf72b/821q1bl91n7ty50bp16zj++OMLtw0ePDiKiopi3rx5e3zsOl8U6ouqq6vjJz/5SQwcODCOPvroiIhYu3ZtNGvWbJd/56uoqIi1a9fW+Vj//u//Hi+//HL85je/yW6zdu3aGn+4dx5353110dBmjIi46aab4q/+6q+iRYsW8cwzz8Sll14amzZtih//+Md7fcyvY76/+Iu/iNdeey0qKyvj4osvjptuuim77f4+hw1tvoj9e/4iDuyMn3zySVxwwQXx8MMP7/EFahrTOazLfBGN5zlYWload955ZwwcODCKiopi2rRpMWzYsPjd734XZ5xxRq37NLbvo3WZMaLxnMO33nor3nnnnZg6dWo89NBDsWPHjrjiiivi7LPPjlmzZtW6z9q1a6NDhw41bisuLo62bdvu1bH3SyiMGzcu3njjjXj++ef3x8NlzZ49O8aOHRsTJ06Mo4466oAe68sa4oz/9E//VPjvb3/727F58+a4/fbb6/QH/OuYb8qUKbFx48ZYuHBhXHXVVXHHHXfE1VdffcCO90UNcb79ef4iDuyMF110Ufz1X/91DBo0aL8/9p5qiPM1ludg+/bt48orryx8/p3vfCfWrFkTt99++1e+iO5vDXHGxnIOq6uro7KyMh566KHo0aNHRETcf//90a9fv1i2bFn07Nlzvx9zp33+p4fLLrssnnjiiZg9e3aNy0537Ngxtm3bFp9++mmN7T/88MPo2LHjXh/nueeei9NPPz3uuuuuOP/8879y244dO+7yjtKdn9fl2A1xxtr0798/3nvvvaisrNyr/b6u+bp06RJHHnlkjBo1Kn7xi1/EDTfcEDt27Kh12/15DhvifLWp6/mLOPAzzpo1K+64444oLi6O4uLiuPDCC2P9+vVRXFwcDzzwQK37NKZzWJf5atPQn4NfXuuKFSuy9zfW76NftLsZc/s0xHPYqVOnKC4uLkRCRETv3r0j4vPfTqlNx44d46OPPqpx2/bt22PdunV7dew6h0JKKS677LKYPn16zJo1K7p3717j/n79+kXTpk3j97//feG2ZcuWxerVq2PAgAF7daw5c+bE0KFD47bbbtujd2sOGDAgFi1aVOMLNHPmzCgrK4sjjzxyj4/bkGeszYIFC6JNmzZ7fBGRr3O+L6uuro6qqqqorq6u9f79cQ4b8ny12dvzF/H1zTh37txYsGBB4eOmm26KVq1axYIFC2L48OG17tOYzmFd5qtNY3oOLliwIDp16pS9vzF+H/2y3c2Y26chnsOBAwfG9u3bY+XKlYXbli9fHhERXbt2rXWfAQMGxKeffhqvvvpq4bZZs2ZFdXV19O/ff4+PXeffevi7v/u7VF5enubMmZM++OCDwsdnn31W2OZHP/pROuyww9KsWbPSK6+8kgYMGJAGDBhQ43HefPPNNH/+/HTJJZekHj16pPnz56f58+cX3lU7a9as1KJFi3TNNdfUOM4nn3xSeIzHHnss9ezZs/D59u3b09FHH51OPfXUtGDBgvT000+nQw89NF1zzTUHzYwzZsxIEydOTIsWLUpvvvlm+tWvfpVatGiRrr/++gY338MPP5ymTJmSlixZklauXJmmTJmSOnfunEaPHp2db3+cw4Y83/44f1/njF9W228FNOZzWJf5GtNz8MEHH0yPPvpoWrp0aVq6dGm65ZZbUlFRUXrggQey8zW276N1mbExncMdO3ak4447Lg0aNCi99tpr6ZVXXkn9+/dPp5xySuEx5s2bl3r27Jnee++9wm3f//7307e//e00b9689Pzzz6cjjjgijRo1ao/nSymlOodCRNT6MWnSpMI2W7ZsSZdeemlq06ZNatGiRRo+fHj64IMPajzOiSeeWOvjrFq1KqWU0pgxY2q9/8QTTyw8xqRJk9KXm+ftt99Op512WjrkkENS+/bt009/+tNUVVV10Mz41FNPpb59+6bS0tLUsmXL1KdPn/Qv//IvaceOHQ1uvsmTJ6fjjjuusNYjjzwy/fznP09btmzJzpfSvp/Dhjzf/jh/X+eMX1bbC2ljPod1ma8xPQcffPDB1Lt379SiRYtUVlaWTjjhhBq/rlfbfCk1ru+jdZmxMZ3DlFJ6//3301lnnZVKS0tTRUVFuuCCC2r8hXL27Nm77PPJJ5+kUaNGpdLS0lRWVpbGjh2bNm7cuMfzpZSSy0wDAFmu9QAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQJRQAgCyhAI3YhAkTokuXLlFUVBR33333Xu17wQUXxLBhww7IuupLkyZN4ne/+119LwMOKkIBapFSisGDB8eQIUN2ue9Xv/pVtG7dOt57770DcuwdO3bEXXfdFcccc0w0b9482rRpE6eddlq88MILNbbbsGFDXHbZZfEP//AP8f777xcuJrZly5YYP3589OjRI0pKSqJ9+/ZxzjnnxOLFiw/IeuvDDTfcEH379t3l9g8++CBOO+20r39BcBATClCLJk2axKRJk2LevHnxm9/8pnD7qlWr4uqrr4577723xqVk94eqqqpIKcW5554bN910U1x++eWxdOnSmDNnTnTp0iW+973v1fjb8urVq6OqqiqGDh0anTp1ihYtWkRlZWUMHjw4HnjggfjZz34Wy5cvjyeffDK2b98e/fv3j5deemm/rnlvVVVVHdDH79ix415dfRPYA3t1ZQj4E/Pggw+m0tLS9NZbb6Xq6up00kknpeHDh6dFixal73//+6lly5apQ4cO6bzzzksff/xxYb+nnnoqDRw4MJWXl6e2bdumoUOHphUrVhTuX7VqVYqINHny5DRo0KBUUlKSJk2alCZPnpwiIs2YMWOXtZx11lmpXbt2adOmTYWL28SXLh7zi1/8IjVp0iQtWLCgxr47duxIxx9/fDryyCNTdXV1Sunzi5GdeeaZ6YYbbkjt27dPrVq1SpdcckmNKypOnTo1HX300al58+apbdu26eSTT06bNm0q3D9x4sTUq1evVFJSknr27Jnuu+++r5zxn//5n1Pz5s3Tk08+WWN9jz32WCotLU2bN29OKaV09dVXpyOOOCIdcsghqXv37um6665L27ZtSymlWmffeQGeiEjTp08vPO7rr7+eTjrppML6L7roohoXxNn5Nbj99ttTx44dU9u2bdOll15aOBawD1ePhD8VZ555Zvre976X7rnnnnTooYemjz76qHC53aVLl6bXXnstnXLKKemkk04q7PMf//Efadq0aYVLx55++unpmGOOKVyVbueLaLdu3dK0adPSW2+9ldasWZPOOOOM1KNHj1rX8cILLxReCD/77LP07LPPpohI//3f/50++OCDtH379nTsscemU089tdb9H3nkkRQRaf78+Smlz18kS0tL08iRI9Mbb7yRnnjiiXTooYemf/zHf0wppbRmzZpUXFycfvnLX6ZVq1al119/Pd13332FF9qHH344derUqbD+adOmpbZt26YHH3zwK2c8++yz03nnnVdjbSNGjKhx280335xeeOGFtGrVqjRjxoxUUVGRbrvttpRSSp999ln66U9/mo466qhdLun7xVDYtGlT6tSpUzrrrLPSokWL0u9///vUvXv3NGbMmMJxxowZk8rKytKPfvSjtHTp0vT444+nFi1apAkTJuz2zwX8qRAKsBsffvhhat++fSoqKkrTp09PN9988y4vxu+++26KiLRs2bJaH+Pjjz9OEZEWLVqUUvr/L6J33313je169eqVzjzzzFofY926dSkiCi+Y8+fP3+WSss2bN0+XX355rfu/9tprKSLSlClTUkqfv0i2bdu28Lf4lFL69a9/nUpLS9OOHTvSq6++miIivf3227U+3p//+Z+nRx99tMZtN998cxowYMBXzjh9+vQaPz1Yv359at68eXrqqadqPU5KKd1+++2pX79+hc/Hjx+f+vTps8t2XwyFCRMmpDZt2tT4Cch//ud/pqKiorR27drC16Br165p+/bthW3OOeecNHLkyOxa4E+N9yjAbnTo0CEuueSS6N27dwwbNiwWLlwYs2fPjtLS0sJHr169IiJi5cqVERHx5ptvxqhRo+Lwww+PsrKy6NatW0R8/r6CLzr++ON3OV7axyu/783+ffr0iRYtWhQ+HzBgQGzatCnefffd6NOnT5x88slxzDHHxDnnnBMTJ06M//3f/42IiM2bN8fKlSvjwgsvrPF1+NnPflb4Guz05Rl/8IMfRNOmTWPGjBkRETFt2rQoKyuLwYMHF7aZMmVKDBw4MDp27BilpaVx3XXX7fK1252lS5dGnz59omXLloXbBg4cGNXV1bFs2bLCbUcddVR84xvfKHzeqVOn+Oijj/bqWHAwK67vBUBjUFxcHMXFnz9dNm3aFKeffnrcdtttu2zXqVOniIg4/fTTo2vXrjFx4sTo3LlzVFdXx9FHHx3btm2rsf0XX8QiInr06BFLly6tdQ07b+/Ro0d2nfu6/xd94xvfiJkzZ8aLL74YzzzzTNx7771x7bXXxrx58wpxMXHixOjfv/8u+33Rl2ds1qxZnH322fHoo4/GueeeG48++miMHDmy8PWdO3dujB49Om688cYYMmRIlJeXx+TJk+POO+/co3XvraZNm9b4vEmTJlFdXX1AjgWNkZ8owF467rjjYvHixdGtW7f41re+VeOjZcuW8cknn8SyZcviuuuui5NPPjl69+5d+Jv47px77rnx5ptvxuOPP77LfXfeeWe0a9cuTjnllK/c/9lnn42FCxfWuL26ujruuuuuOPLII6NPnz6F2xcuXBhbtmwpfP7SSy9FaWlpdOnSJSI+f9EcOHBg3HjjjTF//vxo1qxZTJ8+PSoqKqJz587x1ltv7fI16N69+27nHD16dDz99NOxePHimDVrVowePbpw34svvhhdu3aNa6+9No4//vg44ogj4p133qmxf7NmzWLHjh1feYzevXvHwoULY/PmzYXbXnjhhSgqKoqePXvudo3A54QC7KVx48bFunXrYtSoUfHyyy/HypUr47/+679i7NixsWPHjmjTpk20a9cuJkyYECtWrIhZs2bFlVdeuUePfe6558bw4cNjzJgxcf/998fbb78dr7/+elxyySUxY8aM+O1vf7vL39C/6IorrogTTjghTj/99Jg6dWqsXr06Xn755RgxYkQsXbo07r///mjSpElh+23btsWFF14YS5YsiSeffDLGjx8fl112WRQVFcW8efPi5z//ebzyyiuxevXqeOyxx+Ljjz+O3r17R0TEjTfeGLfeemvcc889sXz58li0aFFMmjQpfvnLX+52zkGDBkXHjh1j9OjR0b179xo/lTjiiCNi9erVMXny5Fi5cmXcc889MX369Br7d+vWLVatWhULFiyI//mf/4nKyspdjjF69Oho3rx5jBkzJt54442YPXt2/P3f/3388Ic/jIqKit2uEfh/6vtNEtAYfPnNc8uXL0/Dhw9PrVu3Toccckjq1atX+slPflL41cOZM2em3r17p5KSknTsscemOXPm1Hij3c43+u38DYQvqqqqSrfffns66qijUrNmzVJZWVkaMmRIev7552tsV9ubGVNKafPmzenaa69N3/rWt1LTpk1T27Zt04gRIwpvpNxp568GXn/99aldu3aptLQ0XXTRRWnr1q0ppZSWLFmShgwZkg499NBUUlKSevToke69994aj/HII4+kvn37pmbNmqU2bdqkQYMGpccee2y3M6b0+a9ARkS6/vrrd7nvqquuKqxp5MiR6a677krl5eWF+7du3ZpGjBiRWrduvV9+PfKLLr/88nTiiSfWumb4U9QkpX185xQAcNDyTw8AQJZQAACyhAIAkCUUAIAsoQAAZAkFACBLKAAAWUIBAMgSCgBAllAAALKEAgCQ9X9edrstsHnpgwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='YearOfObservation', ylabel='Count'>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=df_insurance, x='YearOfObservation')\n",
    "plt.show()\n",
    "sns.histplot(data=df_insurance, x='YearOfObservation')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Insured_Period"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Insured_Period', ylabel='Count'>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(data=df_insurance, x='Insured_Period')\n",
    "plt.show()\n",
    "sns.histplot(data=df_insurance, x='Insured_Period')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance['Residential'].unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Number of windows transformation: 'without' --> 0, '>=10' --> 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building_Painted</th>\n",
       "      <th>Building_Fenced</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Building_Painted  \\\n",
       "0                  2012             1.0            1                N   \n",
       "1                  2012             1.0            0                N   \n",
       "2                  2013             1.0            1                V   \n",
       "3                  2016             0.5            0                N   \n",
       "4                  2012             1.0            0                N   \n",
       "...                 ...             ...          ...              ...   \n",
       "5007               2013             1.0            0                N   \n",
       "5008               2012             0.5            0                V   \n",
       "5009               2015             1.0            1                V   \n",
       "5010               2012             0.5            0                V   \n",
       "5011               2013             1.0            1                V   \n",
       "\n",
       "     Building_Fenced Garden Settlement  Building Dimension    Building_Type  \\\n",
       "0                  V      V          U            0.100581      Wood-framed   \n",
       "1                  V      V          U           -0.097093  Non-combustible   \n",
       "2                  N      O          R            2.187791  Non-combustible   \n",
       "3                  V      V          U           -0.271512      Wood-framed   \n",
       "4                  V      V          U           -0.097093  Non-combustible   \n",
       "...              ...    ...        ...                 ...              ...   \n",
       "5007               V      V          U           -0.300581         Ordinary   \n",
       "5008               N      O          R           -0.038953   Fire-resistive   \n",
       "5009               N      O          R           -0.341279         Ordinary   \n",
       "5010               N      O          R           -0.308721   Fire-resistive   \n",
       "5011               V      V          U            0.000000      Wood-framed   \n",
       "\n",
       "      NumberOfWindows Claim  \n",
       "0                   0   non  \n",
       "1                   0   non  \n",
       "2                   4   oui  \n",
       "3                   0   oui  \n",
       "4                   0   non  \n",
       "...               ...   ...  \n",
       "5007                0   oui  \n",
       "5008                4   non  \n",
       "5009                3   non  \n",
       "5010                4   non  \n",
       "5011                0   non  \n",
       "\n",
       "[5012 rows x 11 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def transform_num_windows(df:pd.DataFrame)->None:\n",
    "  df['NumberOfWindows'] = np.where(df['NumberOfWindows'] == 'without', '0', np.where(df['NumberOfWindows'] == '>=10', '10', df['NumberOfWindows']))\n",
    "  ## casting to int\n",
    "  df['NumberOfWindows'] = pd.to_numeric(df['NumberOfWindows'])\n",
    "\n",
    "transform_num_windows(df_insurance)\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Change the 'insured_period' from 0.5 to 0 and cast to int"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 0.])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def transform_insured_period(df:pd.DataFrame):\n",
    "  df['Insured_Period'] = df['Insured_Period'].apply(lambda x: 0 if x == 0.5 else x)\n",
    "\n",
    "transform_insured_period(df_insurance)\n",
    "df_insurance['Insured_Period'].unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoding non numerical feautres (some ordinal: Building type, and some one hot)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building_Painted</th>\n",
       "      <th>Building_Fenced</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>N</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Building_Painted  \\\n",
       "0                  2012             1.0            1                N   \n",
       "1                  2012             1.0            0                N   \n",
       "2                  2013             1.0            1                V   \n",
       "3                  2016             0.0            0                N   \n",
       "4                  2012             1.0            0                N   \n",
       "...                 ...             ...          ...              ...   \n",
       "5007               2013             1.0            0                N   \n",
       "5008               2012             0.0            0                V   \n",
       "5009               2015             1.0            1                V   \n",
       "5010               2012             0.0            0                V   \n",
       "5011               2013             1.0            1                V   \n",
       "\n",
       "     Building_Fenced Garden Settlement  Building Dimension    Building_Type  \\\n",
       "0                  V      V          U            0.100581      Wood-framed   \n",
       "1                  V      V          U           -0.097093  Non-combustible   \n",
       "2                  N      O          R            2.187791  Non-combustible   \n",
       "3                  V      V          U           -0.271512      Wood-framed   \n",
       "4                  V      V          U           -0.097093  Non-combustible   \n",
       "...              ...    ...        ...                 ...              ...   \n",
       "5007               V      V          U           -0.300581         Ordinary   \n",
       "5008               N      O          R           -0.038953   Fire-resistive   \n",
       "5009               N      O          R           -0.341279         Ordinary   \n",
       "5010               N      O          R           -0.308721   Fire-resistive   \n",
       "5011               V      V          U            0.000000      Wood-framed   \n",
       "\n",
       "      NumberOfWindows Claim  \n",
       "0                   0   non  \n",
       "1                   0   non  \n",
       "2                   4   oui  \n",
       "3                   0   oui  \n",
       "4                   0   non  \n",
       "...               ...   ...  \n",
       "5007                0   oui  \n",
       "5008                4   non  \n",
       "5009                3   non  \n",
       "5010                4   non  \n",
       "5011                0   non  \n",
       "\n",
       "[5012 rows x 11 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "encoding Building_Fenced - one hot encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building_Painted</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Building_Painted Garden  \\\n",
       "0                  2012             1.0            1                N      V   \n",
       "1                  2012             1.0            0                N      V   \n",
       "2                  2013             1.0            1                V      O   \n",
       "3                  2016             0.0            0                N      V   \n",
       "4                  2012             1.0            0                N      V   \n",
       "...                 ...             ...          ...              ...    ...   \n",
       "5007               2013             1.0            0                N      V   \n",
       "5008               2012             0.0            0                V      O   \n",
       "5009               2015             1.0            1                V      O   \n",
       "5010               2012             0.0            0                V      O   \n",
       "5011               2013             1.0            1                V      V   \n",
       "\n",
       "     Settlement  Building Dimension    Building_Type  NumberOfWindows Claim  \\\n",
       "0             U            0.100581      Wood-framed                0   non   \n",
       "1             U           -0.097093  Non-combustible                0   non   \n",
       "2             R            2.187791  Non-combustible                4   oui   \n",
       "3             U           -0.271512      Wood-framed                0   oui   \n",
       "4             U           -0.097093  Non-combustible                0   non   \n",
       "...         ...                 ...              ...              ...   ...   \n",
       "5007          U           -0.300581         Ordinary                0   oui   \n",
       "5008          R           -0.038953   Fire-resistive                4   non   \n",
       "5009          R           -0.341279         Ordinary                3   non   \n",
       "5010          R           -0.308721   Fire-resistive                4   non   \n",
       "5011          U            0.000000      Wood-framed                0   non   \n",
       "\n",
       "      Fenced=T  Fenced=F  \n",
       "0          1.0       0.0  \n",
       "1          1.0       0.0  \n",
       "2          0.0       1.0  \n",
       "3          1.0       0.0  \n",
       "4          1.0       0.0  \n",
       "...        ...       ...  \n",
       "5007       1.0       0.0  \n",
       "5008       0.0       1.0  \n",
       "5009       0.0       1.0  \n",
       "5010       0.0       1.0  \n",
       "5011       1.0       0.0  \n",
       "\n",
       "[5012 rows x 12 columns]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def encode_true_false(df:pd.DataFrame, feature:str,categories:list, result_cols:list):\n",
    "  # this encoder will transform N to 0 and V to 1 in the given feature\n",
    "  # save the result into new columns of which the names are given in result_cols\n",
    "  enc = OneHotEncoder(sparse_output=False, categories=[categories])\n",
    "  df[result_cols] = enc.fit_transform(df.loc[:, [feature]]) # will be added at the end\n",
    "  df.drop(axis=1, labels=[feature], inplace=True)\n",
    "\n",
    "encode_true_false(df_insurance, 'Building_Fenced',['V', 'N'], ['Fenced=T', 'Fenced=F'])\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoding Building_Painted - one hot encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "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>YearOfObservation</th>\n",
       "      <th>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>2015</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>2012</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>2013</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      YearOfObservation  Insured_Period  Residential Garden Settlement  \\\n",
       "0                  2012             1.0            1      V          U   \n",
       "1                  2012             1.0            0      V          U   \n",
       "2                  2013             1.0            1      O          R   \n",
       "3                  2016             0.0            0      V          U   \n",
       "4                  2012             1.0            0      V          U   \n",
       "...                 ...             ...          ...    ...        ...   \n",
       "5007               2013             1.0            0      V          U   \n",
       "5008               2012             0.0            0      O          R   \n",
       "5009               2015             1.0            1      O          R   \n",
       "5010               2012             0.0            0      O          R   \n",
       "5011               2013             1.0            1      V          U   \n",
       "\n",
       "      Building Dimension    Building_Type  NumberOfWindows Claim  Fenced=T  \\\n",
       "0               0.100581      Wood-framed                0   non       1.0   \n",
       "1              -0.097093  Non-combustible                0   non       1.0   \n",
       "2               2.187791  Non-combustible                4   oui       0.0   \n",
       "3              -0.271512      Wood-framed                0   oui       1.0   \n",
       "4              -0.097093  Non-combustible                0   non       1.0   \n",
       "...                  ...              ...              ...   ...       ...   \n",
       "5007           -0.300581         Ordinary                0   oui       1.0   \n",
       "5008           -0.038953   Fire-resistive                4   non       0.0   \n",
       "5009           -0.341279         Ordinary                3   non       0.0   \n",
       "5010           -0.308721   Fire-resistive                4   non       0.0   \n",
       "5011            0.000000      Wood-framed                0   non       1.0   \n",
       "\n",
       "      Fenced=F  Painted=T  Painted=F  \n",
       "0          0.0        0.0        1.0  \n",
       "1          0.0        0.0        1.0  \n",
       "2          1.0        1.0        0.0  \n",
       "3          0.0        0.0        1.0  \n",
       "4          0.0        0.0        1.0  \n",
       "...        ...        ...        ...  \n",
       "5007       0.0        0.0        1.0  \n",
       "5008       1.0        1.0        0.0  \n",
       "5009       1.0        1.0        0.0  \n",
       "5010       1.0        1.0        0.0  \n",
       "5011       0.0        1.0        0.0  \n",
       "\n",
       "[5012 rows x 13 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encode_true_false(df_insurance, 'Building_Painted',['V', 'N'], ['Painted=T', 'Painted=F'])\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Consider Discretisation of num windows (Depending on a strong argument and interpertation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Remove the YearOfObservation feature, because all new observations after depolyment will be different (new years)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "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>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>Non-combustible</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>Ordinary</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>Fire-resistive</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Wood-framed</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Insured_Period  Residential Garden Settlement  Building Dimension  \\\n",
       "0                1.0            1      V          U            0.100581   \n",
       "1                1.0            0      V          U           -0.097093   \n",
       "2                1.0            1      O          R            2.187791   \n",
       "3                0.0            0      V          U           -0.271512   \n",
       "4                1.0            0      V          U           -0.097093   \n",
       "...              ...          ...    ...        ...                 ...   \n",
       "5007             1.0            0      V          U           -0.300581   \n",
       "5008             0.0            0      O          R           -0.038953   \n",
       "5009             1.0            1      O          R           -0.341279   \n",
       "5010             0.0            0      O          R           -0.308721   \n",
       "5011             1.0            1      V          U            0.000000   \n",
       "\n",
       "        Building_Type  NumberOfWindows Claim  Fenced=T  Fenced=F  Painted=T  \\\n",
       "0         Wood-framed                0   non       1.0       0.0        0.0   \n",
       "1     Non-combustible                0   non       1.0       0.0        0.0   \n",
       "2     Non-combustible                4   oui       0.0       1.0        1.0   \n",
       "3         Wood-framed                0   oui       1.0       0.0        0.0   \n",
       "4     Non-combustible                0   non       1.0       0.0        0.0   \n",
       "...               ...              ...   ...       ...       ...        ...   \n",
       "5007         Ordinary                0   oui       1.0       0.0        0.0   \n",
       "5008   Fire-resistive                4   non       0.0       1.0        1.0   \n",
       "5009         Ordinary                3   non       0.0       1.0        1.0   \n",
       "5010   Fire-resistive                4   non       0.0       1.0        1.0   \n",
       "5011      Wood-framed                0   non       1.0       0.0        1.0   \n",
       "\n",
       "      Painted=F  \n",
       "0           1.0  \n",
       "1           1.0  \n",
       "2           0.0  \n",
       "3           1.0  \n",
       "4           1.0  \n",
       "...         ...  \n",
       "5007        1.0  \n",
       "5008        0.0  \n",
       "5009        0.0  \n",
       "5010        0.0  \n",
       "5011        0.0  \n",
       "\n",
       "[5012 rows x 12 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_insurance.drop(axis=1, labels=['YearOfObservation'], inplace=True)\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encode Building_Type - using ordinal encoder, because there's a degradation of types (it's like: weak(Wood-framed):0, average(Ordinary):1, strong:2, very strong:3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "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>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Garden</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>O</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>V</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Insured_Period  Residential Garden Settlement  Building Dimension  \\\n",
       "0                1.0            1      V          U            0.100581   \n",
       "1                1.0            0      V          U           -0.097093   \n",
       "2                1.0            1      O          R            2.187791   \n",
       "3                0.0            0      V          U           -0.271512   \n",
       "4                1.0            0      V          U           -0.097093   \n",
       "...              ...          ...    ...        ...                 ...   \n",
       "5007             1.0            0      V          U           -0.300581   \n",
       "5008             0.0            0      O          R           -0.038953   \n",
       "5009             1.0            1      O          R           -0.341279   \n",
       "5010             0.0            0      O          R           -0.308721   \n",
       "5011             1.0            1      V          U            0.000000   \n",
       "\n",
       "      Building_Type  NumberOfWindows Claim  Fenced=T  Fenced=F  Painted=T  \\\n",
       "0               0.0                0   non       1.0       0.0        0.0   \n",
       "1               2.0                0   non       1.0       0.0        0.0   \n",
       "2               2.0                4   oui       0.0       1.0        1.0   \n",
       "3               0.0                0   oui       1.0       0.0        0.0   \n",
       "4               2.0                0   non       1.0       0.0        0.0   \n",
       "...             ...              ...   ...       ...       ...        ...   \n",
       "5007            1.0                0   oui       1.0       0.0        0.0   \n",
       "5008            3.0                4   non       0.0       1.0        1.0   \n",
       "5009            1.0                3   non       0.0       1.0        1.0   \n",
       "5010            3.0                4   non       0.0       1.0        1.0   \n",
       "5011            0.0                0   non       1.0       0.0        1.0   \n",
       "\n",
       "      Painted=F  \n",
       "0           1.0  \n",
       "1           1.0  \n",
       "2           0.0  \n",
       "3           1.0  \n",
       "4           1.0  \n",
       "...         ...  \n",
       "5007        1.0  \n",
       "5008        0.0  \n",
       "5009        0.0  \n",
       "5010        0.0  \n",
       "5011        0.0  \n",
       "\n",
       "[5012 rows x 12 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def encode_Building_type(df):\n",
    "  enc = OrdinalEncoder(categories=[['Wood-framed', 'Ordinary', 'Non-combustible', 'Fire-resistive']])\n",
    "  df['Building_Type'] = enc.fit_transform(df.loc[:, ['Building_Type']])\n",
    "\n",
    "encode_Building_type(df_insurance)\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encode Garden"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "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>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Settlement</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "      <th>Garden=T</th>\n",
       "      <th>Garden=F</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>U</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>U</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>R</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>U</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 13 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Insured_Period  Residential Settlement  Building Dimension  \\\n",
       "0                1.0            1          U            0.100581   \n",
       "1                1.0            0          U           -0.097093   \n",
       "2                1.0            1          R            2.187791   \n",
       "3                0.0            0          U           -0.271512   \n",
       "4                1.0            0          U           -0.097093   \n",
       "...              ...          ...        ...                 ...   \n",
       "5007             1.0            0          U           -0.300581   \n",
       "5008             0.0            0          R           -0.038953   \n",
       "5009             1.0            1          R           -0.341279   \n",
       "5010             0.0            0          R           -0.308721   \n",
       "5011             1.0            1          U            0.000000   \n",
       "\n",
       "      Building_Type  NumberOfWindows Claim  Fenced=T  Fenced=F  Painted=T  \\\n",
       "0               0.0                0   non       1.0       0.0        0.0   \n",
       "1               2.0                0   non       1.0       0.0        0.0   \n",
       "2               2.0                4   oui       0.0       1.0        1.0   \n",
       "3               0.0                0   oui       1.0       0.0        0.0   \n",
       "4               2.0                0   non       1.0       0.0        0.0   \n",
       "...             ...              ...   ...       ...       ...        ...   \n",
       "5007            1.0                0   oui       1.0       0.0        0.0   \n",
       "5008            3.0                4   non       0.0       1.0        1.0   \n",
       "5009            1.0                3   non       0.0       1.0        1.0   \n",
       "5010            3.0                4   non       0.0       1.0        1.0   \n",
       "5011            0.0                0   non       1.0       0.0        1.0   \n",
       "\n",
       "      Painted=F  Garden=T  Garden=F  \n",
       "0           1.0       1.0       0.0  \n",
       "1           1.0       1.0       0.0  \n",
       "2           0.0       0.0       1.0  \n",
       "3           1.0       1.0       0.0  \n",
       "4           1.0       1.0       0.0  \n",
       "...         ...       ...       ...  \n",
       "5007        1.0       1.0       0.0  \n",
       "5008        0.0       0.0       1.0  \n",
       "5009        0.0       0.0       1.0  \n",
       "5010        0.0       0.0       1.0  \n",
       "5011        0.0       1.0       0.0  \n",
       "\n",
       "[5012 rows x 13 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encode_true_false(df_insurance, 'Garden', ['V','O'], ['Garden=T', 'Garden=F'])\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encode Settelement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "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>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "      <th>Garden=T</th>\n",
       "      <th>Garden=F</th>\n",
       "      <th>Settle=R</th>\n",
       "      <th>Settle=U</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.100581</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.187791</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.271512</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.097093</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5007</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.300581</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5008</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.038953</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5009</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.341279</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5010</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.308721</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5011</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5012 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Insured_Period  Residential  Building Dimension  Building_Type  \\\n",
       "0                1.0            1            0.100581            0.0   \n",
       "1                1.0            0           -0.097093            2.0   \n",
       "2                1.0            1            2.187791            2.0   \n",
       "3                0.0            0           -0.271512            0.0   \n",
       "4                1.0            0           -0.097093            2.0   \n",
       "...              ...          ...                 ...            ...   \n",
       "5007             1.0            0           -0.300581            1.0   \n",
       "5008             0.0            0           -0.038953            3.0   \n",
       "5009             1.0            1           -0.341279            1.0   \n",
       "5010             0.0            0           -0.308721            3.0   \n",
       "5011             1.0            1            0.000000            0.0   \n",
       "\n",
       "      NumberOfWindows Claim  Fenced=T  Fenced=F  Painted=T  Painted=F  \\\n",
       "0                   0   non       1.0       0.0        0.0        1.0   \n",
       "1                   0   non       1.0       0.0        0.0        1.0   \n",
       "2                   4   oui       0.0       1.0        1.0        0.0   \n",
       "3                   0   oui       1.0       0.0        0.0        1.0   \n",
       "4                   0   non       1.0       0.0        0.0        1.0   \n",
       "...               ...   ...       ...       ...        ...        ...   \n",
       "5007                0   oui       1.0       0.0        0.0        1.0   \n",
       "5008                4   non       0.0       1.0        1.0        0.0   \n",
       "5009                3   non       0.0       1.0        1.0        0.0   \n",
       "5010                4   non       0.0       1.0        1.0        0.0   \n",
       "5011                0   non       1.0       0.0        1.0        0.0   \n",
       "\n",
       "      Garden=T  Garden=F  Settle=R  Settle=U  \n",
       "0          1.0       0.0       0.0       1.0  \n",
       "1          1.0       0.0       0.0       1.0  \n",
       "2          0.0       1.0       1.0       0.0  \n",
       "3          1.0       0.0       0.0       1.0  \n",
       "4          1.0       0.0       0.0       1.0  \n",
       "...        ...       ...       ...       ...  \n",
       "5007       1.0       0.0       0.0       1.0  \n",
       "5008       0.0       1.0       1.0       0.0  \n",
       "5009       0.0       1.0       1.0       0.0  \n",
       "5010       0.0       1.0       1.0       0.0  \n",
       "5011       1.0       0.0       0.0       1.0  \n",
       "\n",
       "[5012 rows x 14 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encode_true_false(df_insurance, 'Settlement', ['R', 'U'], ['Settle=R', 'Settle=U'])\n",
    "df_insurance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Seperate data - X and Y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'> (5012, 13)\n",
      "<class 'numpy.ndarray'> (5012, 1)\n"
     ]
    }
   ],
   "source": [
    "# sepearate data\n",
    "def seperate_data(df:pd.DataFrame, class_name:str):\n",
    "  df_cp = df.copy(deep=True)\n",
    "  Y_train = df_cp.loc[:, [class_name]].values\n",
    "  df_cp.drop(axis=1, inplace=True, labels=[class_name])\n",
    "  X_train = df_cp.values\n",
    "  return X_train, Y_train\n",
    "\n",
    "X_train, Y_train = seperate_data(df_insurance, 'Claim')\n",
    "print(type(X_train), X_train.shape)\n",
    "print(type(Y_train), Y_train.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Apply ml algorithms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Claim', ylabel='Count'>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.histplot(df_insurance, x='Claim')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Decision Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree Classifier model accuracy (in %): 92.75738228252195\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cls = DecisionTreeClassifier()\n",
    "cls.fit(X_train, Y_train)\n",
    "\n",
    "Y_p_train = cls.predict(X_train)\n",
    "c_m_train = confusion_matrix(Y_train, Y_p_train)\n",
    "sns.heatmap(c_m_train, annot=True, cbar=False, xticklabels=cls.classes_, yticklabels=cls.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_train, Y_p_train)\n",
    "print(\"Decision Tree Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Decision tree classifier - testing with test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'> (2147, 13)\n",
      "<class 'numpy.ndarray'> (2147, 1)\n",
      "<class 'numpy.ndarray'> (2147, 13)\n",
      "<class 'numpy.ndarray'> (2147, 1)\n"
     ]
    }
   ],
   "source": [
    "from preprocessing import get_test_dataframe\n",
    "\n",
    "df_test = get_test_dataframe()\n",
    "X_test, Y_test = seperate_data(df_test, 'Claim')\n",
    "print(type(X_test), X_test.shape)\n",
    "print(type(Y_test), Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "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>Insured_Period</th>\n",
       "      <th>Residential</th>\n",
       "      <th>Building Dimension</th>\n",
       "      <th>Building_Type</th>\n",
       "      <th>NumberOfWindows</th>\n",
       "      <th>Claim</th>\n",
       "      <th>Fenced=T</th>\n",
       "      <th>Fenced=F</th>\n",
       "      <th>Painted=T</th>\n",
       "      <th>Painted=F</th>\n",
       "      <th>Garden=T</th>\n",
       "      <th>Garden=F</th>\n",
       "      <th>Settle=R</th>\n",
       "      <th>Settle=U</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.535354</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.203175</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.487157</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>oui</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.675325</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.082828</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2142</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.137374</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2143</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2144</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.213564</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>non</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2145</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.307071</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>oui</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2146</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.213564</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>non</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2147 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Insured_Period  Residential  Building Dimension  Building_Type  \\\n",
       "0                1.0            0            1.535354            3.0   \n",
       "1                1.0            0            0.203175            3.0   \n",
       "2                1.0            1            0.487157            1.0   \n",
       "3                1.0            0            0.675325            2.0   \n",
       "4                0.0            0            1.082828            3.0   \n",
       "...              ...          ...                 ...            ...   \n",
       "2142             0.0            1           -0.137374            0.0   \n",
       "2143             1.0            0            0.000000            2.0   \n",
       "2144             1.0            0           -0.213564            2.0   \n",
       "2145             1.0            1           -0.307071            2.0   \n",
       "2146             0.0            0           -0.213564            2.0   \n",
       "\n",
       "      NumberOfWindows Claim  Fenced=T  Fenced=F  Painted=T  Painted=F  \\\n",
       "0                   0   oui       1.0       0.0        1.0        0.0   \n",
       "1                   5   non       0.0       1.0        1.0        0.0   \n",
       "2                   6   oui       0.0       1.0        1.0        0.0   \n",
       "3                   0   oui       1.0       0.0        0.0        1.0   \n",
       "4                   9   non       0.0       1.0        1.0        0.0   \n",
       "...               ...   ...       ...       ...        ...        ...   \n",
       "2142                2   non       0.0       1.0        1.0        0.0   \n",
       "2143                0   non       1.0       0.0        1.0        0.0   \n",
       "2144                3   non       0.0       1.0        1.0        0.0   \n",
       "2145                0   oui       1.0       0.0        0.0        1.0   \n",
       "2146                0   non       1.0       0.0        0.0        1.0   \n",
       "\n",
       "      Garden=T  Garden=F  Settle=R  Settle=U  \n",
       "0          1.0       0.0       0.0       1.0  \n",
       "1          0.0       1.0       1.0       0.0  \n",
       "2          0.0       1.0       1.0       0.0  \n",
       "3          1.0       0.0       0.0       1.0  \n",
       "4          0.0       1.0       1.0       0.0  \n",
       "...        ...       ...       ...       ...  \n",
       "2142       0.0       1.0       1.0       0.0  \n",
       "2143       1.0       0.0       0.0       1.0  \n",
       "2144       0.0       1.0       1.0       0.0  \n",
       "2145       1.0       0.0       0.0       1.0  \n",
       "2146       1.0       0.0       0.0       1.0  \n",
       "\n",
       "[2147 rows x 14 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision Tree Classifier model accuracy (in %): 70.37727061015372\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Y_p_test = cls.predict(X_test)\n",
    "c_m_test = confusion_matrix(Y_test, Y_p_test)\n",
    "sns.heatmap(c_m_test, annot=True, cbar=False, xticklabels=cls.classes_, yticklabels=cls.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_test, Y_p_test)\n",
    "print(\"Decision Tree Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Logistic Regression "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression model accuracy (in %): 78.69114126097367\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = LogisticRegression(random_state=0)\n",
    "clf.fit(X_train, Y_train)\n",
    "Y_p_train2 = clf.predict(X_train)\n",
    "\n",
    "c_m_train2 = confusion_matrix(Y_train, Y_p_train2)\n",
    "sns.heatmap(c_m_train2, annot=True, cbar=False, xticklabels=clf.classes_, yticklabels=clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_train, Y_p_train2)\n",
    "print(\"Logistic Regression model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Logistic Regression - test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'> (2147, 13)\n",
      "<class 'numpy.ndarray'> (2147, 1)\n"
     ]
    }
   ],
   "source": [
    "from preprocessing import get_test_dataframe\n",
    "\n",
    "df_test = get_test_dataframe()\n",
    "X_test, Y_test = seperate_data(df_test, 'Claim')\n",
    "print(type(X_test), X_test.shape)\n",
    "print(type(Y_test), Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Logistic Regression model accuracy (in %): 77.17745691662785\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "Y_p_test = clf.predict(X_test)\n",
    "c_m_test = confusion_matrix(Y_test, Y_p_test)\n",
    "sns.heatmap(c_m_test, annot=True, cbar=False, xticklabels=clf.classes_, yticklabels=clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_test, Y_p_test)\n",
    "print(\"Logistic Regression model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. RandomForestClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Classifier model accuracy (in %): 92.75738228252195\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train, Y_train = seperate_data(df_insurance, 'Claim')\n",
    "rf_clf = RandomForestClassifier(n_estimators=100)\n",
    "rf_clf.fit(X_train, Y_train)\n",
    "Y_p_train3 = rf_clf.predict(X_train)\n",
    "\n",
    "c_m_train3 = confusion_matrix(Y_train, Y_p_train3)\n",
    "sns.heatmap(c_m_train3, annot=True, cbar=False, xticklabels=rf_clf.classes_, yticklabels=rf_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_train, Y_p_train3)\n",
    "print(\"Random Forest Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Random Forest Classifier - test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Classifier model accuracy (in %): 71.1690731252911\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from preprocessing import get_test_dataframe\n",
    "\n",
    "df_test = get_test_dataframe()\n",
    "X_test, Y_test = seperate_data(df_test, 'Claim')\n",
    "\n",
    "Y_p_test = rf_clf.predict(X_test)\n",
    "c_m_test3 = confusion_matrix(Y_test, Y_p_test)\n",
    "sns.heatmap(c_m_test3, annot=True, cbar=False, xticklabels=rf_clf.classes_, yticklabels=rf_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_test, Y_p_test)\n",
    "print(\"Random Forest Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. Svm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM Classifier model accuracy (in %): 78.63128491620111\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_train, Y_train = seperate_data(df_insurance, 'Claim')\n",
    "svm_clf = svm.SVC(kernel='rbf', C = 1.0)\n",
    "svm_clf.fit(X_train, Y_train)\n",
    "Y_p_train4 = svm_clf.predict(X_train)\n",
    "\n",
    "c_m_train3 = confusion_matrix(Y_train, Y_p_train4)\n",
    "sns.heatmap(c_m_train3, annot=True, cbar=False, xticklabels=svm_clf.classes_, yticklabels=svm_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_train, Y_p_train4)\n",
    "print(\"SVM Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. SVM classifier - test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM Classifier model accuracy (in %): 76.61853749417791\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from preprocessing import get_test_dataframe\n",
    "\n",
    "df_test = get_test_dataframe()\n",
    "X_test, Y_test = seperate_data(df_test, 'Claim')\n",
    "\n",
    "Y_p_test = svm_clf.predict(X_test)\n",
    "c_m_test4 = confusion_matrix(Y_test, Y_p_test)\n",
    "sns.heatmap(c_m_test4, annot=True, cbar=False, xticklabels=svm_clf.classes_, yticklabels=svm_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_test, Y_p_test)\n",
    "print(\"SVM Classifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. MLPClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SVM Classifier model accuracy (in %): 79.66879489225857\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mlp_clf = MLPClassifier(hidden_layer_sizes=(150, 100, 50), max_iter=300, activation='relu', solver='adam', random_state=1)\n",
    "mlp_clf.fit(X_train, Y_train)\n",
    "Y_p_train5 = mlp_clf.predict(X_train)\n",
    "\n",
    "c_m_train4 = confusion_matrix(Y_train, Y_p_train5)\n",
    "\n",
    "\n",
    "sns.heatmap(c_m_train4, annot=True, cbar=False, xticklabels=svm_clf.classes_, yticklabels=svm_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_train, Y_p_train5)\n",
    "print(\"MLPClassifier model accuracy (in %):\", acc*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. MLPClassifier - test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MLPClassifier model accuracy (in %): 77.08430367955287\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from preprocessing import get_test_dataframe\n",
    "\n",
    "df_test = get_test_dataframe()\n",
    "X_test, Y_test = seperate_data(df_test, 'Claim')\n",
    "\n",
    "Y_p_test = mlp_clf.predict(X_test)\n",
    "c_m_test4 = confusion_matrix(Y_test, Y_p_test)\n",
    "sns.heatmap(c_m_test4, annot=True, cbar=False, xticklabels=mlp_clf.classes_, yticklabels=mlp_clf.classes_, fmt='g')\n",
    "acc = accuracy_score(Y_test, Y_p_test)\n",
    "print(\"MLPClassifier model accuracy (in %):\", acc*100)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "machine-learning-2",
   "language": "python",
   "name": "machine-learning-2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}