got_prjt_v4.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import plotly.express as px\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sb\n",
    "\n",
    "from sklearn.linear_model import LinearRegression, LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsRegressor, KNeighborsClassifier\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score\n",
    "\n",
    "pd.set_option('display.max_columns', 50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "battles = pd.read_csv('battles.csv')\n",
    "char_deaths = pd.read_csv('character-deaths.csv')\n",
    "char_pred = pd.read_csv('character-predictions.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>name</th>\n",
       "      <th>year</th>\n",
       "      <th>battle_number</th>\n",
       "      <th>attacker_king</th>\n",
       "      <th>defender_king</th>\n",
       "      <th>attacker_1</th>\n",
       "      <th>attacker_2</th>\n",
       "      <th>attacker_3</th>\n",
       "      <th>attacker_4</th>\n",
       "      <th>defender_1</th>\n",
       "      <th>defender_2</th>\n",
       "      <th>defender_3</th>\n",
       "      <th>defender_4</th>\n",
       "      <th>attacker_outcome</th>\n",
       "      <th>battle_type</th>\n",
       "      <th>major_death</th>\n",
       "      <th>major_capture</th>\n",
       "      <th>attacker_size</th>\n",
       "      <th>defender_size</th>\n",
       "      <th>attacker_commander</th>\n",
       "      <th>defender_commander</th>\n",
       "      <th>summer</th>\n",
       "      <th>location</th>\n",
       "      <th>region</th>\n",
       "      <th>note</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Battle of the Golden Tooth</td>\n",
       "      <td>298</td>\n",
       "      <td>1</td>\n",
       "      <td>Joffrey/Tommen Baratheon</td>\n",
       "      <td>Robb Stark</td>\n",
       "      <td>Lannister</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Tully</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>win</td>\n",
       "      <td>pitched battle</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>4000.0</td>\n",
       "      <td>Jaime Lannister</td>\n",
       "      <td>Clement Piper, Vance</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Golden Tooth</td>\n",
       "      <td>The Westerlands</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Battle at the Mummer's Ford</td>\n",
       "      <td>298</td>\n",
       "      <td>2</td>\n",
       "      <td>Joffrey/Tommen Baratheon</td>\n",
       "      <td>Robb Stark</td>\n",
       "      <td>Lannister</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Baratheon</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>win</td>\n",
       "      <td>ambush</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>120.0</td>\n",
       "      <td>Gregor Clegane</td>\n",
       "      <td>Beric Dondarrion</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Mummer's Ford</td>\n",
       "      <td>The Riverlands</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Battle of Riverrun</td>\n",
       "      <td>298</td>\n",
       "      <td>3</td>\n",
       "      <td>Joffrey/Tommen Baratheon</td>\n",
       "      <td>Robb Stark</td>\n",
       "      <td>Lannister</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Tully</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>win</td>\n",
       "      <td>pitched battle</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>15000.0</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>Jaime Lannister, Andros Brax</td>\n",
       "      <td>Edmure Tully, Tytos Blackwood</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Riverrun</td>\n",
       "      <td>The Riverlands</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          name  year  battle_number             attacker_king  \\\n",
       "0   Battle of the Golden Tooth   298              1  Joffrey/Tommen Baratheon   \n",
       "1  Battle at the Mummer's Ford   298              2  Joffrey/Tommen Baratheon   \n",
       "2           Battle of Riverrun   298              3  Joffrey/Tommen Baratheon   \n",
       "\n",
       "  defender_king attacker_1 attacker_2 attacker_3 attacker_4 defender_1  \\\n",
       "0    Robb Stark  Lannister        NaN        NaN        NaN      Tully   \n",
       "1    Robb Stark  Lannister        NaN        NaN        NaN  Baratheon   \n",
       "2    Robb Stark  Lannister        NaN        NaN        NaN      Tully   \n",
       "\n",
       "  defender_2  defender_3  defender_4 attacker_outcome     battle_type  \\\n",
       "0        NaN         NaN         NaN              win  pitched battle   \n",
       "1        NaN         NaN         NaN              win          ambush   \n",
       "2        NaN         NaN         NaN              win  pitched battle   \n",
       "\n",
       "   major_death  major_capture  attacker_size  defender_size  \\\n",
       "0          1.0            0.0        15000.0         4000.0   \n",
       "1          1.0            0.0            NaN          120.0   \n",
       "2          0.0            1.0        15000.0        10000.0   \n",
       "\n",
       "             attacker_commander             defender_commander  summer  \\\n",
       "0               Jaime Lannister           Clement Piper, Vance     1.0   \n",
       "1                Gregor Clegane               Beric Dondarrion     1.0   \n",
       "2  Jaime Lannister, Andros Brax  Edmure Tully, Tytos Blackwood     1.0   \n",
       "\n",
       "        location           region note  \n",
       "0   Golden Tooth  The Westerlands  NaN  \n",
       "1  Mummer's Ford   The Riverlands  NaN  \n",
       "2       Riverrun   The Riverlands  NaN  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "battles.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>S.No</th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "      <th>alive</th>\n",
       "      <th>plod</th>\n",
       "      <th>name</th>\n",
       "      <th>title</th>\n",
       "      <th>male</th>\n",
       "      <th>culture</th>\n",
       "      <th>dateOfBirth</th>\n",
       "      <th>DateoFdeath</th>\n",
       "      <th>mother</th>\n",
       "      <th>father</th>\n",
       "      <th>heir</th>\n",
       "      <th>house</th>\n",
       "      <th>spouse</th>\n",
       "      <th>book1</th>\n",
       "      <th>book2</th>\n",
       "      <th>book3</th>\n",
       "      <th>book4</th>\n",
       "      <th>book5</th>\n",
       "      <th>isAliveMother</th>\n",
       "      <th>isAliveFather</th>\n",
       "      <th>isAliveHeir</th>\n",
       "      <th>isAliveSpouse</th>\n",
       "      <th>isMarried</th>\n",
       "      <th>isNoble</th>\n",
       "      <th>age</th>\n",
       "      <th>numDeadRelations</th>\n",
       "      <th>boolDeadRelations</th>\n",
       "      <th>isPopular</th>\n",
       "      <th>popularity</th>\n",
       "      <th>isAlive</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.946</td>\n",
       "      <td>Viserys II Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Rhaenyra Targaryen</td>\n",
       "      <td>Daemon Targaryen</td>\n",
       "      <td>Aegon IV Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.605351</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.387</td>\n",
       "      <td>0.613</td>\n",
       "      <td>Walder Frey</td>\n",
       "      <td>Lord of the Crossing</td>\n",
       "      <td>1</td>\n",
       "      <td>Rivermen</td>\n",
       "      <td>208.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Frey</td>\n",
       "      <td>Perra Royce</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.896321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.493</td>\n",
       "      <td>0.507</td>\n",
       "      <td>Addison Hill</td>\n",
       "      <td>Ser</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Swyft</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.267559</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   S.No  actual  pred  alive   plod                  name  \\\n",
       "0     1       0     0  0.054  0.946  Viserys II Targaryen   \n",
       "1     2       1     0  0.387  0.613           Walder Frey   \n",
       "2     3       1     0  0.493  0.507          Addison Hill   \n",
       "\n",
       "                  title  male   culture  dateOfBirth  DateoFdeath  \\\n",
       "0                   NaN     1       NaN          NaN          NaN   \n",
       "1  Lord of the Crossing     1  Rivermen        208.0          NaN   \n",
       "2                   Ser     1       NaN          NaN          NaN   \n",
       "\n",
       "               mother            father                heir        house  \\\n",
       "0  Rhaenyra Targaryen  Daemon Targaryen  Aegon IV Targaryen          NaN   \n",
       "1                 NaN               NaN                 NaN   House Frey   \n",
       "2                 NaN               NaN                 NaN  House Swyft   \n",
       "\n",
       "        spouse  book1  book2  book3  book4  book5  isAliveMother  \\\n",
       "0          NaN      0      0      0      0      0            1.0   \n",
       "1  Perra Royce      1      1      1      1      1            NaN   \n",
       "2          NaN      0      0      0      1      0            NaN   \n",
       "\n",
       "   isAliveFather  isAliveHeir  isAliveSpouse  isMarried  isNoble   age  \\\n",
       "0            0.0          0.0            NaN          0        0   NaN   \n",
       "1            NaN          NaN            1.0          1        1  97.0   \n",
       "2            NaN          NaN            NaN          0        1   NaN   \n",
       "\n",
       "   numDeadRelations  boolDeadRelations  isPopular  popularity  isAlive  \n",
       "0                11                  1          1    0.605351        0  \n",
       "1                 1                  1          1    0.896321        1  \n",
       "2                 0                  0          0    0.267559        1  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "char_pred.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Name</th>\n",
       "      <th>Allegiances</th>\n",
       "      <th>Death Year</th>\n",
       "      <th>Book of Death</th>\n",
       "      <th>Death Chapter</th>\n",
       "      <th>Book Intro Chapter</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Nobility</th>\n",
       "      <th>GoT</th>\n",
       "      <th>CoK</th>\n",
       "      <th>SoS</th>\n",
       "      <th>FfC</th>\n",
       "      <th>DwD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Addam Marbrand</td>\n",
       "      <td>Lannister</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>56.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Aegon Frey (Jinglebell)</td>\n",
       "      <td>None</td>\n",
       "      <td>299.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>49.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Aegon Targaryen</td>\n",
       "      <td>House Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      Name      Allegiances  Death Year  Book of Death  \\\n",
       "0           Addam Marbrand        Lannister         NaN            NaN   \n",
       "1  Aegon Frey (Jinglebell)             None       299.0            3.0   \n",
       "2          Aegon Targaryen  House Targaryen         NaN            NaN   \n",
       "\n",
       "   Death Chapter  Book Intro Chapter  Gender  Nobility  GoT  CoK  SoS  FfC  \\\n",
       "0            NaN                56.0       1         1    1    1    1    1   \n",
       "1           51.0                49.0       1         1    0    0    1    0   \n",
       "2            NaN                 5.0       1         1    0    0    0    0   \n",
       "\n",
       "   DwD  \n",
       "0    0  \n",
       "1    0  \n",
       "2    1  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "char_deaths.head(3)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Exploration"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since our task is to predict characters death. We will focus on Character dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>S.No</th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "      <th>alive</th>\n",
       "      <th>plod</th>\n",
       "      <th>name</th>\n",
       "      <th>title</th>\n",
       "      <th>male</th>\n",
       "      <th>culture</th>\n",
       "      <th>dateOfBirth</th>\n",
       "      <th>DateoFdeath</th>\n",
       "      <th>mother</th>\n",
       "      <th>father</th>\n",
       "      <th>heir</th>\n",
       "      <th>house</th>\n",
       "      <th>spouse</th>\n",
       "      <th>book1</th>\n",
       "      <th>book2</th>\n",
       "      <th>book3</th>\n",
       "      <th>book4</th>\n",
       "      <th>book5</th>\n",
       "      <th>isAliveMother</th>\n",
       "      <th>isAliveFather</th>\n",
       "      <th>isAliveHeir</th>\n",
       "      <th>isAliveSpouse</th>\n",
       "      <th>isMarried</th>\n",
       "      <th>isNoble</th>\n",
       "      <th>age</th>\n",
       "      <th>numDeadRelations</th>\n",
       "      <th>boolDeadRelations</th>\n",
       "      <th>isPopular</th>\n",
       "      <th>popularity</th>\n",
       "      <th>isAlive</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.946</td>\n",
       "      <td>Viserys II Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Rhaenyra Targaryen</td>\n",
       "      <td>Daemon Targaryen</td>\n",
       "      <td>Aegon IV Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.605351</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.387</td>\n",
       "      <td>0.613</td>\n",
       "      <td>Walder Frey</td>\n",
       "      <td>Lord of the Crossing</td>\n",
       "      <td>1</td>\n",
       "      <td>Rivermen</td>\n",
       "      <td>208.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Frey</td>\n",
       "      <td>Perra Royce</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.896321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.493</td>\n",
       "      <td>0.507</td>\n",
       "      <td>Addison Hill</td>\n",
       "      <td>Ser</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Swyft</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.267559</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   S.No  actual  pred  alive   plod                  name  \\\n",
       "0     1       0     0  0.054  0.946  Viserys II Targaryen   \n",
       "1     2       1     0  0.387  0.613           Walder Frey   \n",
       "2     3       1     0  0.493  0.507          Addison Hill   \n",
       "\n",
       "                  title  male   culture  dateOfBirth  DateoFdeath  \\\n",
       "0                   NaN     1       NaN          NaN          NaN   \n",
       "1  Lord of the Crossing     1  Rivermen        208.0          NaN   \n",
       "2                   Ser     1       NaN          NaN          NaN   \n",
       "\n",
       "               mother            father                heir        house  \\\n",
       "0  Rhaenyra Targaryen  Daemon Targaryen  Aegon IV Targaryen          NaN   \n",
       "1                 NaN               NaN                 NaN   House Frey   \n",
       "2                 NaN               NaN                 NaN  House Swyft   \n",
       "\n",
       "        spouse  book1  book2  book3  book4  book5  isAliveMother  \\\n",
       "0          NaN      0      0      0      0      0            1.0   \n",
       "1  Perra Royce      1      1      1      1      1            NaN   \n",
       "2          NaN      0      0      0      1      0            NaN   \n",
       "\n",
       "   isAliveFather  isAliveHeir  isAliveSpouse  isMarried  isNoble   age  \\\n",
       "0            0.0          0.0            NaN          0        0   NaN   \n",
       "1            NaN          NaN            1.0          1        1  97.0   \n",
       "2            NaN          NaN            NaN          0        1   NaN   \n",
       "\n",
       "   numDeadRelations  boolDeadRelations  isPopular  popularity  isAlive  \n",
       "0                11                  1          1    0.605351        0  \n",
       "1                 1                  1          1    0.896321        1  \n",
       "2                 0                  0          0    0.267559        1  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "char_pred.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count       433.000000\n",
       "mean      -1293.563510\n",
       "std       19564.340993\n",
       "min     -298001.000000\n",
       "25%          18.000000\n",
       "50%          27.000000\n",
       "75%          50.000000\n",
       "max         100.000000\n",
       "Name: age, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# viewing age properties\n",
    "char_pred['age'].describe()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are negative value for date of birth. Negative mean for age.\n",
    "Let's investigate that."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Research was made and according to [Fandom.com](https://gameofthrones.fandom.com/wiki/Doreah) Doreah is 25 years old.\n",
    "\n",
    "According to [Fanthom.com](https://gameofthrones.fandom.com/wiki/Rhaego) Rhaego was never born."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Engineering\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "char = char_pred.copy()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cleaning Feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Replacing the age\n",
    "char.loc[1684, 'age'] = 25.0\n",
    "char.loc[1868, 'age'] = 0.0"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Removing NaNs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title            1008\n",
       "culture          1269\n",
       "dateOfBirth      1513\n",
       "DateoFdeath      1502\n",
       "mother           1925\n",
       "father           1920\n",
       "heir             1923\n",
       "house             427\n",
       "spouse           1670\n",
       "isAliveMother    1925\n",
       "isAliveFather    1920\n",
       "isAliveHeir      1923\n",
       "isAliveSpouse    1670\n",
       "age              1513\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "miss = char.isna().sum()\n",
    "miss[miss > 0]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have 14 columns with NaN vales. We will replace that of age with the average age value.\n",
    "\n",
    "Replace date of birth and date of death with the median value.\n",
    "\n",
    "Replace other categorical variable with unknown\n",
    "\n",
    "Replace other numerical variable with -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fill missing age with average age\n",
    "char[\"age\"].fillna(round(char[\"age\"].mean()), inplace=True)\n",
    "char['dateOfBirth'].fillna(char['dateOfBirth'].median(), inplace=True)\n",
    "char['DateoFdeath'].fillna(char['DateoFdeath'].median(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat = ['title', 'culture', 'mother', 'father', 'heir', 'house', 'spouse']\n",
    "for col in cat:\n",
    "    char[col].fillna('unknown', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "char.fillna(-1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['unknown', 'Rivermen', 'Dornish', 'Valyrian', 'Ironborn',\n",
       "       'Free Folk', 'Northmen', 'Summer Isles', 'Braavosi', 'Dothraki',\n",
       "       'Ghiscari', 'Vale mountain clans', 'Reach', 'Tyroshi', 'Lhazarene',\n",
       "       'Free folk', 'Ironmen', 'Qartheen', 'Lysene', 'westermen',\n",
       "       'Westerman', 'Qarth', 'Lyseni', 'northmen', 'Qohor', 'Westeros',\n",
       "       'Norvoshi', 'First Men', 'Meereenese', 'Andal', 'Astapori',\n",
       "       'Westermen', 'ironborn', 'Ghiscaricari', 'Braavos', 'Stormlands',\n",
       "       'Valemen', 'Myrish', 'Lhazareen', 'Dornishmen', 'Sistermen',\n",
       "       'Northern mountain clans', 'Andals', 'Vale', 'Crannogmen',\n",
       "       'Wildling', 'Dorne', 'Pentoshi', 'free folk', 'Summer Islander',\n",
       "       'Westerlands', 'Summer Islands', 'Asshai', 'Riverlands', 'Naathi',\n",
       "       'Rhoynar', 'Meereen', 'Norvos', 'Stormlander', 'Wildlings',\n",
       "       'Astapor', 'Reachmen', \"Asshai'i\", 'Ibbenese', 'The Reach'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "char['culture'].unique()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some culture are spelt wrongly thus making duplicate. E.g. 'Westermen', 'Westerman' and 'westermen' are all the same."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From  Shail Daliwala's kernel on Kaggle\n",
    "cult = {\n",
    "    'Summer Islands': ['summer islands', 'summer islander', 'summer isles'],\n",
    "    'Ghiscari': ['ghiscari', 'ghiscaricari',  'ghis'],\n",
    "    'Asshai': [\"asshai'i\", 'asshai'],\n",
    "    'Lysene': ['lysene', 'lyseni'],\n",
    "    'Andal': ['andal', 'andals'],\n",
    "    'Braavosi': ['braavosi', 'braavos'],\n",
    "    'Dornish': ['dornishmen', 'dorne', 'dornish'],\n",
    "    'Myrish': ['myr', 'myrish', 'myrmen'],\n",
    "    'Westermen': ['westermen', 'westerman', 'westerlands'],\n",
    "    'Westerosi': ['westeros', 'westerosi'],\n",
    "    'Stormlander': ['stormlands', 'stormlander'],\n",
    "    'Norvoshi': ['norvos', 'norvoshi'],\n",
    "    'Northmen': ['the north', 'northmen'],\n",
    "    'Free Folk': ['wildling', 'first men', 'free folk'],\n",
    "    'Qartheen': ['qartheen', 'qarth'],\n",
    "    'Reach': ['the reach', 'reach', 'reachmen'],\n",
    "}\n",
    "\n",
    "def get_cult(value):\n",
    "    value = value.lower()\n",
    "    v = [k for (k, v) in cult.items() if value in v]\n",
    "    return v[0] if len(v) > 0 else value.title()\n",
    "\n",
    "# char.loc[:, \"culture\"] = [get_cult(x) for x in char.culture.fillna(\"\")]\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Link to other data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Names of all Attackers\n",
    "all_attackers = set(battles.attacker_1).union(set(battles.attacker_2))\\\n",
    ".union(set(battles.attacker_3)).union(set(battles.attacker_4))\n",
    "# Names of all Defenders\n",
    "all_defenders = set(battles.defender_1).union(set(battles.defender_2))\n",
    "\n",
    "# Unique attackers and defenders combined\n",
    "all_warlead = all_attackers.union(all_defenders)\n",
    "all_warlead = list(all_warlead)\n",
    "all_warlead.remove(np.nan)\n",
    "\n",
    "renam_warlead = []\n",
    "for x in all_warlead:\n",
    "    house = 'House ' + x\n",
    "    renam_warlead.append(house)\n",
    "    \n",
    "# All attacking commanders\n",
    "all = []\n",
    "for i in battles['attacker_commander'].dropna().values:\n",
    "    splitted = i.split(', ')\n",
    "    all +=splitted\n",
    "    \n",
    "all_att_commanders = set(all)\n",
    "\n",
    "# All defending commanders\n",
    "all = []\n",
    "for i in battles['defender_commander'].dropna().values:\n",
    "    splitted = i.split(', ')\n",
    "    all +=splitted\n",
    "    \n",
    "all_def_commanders = set(all)\n",
    "\n",
    "# All commanders \n",
    "all_commanders = all_att_commanders.union(all_def_commanders)\n",
    "# characters who swore allegiance in character death data\n",
    "char_with_allgnc = char_deaths[char_deaths['Allegiances'] != 'None']\n",
    "\n",
    "# Extracting the above character from all character data\n",
    "all_char_with_allgnc = char_pred[char_pred['name'].isin(char_with_allgnc['Name'])]\n",
    "\n",
    "# house of character with allegiance to another house (house in character dead allegiance column)\n",
    "hus_char_with_allegnc = all_char_with_allgnc['house']\n",
    "\n",
    "# Extract unique house of allegiance from charater death data.\n",
    "husOfAllgnc = list(set(char_deaths['Allegiances']))\n",
    "husOfAllgnc.remove('None')\n",
    "\n",
    "renam_husOfAllgnc = []\n",
    "for x in husOfAllgnc:\n",
    "    house = 'House ' + x\n",
    "    renam_husOfAllgnc.append(house)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# num_features = da.select_dtypes(['int', 'float'])\n",
    "# corr_mat = num_features.corr()\n",
    "# sb.heatmap(data=corr_mat)\n",
    "# plt.rcParams['figure.figsize'] = (12,10)\n",
    "# plt.show()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating Feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "char['isACommander'] = np.where(char['name'].isin(all_commanders), 1, 0)\n",
    "char['inHouseOfWarLead'] = np.where(char['house'].isin(renam_warlead), 1, 0)\n",
    "char['houseSworeAllegiance'] = np.where(char['house'].isin(hus_char_with_allegnc), 1, 0)\n",
    "char['inHouseOfAllegnc'] = np.where(char['house'].isin(renam_husOfAllgnc), 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "char = char.drop(['S.No', 'actual', 'pred', 'alive', 'plod', 'name', 'dateOfBirth',\n",
    "                  'DateoFdeath', 'house'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>title</th>\n",
       "      <th>male</th>\n",
       "      <th>culture</th>\n",
       "      <th>mother</th>\n",
       "      <th>father</th>\n",
       "      <th>heir</th>\n",
       "      <th>spouse</th>\n",
       "      <th>book1</th>\n",
       "      <th>book2</th>\n",
       "      <th>book3</th>\n",
       "      <th>book4</th>\n",
       "      <th>book5</th>\n",
       "      <th>isAliveMother</th>\n",
       "      <th>isAliveFather</th>\n",
       "      <th>isAliveHeir</th>\n",
       "      <th>isAliveSpouse</th>\n",
       "      <th>isMarried</th>\n",
       "      <th>isNoble</th>\n",
       "      <th>age</th>\n",
       "      <th>numDeadRelations</th>\n",
       "      <th>boolDeadRelations</th>\n",
       "      <th>isPopular</th>\n",
       "      <th>popularity</th>\n",
       "      <th>isAlive</th>\n",
       "      <th>isACommander</th>\n",
       "      <th>inHouseOfWarLead</th>\n",
       "      <th>houseSworeAllegiance</th>\n",
       "      <th>inHouseOfAllegnc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>Rhaenyra Targaryen</td>\n",
       "      <td>Daemon Targaryen</td>\n",
       "      <td>Aegon IV Targaryen</td>\n",
       "      <td>unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.605351</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Lord of the Crossing</td>\n",
       "      <td>1</td>\n",
       "      <td>Rivermen</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>Perra Royce</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.896321</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ser</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.267559</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Queen</td>\n",
       "      <td>0</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>Viserys I Targaryen</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.183946</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Greenstone</td>\n",
       "      <td>0</td>\n",
       "      <td>Dornish</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>Eldon Estermont</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.043478</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>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",
       "      <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>1941</th>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>Westeros</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.351171</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1942</th>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.096990</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1943</th>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.030100</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1944</th>\n",
       "      <td>unknown</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.130435</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1945</th>\n",
       "      <td>Casterly Rock</td>\n",
       "      <td>1</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>unknown</td>\n",
       "      <td>Jeyne Marbrand</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>47.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.210702</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1946 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     title  male   culture              mother  \\\n",
       "0                  unknown     1   unknown  Rhaenyra Targaryen   \n",
       "1     Lord of the Crossing     1  Rivermen             unknown   \n",
       "2                      Ser     1   unknown             unknown   \n",
       "3                    Queen     0   unknown             unknown   \n",
       "4               Greenstone     0   Dornish             unknown   \n",
       "...                    ...   ...       ...                 ...   \n",
       "1941               unknown     1  Westeros             unknown   \n",
       "1942               unknown     1   unknown             unknown   \n",
       "1943               unknown     1   unknown             unknown   \n",
       "1944               unknown     1   unknown             unknown   \n",
       "1945         Casterly Rock     1   unknown             unknown   \n",
       "\n",
       "                father                heir               spouse  book1  book2  \\\n",
       "0     Daemon Targaryen  Aegon IV Targaryen              unknown      0      0   \n",
       "1              unknown             unknown          Perra Royce      1      1   \n",
       "2              unknown             unknown              unknown      0      0   \n",
       "3              unknown             unknown  Viserys I Targaryen      0      0   \n",
       "4              unknown             unknown      Eldon Estermont      0      0   \n",
       "...                ...                 ...                  ...    ...    ...   \n",
       "1941           unknown             unknown              unknown      1      1   \n",
       "1942           unknown             unknown              unknown      0      1   \n",
       "1943           unknown             unknown              unknown      1      1   \n",
       "1944           unknown             unknown              unknown      0      0   \n",
       "1945           unknown             unknown       Jeyne Marbrand      0      0   \n",
       "\n",
       "      book3  book4  book5  isAliveMother  isAliveFather  isAliveHeir  \\\n",
       "0         0      0      0            1.0            0.0          0.0   \n",
       "1         1      1      1           -1.0           -1.0         -1.0   \n",
       "2         0      1      0           -1.0           -1.0         -1.0   \n",
       "3         0      0      0           -1.0           -1.0         -1.0   \n",
       "4         0      1      0           -1.0           -1.0         -1.0   \n",
       "...     ...    ...    ...            ...            ...          ...   \n",
       "1941      1      1      1           -1.0           -1.0         -1.0   \n",
       "1942      0      1      1           -1.0           -1.0         -1.0   \n",
       "1943      1      1      1           -1.0           -1.0         -1.0   \n",
       "1944      1      1      1           -1.0           -1.0         -1.0   \n",
       "1945      1      1      1           -1.0           -1.0         -1.0   \n",
       "\n",
       "      isAliveSpouse  isMarried  isNoble   age  numDeadRelations  \\\n",
       "0              -1.0          0        0  37.0                11   \n",
       "1               1.0          1        1  97.0                 1   \n",
       "2              -1.0          0        1  37.0                 0   \n",
       "3               0.0          1        1  23.0                 0   \n",
       "4               1.0          1        1  29.0                 0   \n",
       "...             ...        ...      ...   ...               ...   \n",
       "1941           -1.0          0        0  37.0                 0   \n",
       "1942           -1.0          0        0  37.0                 0   \n",
       "1943           -1.0          0        0  37.0                 0   \n",
       "1944           -1.0          0        0  37.0                 0   \n",
       "1945            1.0          1        1  47.0                 4   \n",
       "\n",
       "      boolDeadRelations  isPopular  popularity  isAlive  isACommander  \\\n",
       "0                     1          1    0.605351        0             0   \n",
       "1                     1          1    0.896321        1             1   \n",
       "2                     0          0    0.267559        1             0   \n",
       "3                     0          0    0.183946        0             0   \n",
       "4                     0          0    0.043478        1             0   \n",
       "...                 ...        ...         ...      ...           ...   \n",
       "1941                  0          1    0.351171        0             0   \n",
       "1942                  0          0    0.096990        0             0   \n",
       "1943                  0          0    0.030100        1             0   \n",
       "1944                  0          0    0.130435        1             0   \n",
       "1945                  1          0    0.210702        0             0   \n",
       "\n",
       "      inHouseOfWarLead  houseSworeAllegiance  inHouseOfAllegnc  \n",
       "0                    0                     0                 0  \n",
       "1                    1                     1                 0  \n",
       "2                    0                     1                 0  \n",
       "3                    0                     1                 1  \n",
       "4                    0                     1                 0  \n",
       "...                ...                   ...               ...  \n",
       "1941                 1                     1                 1  \n",
       "1942                 1                     1                 0  \n",
       "1943                 0                     0                 0  \n",
       "1944                 0                     0                 0  \n",
       "1945                 1                     1                 1  \n",
       "\n",
       "[1946 rows x 28 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "char"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "new = char.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "book3                   0.006693\n",
       "inHouseOfWarLead        0.029822\n",
       "isAliveSpouse           0.031536\n",
       "book5                   0.032846\n",
       "isNoble                 0.042211\n",
       "isMarried               0.050037\n",
       "houseSworeAllegiance    0.052159\n",
       "isACommander            0.057968\n",
       "book2                   0.067200\n",
       "inHouseOfAllegnc        0.113380\n",
       "isAliveHeir             0.116365\n",
       "isAliveFather           0.123615\n",
       "age                     0.136891\n",
       "isAliveMother           0.140661\n",
       "male                    0.146982\n",
       "book1                   0.147401\n",
       "isPopular               0.153860\n",
       "popularity              0.183223\n",
       "numDeadRelations        0.192444\n",
       "boolDeadRelations       0.193736\n",
       "book4                   0.268975\n",
       "isAlive                 1.000000\n",
       "Name: isAlive, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Heatmap of numerical feature in the training dataset\n",
    "num_features = new.select_dtypes(['int', 'float'])\n",
    "corr_mat = num_features.corr()['isAlive'].abs().sort_values()\n",
    "corr_mat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAcQAAAFjCAYAAACjRntJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAACCU0lEQVR4nO2dd1xUx9fGn11YmgjYC9JUxEqRYkHAKNbYElQEG2pMYjSWqDGJRo0mRo3GFiuIXVEs0RTFQhVRelHAqGDB3sVgocz7B+/eHyvbBoEFcr75nE/ce5+ZO/fusmennSMCwEAQBEEQ/3HEmm4AQRAEQVQFyCESBEEQBMghEgRBEAQAcogEQRAEAYAcIkEQBEEAIIdIEARBEADIIRIEQRDVjK1bt+L+/ftIS0tTqFmzZg2uXLmClJQUODg4qFUvOUSCIAiiWrF9+3b07dtX4fl+/frB2toa1tbW+PTTT7Fx40a16iWHSBAEQVQroqKi8OTJE4XnBw8ejJ07dwIALly4ABMTEzRu3FhlveQQCYIgiBqFqakpbt26JbzOycmBqampynLaFdkoomLx8PDArFmzMHDgwDKVf/vwGpd+rONMLv1rFHLpXxa95dIDwK03in8lyqONvupfiSXRgohLr82pF3HqU1/f5dIDwJuifC794FqtuPTpRc+59GLOe+Z9D0zFBlx6gP992/UogUvvU78jl/5ywTMuva6Y76v8SUEelx4ALtyN4C5TEp7vm8nfLcWnn34qvN6yZQv8/f3VLi8SlX4/GVMdpZQcIkEQBFHxFKr/w8zf35/LAb5LTk4OzMzMhNfNmjXDnTt3VJajIVMNY2FhgYyMDPj7+yMtLQ27d+9Gz549cfbsWfzzzz9wdnaGs7MzoqOjkZiYiOjoaLRqVfoXvIGBAbZu3YrY2FgkJiZi0KBBGrgbgiAIBRQVqW/vybFjxzBmzBgAQKdOnfD8+XPcu3dPZTnqIVYBWrZsiWHDhuHTTz9FXFwcfH190a1bNwwaNAjfffcdxowZA3d3dxQWFqJnz55YsmQJhg4dKlPH3LlzERoaigkTJsDY2BixsbE4ffo08vL4h0YIgiDKG8be39FJ2bt3L7p374769evj1q1bWLBgASQSCQBg8+bN+Pvvv9G/f39cvXoVeXl5GDdunFr1kkOsAmRnZ+PixYsAgEuXLuHMmTMAgLS0NFhaWsLY2Bg7duyAtbU1GGPCG1+S3r17Y9CgQZg1axYAQE9PD+bm5sjMzKy8GyEIglBEOfT8pPj6+qrUTJkyhbtecohVgDdv3gj/LioqEl4XFRVBW1sbixcvRlhYGD7++GNYWFggPDy8VB0ikQheXl74559/lF5r4sSJwmS1SLc22Jvc8rsRgiAIRZRjD7GioDnEaoCxsTFu374NAPDz85OrCQkJwZdffim8tre3l6vz9/cX5iXJGRIEUWkUFapvGoIcYjVg+fLl+Pnnn3H27FloaWnJ1SxevBgSiQSpqalIS0vD4sWLK7mVBEEQSigsUN80hAiA6s0ZRI2E9iGqhvYhqob2IaqG9iECr6/GqK3Va9nlva5VVmgOkSAIgqh4ynFRTUVBDvE/DG+Pb0fCSi59gu0sLn2clhGXHgDO6etx6XsUGnLp1xXw9aJ5e3yMc4DGQa8plx4ACjmv8RB8PfXcwjeqRSXoqc3XS69XxNmjLMMUlGk+35f1Ds5nalOkw6Wvq12fS2/I+Ga/3krqcOkB4AJ3iXeoBotqyCESBEEQFY8GF8uoCzlEgiAIouLR4GIZdSGHSBAEQVQ8NGRKEARBEKBFNQRBEAQBAIzRHCJBEARB0JApQRAEQQCgIVOCIAiCAMCVIFhTkEMkCIIgKh4aMiWkWFhY4M8//0SHDh3eq57s7Gw4OTnh8ePHwrFmzZph586daNy4MYqKirBlyxasXbtWZV28sUZ5I884pq7g0tfpNplLDwD3XvNF3HjOGc7+VSF/fNWKpJFIt8KvkVr4lEt/49UDLn1G7dpc+rpafFFeJJzRggCgCKVzjCqjqUE9Lv1jMV9km1b5fB9U/SK++sN0NdBboyFTojIoKCjAzJkzkZSUBENDQyQkJODUqVPIyMjQdNMIgiCKqQY9REr/VIloa2tj+/btSElJQXBwMPT19dGjRw8kJiYiNTUVW7duhY5O8a9hRcel6Onp4fjx4/jkk09w7949JCUlAQBevnyJjIwMmJqaVvr9EQRBKKSoSH3TEOQQK5HWrVtjy5YtsLOzw4sXL/DVV19h+/bt8Pb2hq2tLbS1tTFp0iTo6urKPS7F0NAQf/zxB/bu3YuAgACZa1hYWMDBwQEXLrx3KF6CIIjygxwiUZKbN2/i3LlzAIDdu3ejZ8+eyM7OxpUrVwAAO3bsgLu7O2xsbOQel3L06FFs27YNu3btkqm/Vq1aOHToEKZPn47c3Fy5bZg4cSLi4uIQFxeHXj59KuI2CYIgSsEK89U2TUEOsRJhTL2Jb5FI+aKA6Oho9OvXT+aYtrY2Dh06hD179uDIkSMKy/r7+8PZ2RnOzs44tS9ErfYQBEG8N6xIfdMQ5BArEQsLC3Tu3BkA4OPjg9OnT8PS0hItWrQAAIwePRoRERHIzMyUe1zK/Pnz8fjxY2zYsEE4tnXrVmRkZGDVqlWVeEcEQRBqQkOmREnS09MxduxYpKSkoG7duli1ahXGjRuH4OBgpKamoqioCJs2bcKbN2/kHi/J9OnToaenh2XLlsHV1RVjxoxBjx49kJSUhKSkpFI9SIIgCI1SDXqItO2ikrhx4wbatWtX6nhoaCg6duyo9nErKyvh3+PHjxf+rWqYlSAIQqPQPkSiKvOyiG/TeZyWEZeed6N987PrufQA0Kf9HC79dS2+je1vi/gm+As5f92KOTeRF4FvA3ZlwHvPrxlfotgiEd+m+Tzwf/G+FPNdo5EO39/Cc84gGKe0+f42OzIDLn3c23tc+nKBEgQTBEEQBKiHSBAEQRAAqkWkGnKIBEEQRMVTDXqItMqUIAiCqHjKeZVpnz59kJmZiStXrmDOnNJrCYyMjHDs2DEkJyfj4sWL8PPzU1knOUSCIAii4inHfYhisRjr169Hv3790LZtW/j4+KBNmzYymsmTJyM9PR329vbo3r07Vq5cCYlE+eIpcogEQRBExVNYoL6pwMXFBVevXkV2djby8/MRFBSEwYMHy2gYY6j9/6nGDA0N8eTJExQUKK+bHCJBEARR8XD0EEvGXI6Li8PEiRNlqjI1NcWtW7eE1zk5OaUy/Pz2229o06YN7ty5g7S0NEybNk1l+ExaVEMQBEFUPGrGcgaKYy77+/srPC8vEMm7zq5Pnz5ITk5Gjx490KJFC5w6dQp2dnYKEx8A1EMkCIIgKoNynEPMycmBmZmZ8LpZs2a4c+eOjGbcuHE4fPgwAODatWvIzs5G69atldZLPcT/MLfePOHSn9PX49Lfe12HS88bdQYAulxcxqXPbTeXS/+GMxVNRUeqKShDpJp8zjL1xHzvM+89PyrM49LnlyHyDC8iMd/7MBqNufSReMWlf1DEp3+qpc+lf1ukgagx5bjtIi4uDtbW1rC0tMTt27cxYsQI+Pr6ymhu3ryJnj174uzZs2jYsCFsbGyQlZWltF7qIVYSFhYWSEtLe+96srOzUa9ePbnnxGIxEhMT8ccff7z3dQiCIMqVctx2UVhYiClTpiAkJAQZGRk4cOAA0tPT8dlnn+Gzzz4DACxevBhdu3ZFamoqzpw5gzlz5uDx48dK66UeYg1i2rRpyMjIgJERX5xFgiCICqeQL56rKo4fP47jx4/LHNu8ebPw77t376JPH74k6NRDrES0tbWxfft2pKSkIDg4GPr6+ujRowcSExORmpqKrVu3QkdHBwAUHpeip6eH48eP45NPPgFQvOrqww8/REBAQKXfF0EQhEooHyJRktatW2PLli2ws7PDixcv8NVXX2H79u3w9vaGra0ttLW1MWnSJOjq6so9LsXQ0BB//PEH9u7dKzjA1atX4+uvv0ZRNQiPRBDEfxByiERJbt68iXPnzgEAdu/ejZ49eyI7OxtXrlwBAOzYsQPu7u6wsbGRe1zK0aNHsW3bNuzatQsA8OGHH+LBgwdITExU2YaS+3uGjR5SzndIEAShgGqQIJgcYiWialOoFFXJfqOjo9GvXz/htaurKwYNGoTs7GwEBQWhR48egrN8F39/fzg7O8PZ2RnBu35Xu+0EQRDvAytiapumIIdYiVhYWKBz584AAB8fH5w+fRqWlpZo0aIFAGD06NGIiIhAZmam3ONS5s+fj8ePH2PDhg0AgO+++w5mZmawsrLCiBEjEBoaitGjR1fy3REEQSihHEO3VRTkECuR9PR0jB07FikpKahbty5WrVqFcePGITg4GKmpqSgqKsKmTZvw5s0bucdLMn36dOjp6WHZMr59eARBEBqhiKlvGoK2XVQSN27cQLt27UodDw0NRceOHdU+bmVlJfx7/Pjxpc5HRETI9CYJgiCqBNVgwR85xP8wbfT5om30KDTk0j/nHH+4rqXLVwD8kWd6X/qJS1+7zTAuPW/UFhFnpJqX4N/LVcgZqUZPpMWlNzdoyKVvJeGLYNQYfJ+LZ+AfcrNgOqpFJTgoUr7B+11aw5hL/0LE99VswPg+R5Y6dbn0AJDKXeIdyCESBEEQBLiCe2sKcogEQRBExUM9RIIgCIJAuYduqwjIIRIEQRAVjwZXj6oLOUSCIAiiwmE0ZEoQBEEQoB4iQRAEQQDQaIxSdSGHSBAEQVQ8BbSohiAIgiBoyFRTREdHw9XVVanG3t4eSUlJ6NOnD06ePCkcz83NRe3atdGkSROsXbsWw4bxRSoBAA8PD4SHh2PChAkIDAyUud6sWbOwcuVKhWXHjh2LkydP4u7duwCA7OxsODk54fFjvsgY6qDFGSVlXcE1Lv2rwrdc+rdF+Vx6AHhTyFeGN/JMekYwl74gMohLX3TxEpfebeMNLj2gfpYVKc/yX3LpzfTqc+k/e8vXnhucUVialmFoTiLO49Ivz73Kpdeq04pLf/PtEy79uVdXuPSN9fkj1bw31WDItEYG91blDIHibBNRUVHw8fGRe/7u3btlcoZSUlNT4e3tLbweMWIEkpOTVZbz8/ND06ZNy3zdkmhp8YXgIgiCqDCqQXDvGukQc3NzAQCNGzdGREQEkpKSkJaWhm7dugmaoUOHws/PD71794aubulYiRYWFkhLSwMAnD9/Hm3bthXOhYWFoWPHjjAwMMDWrVsRGxuLxMREDBo0SNDcvHkTenp6aNiwOM5j3759cfz4ceG8nZ0dYmJikJKSgsOHD8PExAReXl5wcnLCnj17kJSUBD09PQDAl19+iYSEBKSmpsLGxgYAFF577NixOHDgAI4dOybT8yUIgtAkrKhIbdMUNdIhSvH19UVISAgcHBxgZ2cn9NBcXV2RnZ2NrKwshIeHo3///krrCQoKwvDhwwEUO9mmTZsiMTERc+fORWhoKFxcXPDBBx/gl19+gYGBgVDu4MGDGDZsGLp27YrExES8efNGOLdz507MmTMHdnZ2SEtLw4IFC3Do0CHEx8dj5MiRcHBwwOvXrwEAjx49gqOjIzZu3IhZs2YBgNJrd+nSBWPHjkXPnj3L7VkSBEG8F9RD1CxxcXEYN24cFixYgA4dOuDly+K5ER8fHwQFFc/1BAUFKRw2lXLgwAFh+HT48OEIDi6eV+rduze++eYbJCUlITw8HHp6ejA3Ny9VzsfHB/v27ROOGxkZwcTEBJGRkQCAHTt2wN3dXeH1Dx8+DABISEiApaWlymufOnUKT58+Vfs5EQRBVDiFheqbhqjRDjEqKgru7u64ffs2du3ahdGjR0MsFsPLywvz589HdnY21q1bh379+sHQUHFqozt37uDx48fo0KEDvL29BWcqEong5eUFBwcHODg4wMLCApmZmUK5+/fvIz8/H7169cKZM2fKfB/SnmVhYSG0tbVVXvvff/9VWNfEiRMRFxeHuLg4ePr0LnObCIIguKAeomYxNzfHgwcPEBAQgK1bt6Jjx47w9PRESkoKzM3NYWVlBUtLSxw6dAhDhgxRWldQUBC+/vprGBsb4+LFiwCAkJAQfPnll4LG3t6+VLn58+djzpw5KCoxLv7ixQs8ffpUmNMcPXq0kNRXuspVFepcWx7+/v5wdnaGs7MzTu+jOUaCICoHVsTUNk1Rox1i9+7dkZycjMTERHh5eWHNmjXw8fHBkSNHZHSHDh2Cr6+v0roOHjyIESNG4MCBA8KxxYsXQyKRIDU1FWlpaVi8eHGpcjExMTh69Gip42PHjsUvv/yClJQU2NvbY9GiRQCA7du3Y9OmTTKLauShzrUJgiCqDNWghygCONNpEzWGoeaDVItKcPnNAy59ldyHKDFQLSoB7UNUDe8+xOWML3v8DabPpW/K3qgWvYNEzLeysX9uGpe+awXvQ7z/im/NQFn2Iabej+EuU5IXk/uprTVaf1y1qAKokRvzCYIgiCoGRaohqjLanJFqRJx6XgrLEMmCtwyvnrfHp+0+gq/+l/5cesauc+kBoIhzEKgs7wMP94sUTwXI44U23+fOMF/CpQeAWoUFXHodLb6vTt73oIizV8/7njENDAyywvL9XPXp0wdr1qyBlpYWAgICsGzZslIaDw8PrF69GhKJBI8ePUL37t2V1kkOkSAIgqh4yrGHKBaLsX79evTq1Qs5OTmIi4vDsWPHkJGRIWiMjY2xYcMG9O3bF7du3UKDBg1U11tuLSQIgiAIRZTjohoXFxdcvXoV2dnZyM/PR1BQEAYPHiyj8fX1xeHDh3Hr1i0AwMOHD1XWSw6RIAiCqHDKc9uFqamp4OgAICcnB6ampjKaVq1aoU6dOggLC0N8fDxGjx6tsl4aMiUIgiAqHo4h04kTJ+LTTz8VXm/ZsgX+/v+bbxeJSs8rv7uaWltbG46OjujZsyf09fURExOD8+fP48oVxZlByCESBEEQFQ4rUN8h+u/0l3GA75KTkwMzMzPhdbNmzXDnzp1SmkePHiEvLw95eXmIjIyEnZ2dUodIQ6YEQRBExVOOc4hxcXGwtraGpaUlJBIJRowYgWPHjslojh49Cjc3N2hpaUFfXx+dOnWSWXQjD+ohEgRBEBVPOe66KCwsxJQpUxASEgItLS0EBgYiPT0dn332GQBg8+bNyMzMxIkTJ5CamoqioiIEBATg0iXlgTDIIRIEQRAVTnnHKD1+/LhMjlmg2BGWZMWKFVixYoXadZJDJAiCICoezeX9VRtyiCWIjo6Gq6urUo29vT2SkpLQp08fmYz00iwVTZo0wdq1a4X8iTx4eHhg1qxZGDhwoHBs27Zt+PPPP3Ho0CGF5X744QdERkZyp5jijTxT0dEtxGWIhMNbhveeeWON8kae0e4/kUuPryP59CjDM5Kzgq880eeMqqLN+JY6SMrwOdUS8ZXh/RzxvgdizveA9z2r6KhT8tBkFgt1IYdYAlXOEChOLhwVFQUfHx8Zhyjl7t27ZXKG78OCBQvkHheLxTJppwiCIDQF44uOpxFolWkJcnNzAQCNGzdGREQEkpKSkJaWJuQtBIChQ4fCz88PvXv3hq6ubqk6LCwskJZWHAn//PnzaNu2rXAuLCwMHTt2hIGBAbZu3YrY2FgkJiZi0CD1sk507NgR4eHhiI+Px4kTJ9C4cWMAxb1ILy8vAEB2dja+//57REVFVbpjJgiCUEgRh2kIcohy8PX1RUhICBwcHGBnZ4fk5GQAxT3I7OxsZGVlITw8HP3791daT1BQEIYPHw6g2Mk2bdoUiYmJmDt3LkJDQ+Hi4oIPPvgAv/zyCwwMitMSubm5ISkpSTCps9TW1sa6deswdOhQODk5ITAwED/99JPc675+/Rpubm7Yv39/OT0RgiCI94MVqW+agoZM5RAXF4fAwEBIJBL8/vvvSElJAVA8XBoUVJz9ICgoCKNHjy6VbLgkBw4cwKlTp7Bw4UIMHz4cwcHFufV69+6NQYMGYdasWQAAPT09mJubAwCioqJKzSECgI2NDdq3b49Tp04BALS0tHD37l251yVHSBBElaMazN6QQ5RDVFQU3N3d8eGHH2LXrl345ZdfsGfPHnh5eWHQoEGYO3cuRCIR6tWrB0NDQ7x8KT+h6p07d/D48WN06NAB3t7ewh4ZkUgELy8v/PPPPzL6Ro0aKWyTSCTCpUuX0LVrV5Xt//fffxWeKxkSKflgLEL3nVJZH0EQxPuiyZ6futCQqRzMzc3x4MEDBAQEYOvWrejYsSM8PT2RkpICc3NzWFlZwdLSEocOHcKQIUOU1hUUFISvv/4axsbGuHjxIgAgJCQEX375paCxt7dX2abLly+jQYMG6Ny5M4DiIdSS85Pq4u/vD2dnZzg7O5MzJAii0qgOQ6bkEOXQvXt3JCcnIzExEV5eXlizZg18fHxKDY8eOnQIvr6+Sus6ePAgRowYgQMHDgjHFi9eDIlEgtTUVKSlpWHx4sUq25Sfn4+hQ4di2bJlSE5ORnJyslq9RYIgiKoAKxSpbZpCBGggdTJRJfAxH8Klv/jmHpf+VeFbLv0bTn1xmXwuvaHEgEufOrUNl15sac6l592H6NJedQqb9+VJfi6X3kJPdeLVkswr4NPf1uab2TEt4F/fry/iKzP8dTqX3tHIikt/8+0TLv3dV3z6pvr1uPQAkHo/hrtMSe66eaitbRIV8V7XKis0h0gQBEFUOKxIcz0/dSGH+B8m9bX8VaqKcNBryqVvJCq9T1MZRWUYrCjgLPMShVx6t403uPSMXefS80aeib24i69+AOwVX4/vO7elXPpn4Ouln5BwyWHI+R6nctYPADaFelx6UxFfD6uBmK/+VnoWXHp9Pb4eaC7n3wEApOL9eojVYVENOUSCIAiiwmGMeogEQRAEgaICcogEQRAEAVYNlm+SQyQIgiAqHFpUQxAEQRAgh0gQBEEQAGjIlCAIgiAAUA+RIAiCIAAARRoMyaYu5BAJgiCICqeI9iHWHKKjo+Hq6qrwfHZ2Nm7dugV3d3fhWFJSErS1tdGhQ4cKvfa7LFiwAC9fvsTKlSuV6t4U8UUYKayCYW/zOdvEew+Mc+KDN9qOGHxfErxRZwBApF+bS88b8V/EeQ8mTItLr8dZf56I/4tXlzOKylvGF/u0gPNzVCCq2L813vesPKgOG/Mp24WaqOOQateujWbNmgEAWrduzX0NsVgs9zWPMyQIgqiKsCKR2qYpyCGqSW5u8S/zxo0bIyIiAklJSUhLS0O3bt0EzYEDB+Dt7Q0A8PHxwb59+4RzFhYWiIyMREJCAhISEtClSxcAgIeHB0JDQ7Fnzx6kpaWVel3y2gAwa9YsxMbGIiUlBQsXLhSOf/fdd8jMzMSpU6dgY2NTYc+BIAiiLDCmvmkKGjLlxNfXFyEhIViyZAnEYjEMDP6XTujgwYPYvn07Vq5ciYEDB2LkyJEYPbo4Xc+DBw/Qq1cvvHnzBi1btsS+ffvg7OwMAHBxcUH79u1x/fp1eHh4yLwuSa9evWBtbQ0XFxeIRCIcO3YMbm5u+PfffzFixAg4ODhAW1sbiYmJSEhIqLRnQhAEoQpaZVoDiYuLQ2BgICQSCX7//XekpKQI5548eYKnT5/C29sbGRkZyMvLE85JJBL89ttvsLe3R2FhIVq1aiWci42NlXF+776W0rt3b/Tu3RtJSUkAAENDQ1hbW6N27do4cuQIXr16BQA4duxYOd81QRDE+1FYVPUHJKt+C6sYUVFRcHd3x+3bt7Fr1y6hByhl//79WL9+vcxwKQDMmDED9+/fh52dHZycnKCjoyOc+/fff2W0776WIhKJ8PPPP8PBwQEODg6wtrZGYGAgAPUXf0ycOBFxcXGIi4uD9+iP1CpDEATxvlSHIVNyiJyYm5vjwYMHCAgIwNatW9GxY0eZ80eOHMHy5csREhIic9zY2Bh3794FYwyjR4+GNmcWcAAICQnB+PHjUatWLQBA06ZN0aBBA0RGRuKjjz6Cnp4eDA0NMXDgQIV1+Pv7w9nZGc7Ozti/6wh3GwiCIMpCEROpberQp08fZGZm4sqVK5gzZ45CnZOTEwoKCuDl5aWyThoy5aR79+6YPXs28vPz8fLlS4wZM0bm/MuXL7F8+fJS5TZs2IBDhw5h2LBhCAsLw8uXL7mvferUKbRp0wYxMTHCtUaNGoWkpCTs378fycnJuHHjBqKiosp2cwRBEBVEeW67EIvFWL9+PXr16oWcnBzExcXh2LFjyMjIKKVbtmxZqQ6KIkRAFdxcRlQKLet3VC0qgbOBOZe+qYgvSzjvHj4AeAW+DWR5nJnCU1/f49JX9D7EC3EbuPQA/z7EuU5zufRPwbcnz4zpqBaVgHcf4gMRfzb4dvl8eyNXFV7j0tvpNuHS1xFJuPS1wdf+l5x/NwCw9sZ+7jIlSWg2SG2tY47ydRCdO3fGwoUL0bdvXwDAN998AwBYunSpjG7atGnIz8+Hs7Mz/vzzTxw6dEhpvdRD/A8zuFYr1aISPMRbLn1q4VMufVmoJ+Zzunoivi+OZ/l8PflCxvdFI+LcRP6d21LVonfgnRf5Kf4nLv18p3lc+kzkqRaVQIvTIepxOgcASJPw/ZDRKuR7qg1EfD8CjuRmqBaVwMXQikv/tOg1l748KM9FNaamprh165bwOicnB506dZLRNG3aFB999BF69OghrOhXBc0hEgRBEBUOzxxiycV/cXFxmDhxokxd8n5IvruwcPXq1ZgzZw6KitT/kUo9RIIgCKLC4emD+/v7w9/fX+H5nJwcmJmZCa+bNWuGO3fuyGicnJwQFBQEAKhfvz769++PgoICHD16VGG95BAJgiCICqc8g3vHxcXB2toalpaWuH37NkaMGAFfX18ZTfPmzYV/b9u2DX/++adSZwiQQyQIgiAqgfJcZVpYWIgpU6YgJCQEWlpaCAwMRHp6Oj777DMAwObNm8tULzlEgiAIosLhX9eqnOPHj+P48eMyxxQ5wnHjxqlVJzlEgiAIosIprAbpn8ghEgRBEBVOkQZyMPJCDpEgCIKocBg5RIIgCIIo/znEioAc4n+Y9KLnXPrcwjdc+huvHnDpeaO8lKWMuUFDLr2ZXn0ufUXzDPncZUScv8x5I88siv+RS9/b/jMu/asivns21NLl0peFW3kPufTxWnwRlXjJeHOfS2+iZaBaVM5QD5EgCIIgAM6It5qBHCJBEARR4VSHHiLFMq3CHDlyBPHx8bh48aIQy2/8+PG4fPkywsLCsGXLFqxbtw5AcWiigwcPIjY2FrGxsejatasmm04QBCFDkUh90xTUQ6zCjB8/Hk+fPoWenh7i4uLw119/4fvvv0fHjh2Rm5uL0NBQpKSkAADWrFmDVatWITo6GmZmZggJCUHbtm01fAcEQRDF0LYL4r2YOnUqPvroIwCAmZkZRo8ejYiICDx9WpxWKTg4GK1aFadw8vT0lHGARkZGMDQ0LJWIeOLEifj0008BAOcOnsOJvScq41YIgviPUx0S75JDrKJ4eHjA09MTXbp0watXrxAWFobLly+jTZs2cvVisRhdunTB69fK85yVjCLfz6xfubebIAhCHtVh2wXNIVZRjI2N8fTpU7x69Qo2Njbo3LkzDAwM4OHhARMTE2hpacHLy0vQnzx5ElOmTBFe29nZaaLZBEEQcikUidQ2TUEOsYpy4sQJaGtrIyUlBYsXL8b58+dx+/ZtLFmyBBcuXMDp06eRnp6O58+L9xJOnToVTk5OSElJwaVLl/D5559r+A4IgiD+RxGHaQoaMq2ivH37Fv379y91PD4+Hv7+/tDS0sKRI0dw8uRJAMDjx48xYsSIym4mQRCEWmhy9ai6kEOsZixcuBCenp7Q09PDyZMn8fvvv5e5LjHnqq+e2o259Bm1a3PpXzP+rbuPCvO49K0kdbj0n73lWwpwv4gvIok+Z6SdExIuOQDAhGlx6TPB90x5I8+cTObLVVdwYiuXPnfbWS49AOi15Hvf6m77l0vvWLcBl76ZthGX/l/GF82nDzPh0gNAFHcJWWiVKVHuzJ49W9NNIAiC4IZWmRIEQRAEaMiUIAiCIAAAhZpugBqQQyQIgiAqHOohEgRBEASqx8Z8cogEQRBEhUMOkSAIgiAAMBoyJQiCIAhKEEwQBEEQAGgfIlHF0eKMHFGPc5lYXS0dLn2RiD8MSz7nzERj6HLpb3CO87zQ5tNrM75wwoZl+FrR43yfeT8Xr4r4oqTwRp7R7juBr/4N57j0AMDyKrb/YlnE91WbJ+bbpKAn4otGpIk9ENVhlWmlBve2sLBAWlrae9cTFhYGR0dHAEB2djZSU1ORmpqKS5cuYfHixdDR4fsiVsTYsWOFjPQLFixATk4OkpKScOnSJbXihubm5io9b2xsjEmTJgmvmzRpguDg4PdrNEEQRBWkOgT3rhHZLj744APY2trCxcUFzZs3x5YtWyrkOqtWrYKDgwMGDx6MzZs3Q1v7/TrYJiYm+OKLL4TXd+/exbBhw963mQRBEFUOcohy0NbWxvbt25GSkoLg4GDo6+ujR48eSExMRGpqKrZu3Sr08BQdV8S///6Lzz//HEOGDEGdOsVBnGfNmoXY2FikpKRg4cKFgvbIkSOIj4/HxYsXMXHiROG4n58fLl++jPDwcLi6usq9ztWrV5GXl6fyGlJq1aqF06dPIyEhAampqRg0aBAAYOnSpWjRogWSkpKwfPlymR60rq4uAgMDkZqaisTERHTv3h1Aca/10KFDOH78OP755x8sW7YMQHGC4G3btiEtLQ2pqamYPn268jeCIAiiEmEcpikq3SG2bt0aW7ZsgZ2dHV68eIGvvvoK27dvh7e3N2xtbaGtrY1JkyZBV1dX7nFV5ObmIjs7G9bW1ujVqxesra3h4uICe3t7ODo6ws3NDQAwfvx4ODk5wcnJCVOnTkXdunXRuHFj/PDDD3B1dUWvXr3Qtm1buddwcHDAlStX8PDhQ6XXkPL69Wt89NFHcHR0xAcffICVK1cCAL755htcu3YNDg4O+Prrr2XKTJ48GQBga2sLHx8f7NixA7q6xfNf9vb28Pb2RocOHeDt7Y1mzZrB3t4epqam6NChA2xtbbFt2za+N4YgCKICKRCpb+rQp08fZGZm4sqVK5gzZ06p876+vkhJSUFKSgqio6Nha2urss5Kd4g3b97EuXPFk967d+9Gz549kZ2djStXrgAAduzYAXd3d9jY2Mg9rg6i/8+43Lt3b/Tu3RtJSUlITExE69atYW1tDaA4oW5ycjLOnz8PMzMzWFtbo1OnTggPD8ejR4+Qn5+P/fv3y9Q7Y8YMZGZm4sKFC0JPUNk1SrZnyZIlSElJwenTp2FqaopGjRopvYdu3bph165dAIDLly/jxo0baNWqFQDgzJkzePHiBd68eYP09HRYWFggKysLzZs3x9q1a9GnTx+8ePFCbr0TJ05EXFwc4uLi0Nu3j1rPkyAI4n0pzx6iWCzG+vXr0a9fP7Rt2xY+Pj5o06aNjCY7OxseHh6ws7PD4sWL1ZpKq/RVpoyp1yGWOjVeDA0NYWlpiX/++QcikQg///xzqQfh4eEBT09PdOnSBa9evUJYWBj09PRUtm/VqlVYuXIlPvroI+zcuRMtWrRQeI2SjBw5Eg0aNICjoyMKCgqQnZ0tXE8Ryu7/zZs3wr8LCwuhra2NZ8+ewc7ODn369MHkyZMxfPhwTJhQenWev78//P39AQCDzD5U2gaCIIjyoqgcB0NdXFxw9epVZGdnAwCCgoIwePBgZGRkCJqYmBjh3+fPn0ezZs1U1lvpPUQLCwt07twZAODj44PTp0/D0tISLVq0AACMHj0aERERyMzMlHtcGbVq1cKGDRvw+++/49mzZwgJCcH48eNRq1YtAEDTpk3RoEEDGBsb4+nTp3j16hVsbGyE9ly4cAHdu3dH3bp1oa2trXCBi3T+cezYsQqvURJjY2M8ePAABQUF6N69OywtLQEUD+/WVpBENzIyEiNHjgQAWFtbw9zcHJcvX1Z47/Xq1YNYLMbhw4fx/fffo2PHjkqfFUEQRGXCs6im5EhWXFyczDoPADA1NcWtW7eE1zk5OTA1NVV47QkTJuD48eMq21jpPcT09HSMHTsWmzdvxpUrVzBt2jScP38ewcHB0NbWRlxcHDZt2oS3b99i3LhxpY7LIywsDCKRCGKxGEeOHMHixYsBAKdOnUKbNm2EXwovX77EqFGjcOLECXz++edISUnB5cuXcf78eQDAvXv3sHDhQsTExODu3btITEyElpb8/T2LFi3C3r170aZNG7nXePjwoaDds2cP/vjjD8TFxSE5OVn4FfPkyRNER0cjLS0Nx48fx/r164UyGzZswKZNm5CamoqCggL4+fnh7du3Cp+rqakptm3bBrG4+DfOt99+q9b7QRAEURnw9A9LjmTJQ94ImqLRve7du2PChAno1q2byuuKUD0CCBAVAO+QqScz4dKnayl24PIoy5DKraJ/ufR2YhMuvf0bzo35Wrwb87nkyJbwPyNDzuACacjj0l/Jf8KlD1suf/W2Ing35j8cxKcHgFot+AbL6u3P5NKvbPwBlz5d/Ea1qAT5nH87ToXKp2zkMTlnD3eZkiww91Vb+8PNvUrPd+7cGQsXLkTfvn0BFC9QBIpX7pekQ4cOOHLkCPr16yesR1EGRar5D2MqNuDSa3FGt5BwRjzJq4QdSM84Iyo2ZXxtMszni7Yj4fwiS+UP5oM8zvl4PfBFPTHU4ov+k7vtLJeeN/JMg2N8kXAAoCAtlEtvcHgJl74eZyCcFxK+6D95jO8CV7Qq/6u/QFR+fa+4uDhYW1vD0tISt2/fxogRI+DrK+twzczMcPjwYYwePVotZwiQQyQIgiAqgfIciiwsLMSUKVMQEhICLS0tBAYGIj09HZ999hkAYPPmzZg/fz7q1auHDRs2AAAKCgrg7OystF5yiARBEESFU97jP8ePHy+1UGbz5s3CvydOnFhqMY4qyCESBEEQFU55bruoKMghEgRBEBVO1XeH5BAJgiCISqCgGrhEcogEQRBEhVP13SE5RIIgCKIS0GRaJ3Uhh0gQBEFUOKwa9BHJIRIEQRAVDvUQCYVYWFjgzz//RIcOHbjKRUdHw9XVFRYWFujatSv27dtX5jZoc0aSMc3n+0gXgS+syksxfxgWkZjvHiyY8iTT7yIR84Uxq1XIFzFEizN6h00ZQm7pcn4TpZUhPBwPei357oHl8T1T3qgzAKDdoQeX/nXBQi59s0K+UGwNdfie0WsRXxip+owvGlF5UB22XVR6tguibEiDdru6FseBtLS0LBWqiCAIoqpSCKa2aQpyiAqwsLBARkYGtm/fjpSUFAQHB0NfXx89evRAYmIiUlNTsXXrVujoFPc4srOzsXTpUly4cAEXLlwQ0lZt27YNXl5eQr25ublyrxUZGYmEhAQkJCSgS5cuAIrzNoaGhmLPnj1IS0uTKb906VK4ubkhKSkJ06dPR2RkJOzs7IQ6z549y937JAiCqCh40j9pCnKISmjdujW2bNkCOzs7vHjxAl999RW2b98Ob29v2NraQltbG5MmTRL0L168QKdOnfDbb79h9erVal/nwYMH6NWrFxwdHeHt7Y21a9cK51xcXDB37ly0a9dOpsw333yDqKgoODg4YPXq1QgICICfnx+A4vyJurq6ghMlCILQNIzjP01BDlEJN2/exLlzxZH2d+/ejZ49eyI7O1uInL5jxw64u7sLeul83r59+4RenjpIJBL4+/sjNTUVwcHBaNu2rXAuNjYW169fV1lHcHAwBgwYAG1tbYwfPx7bt29X+/oEQRAVTXXoIdKiGiUoSjipjl7674KCAmH+D4AwxFqSGTNm4P79+7Czs4NYLMbr16+Fc//+q16+v1evXuHUqVMYPHgwhg8fDicnJ7m6iRMn4tNPPwUAXD2UiHP7zqhVP0EQxPtQHbZdUA9RCRYWFujcuTMAwMfHB6dPn4alpaUwPzh69GhEREQIem9vb+H/MTExAIDr16/D0dERADB48GC5DtHY2Bh3794FYwyjR4+Gtrbq3ym5ubmoXbu2zLGAgACsXbsWcXFxePr0qdxy/v7+cHZ2hrOzMzlDgiAqjerQQySHqIT09HSMHTsWKSkpqFu3LlatWoVx48YhODgYqampKCoqwqZNmwS9rq4uzp8/j2nTpmHGjBkAih2Qh4cHLly4gE6dOuHly5elrrNhwwaMHTsWMTExaNWqlVzNu6SmpqKgoADJycmYPn06ACAxMREvXrzAtm3byucBEARBlBOFjKltmkKE6hFirtLh3SeYnZ0NJycnPH78uIJbppgmTZogPDwcrVu3Vmu490uL4Vz193zNt3fploTv99bLMvw8uybi29/VgvFld++ez7cP8XUh3zPi3Yd4SVIZ+xD59v2lFsofjVDE4f5ccu59iDrjffguAP59iLVM3VWLSnDSRP01BQBwRJ/vc/QafPsQzTn34wLAvJt7ucuUxMd8iNrafTd/f69rlRXqIdYQRo8ejQsXLmDu3Lncc58EQRAVTXVYZUo9xP8wJrVacOl5I000NajHpW+kY8SlB4DRaMylPyji68FfeHaVS6+jxbdOTcQbLUif75kCwFvGGT2H83fyrbyHXPpnr9VbKFZWDCR8owAA8LrgLZf+39uRXHqHdnxBNPS1+Hpwb4ryufSP37zg0gPAnWfp3GVKMtx8sNraAzePvte1ygqtMiUIgiAqnOoQuo0cIkEQBFHhaDIkm7qQQyQIgiAqnOqwtoEcIkEQBFHh0JApQRAEQYDyIRIEQRAEgOoRuo0cIkEQBFHh0JApQRAEQQAaDcmmLuQQCYIgiAqHhkwriOjoaLi6uirV2NvbIykpCX369MHJkyeF49bW1li9ejVatWqF/Px8pKWl4csvv8SDBw8qutnlgrwsF2XFp35HLr1NEV/0jMdivj+A55zxGAEgEq+49K1hzKXXqtOKS887LCTmjFTTQMwfy7SA85d5AxHf+xyvxdcmx7oNuPSWRXxfU/X4AvMAAJoV8sXE5Y08k3SJLw7oOMdZXHqJiO9z9FCnLpceeP9INeU9ZNqnTx+sWbMGWlpaCAgIwLJly0pp1qxZg/79+yMvLw9+fn5ISkpSWme1jGWqyhkCxemaoqKi4OPzv0C/urq6+Ouvv7Bx40ZYW1ujbdu22LhxIxo04PsDrU6IOP9QCIIgKgLGmNqmCrFYjPXr16Nfv35o27YtfHx80KZNGxlNv379YG1tDWtra3z66afYuHGj6nrLfHcaJDc3FwDQuHFjREREICkpCWlpaejWrZugGTp0KPz8/NC7d2/o6hbHNvT19UVMTAz+/PNPQRceHo5Lly5BV1cXgYGBSE1NRWJiIrp37w4AGDt2LI4cOYJjx44hKysLkydPxowZM5CYmIiYmBjUqVMHABAWFoZff/0VERERSE9Ph5OTEw4dOoR//vkHixcvFq535MgRxMfH4+LFi5g4caLMPf34449ITk5GTEwMGjZsCACwtLTEuXPnEBsbi0WLFsk8h1mzZiE2NhYpKSlYuHAhgOIsHenp6Vi/fj0SExNhZmZWTk+dIAii7BSBqW2qcHFxwdWrV5GdnY38/HwEBQVh8GDZWKmDBw/Gzp07AQAXLlyAiYkJGjdWHvu4WjpEKb6+vggJCYGDgwPs7OyQnJwMoLgHmZ2djaysLISHh6N//+J8M+3bt0dCQoLcuiZPngwAsLW1hY+PD3bs2CE40vbt28PX1xcuLi746aefkJeXh44dOyImJgZjxowR6nj79i08PDywadMmHD16FJMnT0b79u3h5+eHunWLhyjGjx8PJycnODk5YerUqcJxQ0NDnD9/Hvb29oiMjBSc5Zo1a7Bx40a4uLjg3r17wrV69eoFa2truLi4wN7eHo6OjnBzcwMA2NjYYOfOnejYsSNu3rxZXo+bIAiizJRntgtTU1PcunVLeJ2TkwNTU1NuzbtUa4cYFxeHcePGYcGCBejQoYOQWNfHxwdBQUEAgKCgIJlhU0V069YNu3btAgBcvnwZN27cQKtWxfNHYWFhePnyJR49eoTnz5/jjz/+AACkpaXB0tJSqOPYsWPC8UuXLuHevXt4+/YtsrKyhJ7a1KlTkZycjPPnz8PMzAzW1tYAgDdv3gg914SEBKFeV1dX7Nu3DwCE9gFA79690bt3byQlJSExMRGtW7cW6rpx4wYuXLjA+TQJgiAqDp4EwRMnTkRcXJxgJUfTAPlTQe8OtaqjeZdquahGSlRUFNzd3fHhhx9i165d+OWXX7Bnzx54eXlh0KBBmDt3LkQiEerVqwdDQ0NcunQJHh4ecutSNtf25s3/JtyLioqE10VFRdDW1i6lK6kpqfPw8ICnpye6dOmCV69eISwsDHp6xQsS8vP/l76lsLBQpl55b6JIJMLPP/+MLVu2yBy3sLDAv/8qTq8zceJEfPrppwCAy4cSEL3vjEItQRBEecGzqMbf3x/+/v4Kz+fk5MhMBzVr1gx37tzh1rxLte4hmpub48GDBwgICMDWrVvRsWNHeHp6IiUlBebm5rCysoKlpSUOHTqEIUOGYO/evejataswhAoUr1Rq3749IiMjMXLkSADFK1HNzc1x+fLlcm2vsbExnj59ilevXsHGxgadO3dWWSY6OhojRowAAKF9ABASEoLx48ejVq1aAICmTZuqtTjI398fzs7OcHZ2JmdIEESlUZ5ziHFxcbC2toalpSUkEglGjBghjNBJOXbsmDCl1alTJzx//lxm2kke1dohdu/eHcnJyUhMTISXlxfWrFkDHx8fHDlyREZ36NAh+Pr64vXr1xgwYAC+/PJL/PPPP7h06RL8/Pzw4MEDbNiwAVpaWkhNTcX+/fvh5+eHt2/5koaq4sSJE9DW1kZKSgoWL16M8+fPqywzbdo0TJ48GbGxsTA2/t+WgVOnTmHv3r2IiYlBamoqDh48WG7bMQiCIMqb8lxlWlhYiClTpiAkJAQZGRk4cOAA0tPT8dlnn+Gzzz4DAPz999/IysrC1atX4e/vjy+++EJlvSKgGuyWJCqESRbDuPRVcR/iM/BlCq8LCZf+SlEul/4/uQ8x/yGX3lFS/fchTsEt1aISVLl9iEWvufQA8Netv7nLlMS5ibva2ri7ke91rbJSrecQCYIgiOpBEav6+S7IIf6HuVzwjEtfV7s+l75VPt+I/Clt/iHqB0V8kWpeiPg+8jffPuHSF3H2xsScv+xb6Vlw6QGgQMTXpiO5GdzX4KGZthGXPk/MN3LwQsI3agAADXX4et76b/l60bw9vm0JK7j0yXYzufRW7V9y6QGgAV+nuBQU3JsgCIIgoHrLQ1WAHCJBEARR4VAPkSAIgiBA2S4IgiAIAgD//LomIIdIEARBVDiFtMqUIAiCIGjIlCAIgiAA0JApQRAEQQCgHiJRxdEV8739hoxvo71+Ed8fQEdmwKUHgKda+lx6A8a3Ef7cqytcet55EmVZVuShr2fFpS8LLoZ818h4c59L/y/j2zivJ9Li0ucx/thtr0V8m//fFPHdA29oNd6N9vYpK7n0Ue2+4dKXB9Whh6jyG87CwgJpaWmV0RYZxo0bh9TUVKSkpCAtLQ2DBg2q1OsnJydj717Z+IPbtm2Dl5cXgOIciY6OjmWqe+DAgZgzZ857t5EgCKK6UMQK1TZNUSV7iKamppg7dy46duyIFy9eoFatWmqlNlIXLS0tFBYqfuitW7eGWCyGu7s7DAwMkJeXV27XBoA//vhDSDJMEATxX6A6bMxXawxMS0sLW7ZswcWLFxESEgI9PT3Y2dkhJiYGKSkpOHz4MExMTADI9pzq1auH7OxsAEDbtm1x4cIFJCUlISUlBS1btgRQnONPenzTpk0Qi8Vo2LAhcnNz8fJlcby9f//9F9evX0eDBg0QHx8PALC1tQVjTEgAefXqVejr68Pc3BynT59GSkoKTp8+LZzftm0bVq5cidDQUCxbtgzNmzfH8ePHER8fj8jISNjY2Aj36+vri127duHkyZNq9Ux79eqFc+fOISEhAQcOHBByFPbr1w8ZGRmIiorCmjVrBCc4duxYrFu3DgAwYMAAnD9/HomJiTh16hQaNmwIAFiwYAG2bt2KsLAwXLt2DV9++aVwvdGjRyMlJQXJycnYuXMnAKB+/fo4ePAgYmNjERsbi65du6rz1hIEQVQK5Zn+qaJQyyFaW1tj/fr1aN++PZ49ewYvLy/s3LkTc+bMgZ2dHdLS0rBgwQKldXz++edYs2YNHBwc4OTkhJycHLRu3Rre3t5wdXWFg4MDCgsLMXLkSKSkpOD+/fvIzs5GYGAgBgwYAAB4+PAh9PT0ULt2bbi5uSEuLg5ubm5CouBXr17ht99+w86dO2FnZ4c9e/Zg7dq1QhtatWoFT09PzJo1C1u2bMGXX34JJycnzJo1Cxs2bBB03t7e2L9/P/bt2wcfHx+l91WvXj3MmzcPnp6ecHR0RHx8PL766ivo6upi8+bN6NevH9zc3BT2cM+ePYvOnTujY8eOCAoKwtdffy2ca926Nfr06QMXFxcsWLAA2traaNu2LebOnYsePXrA3t4e06ZNAwCsWbMGq1atgouLC7y8vBAQEKD8TSUIgqhEyjNBcEWh1pBpdnY2UlJSAAAJCQlo0aIFTExMEBlZnLNqx44dCA4OVlpHTEwM5s6di2bNmuHw4cO4evUqevbsCUdHR8TFxQEA9PX18eDBAxQVFaFv375wdnZGz549sWrVKjg6OuKHH37AuXPn4OrqCnd3dyxZsgR9+/aFSCRCVFQUAKBLly74+OOPAQC7du3C8uXLhTYEBwejqKgItWrVQteuXWXarKurCwBwcnLCw4cPcfPmTeTk5CAwMBAmJiZ49uyZ3Pvq3Lkz2rZti+joaACAjo4OYmJi0Lp1a2RlZeH69esAgH379uHTTz8tVb5Zs2bYv38/mjRpAh0dHaFHDQB//fUX3r59i8ePH+PBgwdo1KgRevTogYMHD+Lx48cAgKdPnwIAPD090bZtW6GskZERDA0NhV62lIkTJwrtOHfwHE7sPaHsbSMIgigXakxw7zdv/pc8s7CwUBgelUdBQQHE4uKOp57e/1Kq7Nu3DxcuXMCHH36IkJAQfPLJJxCJRNixYwe+++47uXXFxcUhLi4Op06dwrZt2/DDDz8gKioKbm5usLCwwNGjRzFnzhwwxvDnn3/KraPkm/Dvv/8CAMRiMZ49ewYHB4dSeh8fH7Ru3VpwTEZGRvDy8sLWrVvl1i8SiXDq1Cn4+vrKHLe3t1fwhGRZt24dfv31V/zxxx/w8PDAwoULhXPvPndtbW2IRCK5HyyxWIwuXbrg9WvliT/9/f3h7+8PAOhn1k+tNhIEQbwvNWKVqTyeP3+Op0+folu3bgCK57QiIiIAANevXxfmEIcOHSqUsbKyQlZWFtatW4djx47B1tYWZ86cwdChQ4XhxDp16sDc3BxNmjSRcVb29va4ceMGACAyMhKjRo3ClStXwBjDkydP0L9/f6GHdu7cOYwYMQJA8fzk2bNnS7U/NzcX2dnZMu2ztbWFSCTCsGHDYGtrCysrK1hZWWHw4MFKh03Pnz8PV1dXtGjRAkBxL9fa2hqZmZlo3rw5LCyK89d5e3vLLW9sbIzbt28DKJ5bVMWZM2cwfPhw1K1bV3hmAHDy5ElMmTJF0NnZ2amsiyAIorIoYkVqm6Yo8yrTsWPHYtOmTTAwMEBWVhbGjRsHAFixYgUOHDiA0aNHIzQ0VNB7e3tj1KhRyM/Px71797Bo0SI8ffoU8+bNw8mTJyEWi5Gfn4/Jkyfj7t27WLFiBZo2bYrXr1/j4cOH+PzzzwFAxjECxXNwzZo1E4Y0p06disDAQMyePRsPHz4U2vUuI0eOxMaNGzFv3jxIJBIEBQWhTp06uH37Nu7cuSPoIiMj0bZtWzRu3FhuPY8ePYKfnx/27dsnDLvOmzcPV65cwRdffIETJ07g0aNHiI2NlVt+4cKFCA4Oxu3bt3H+/HlYWSnfA5aeno6ffvoJERERKCwsRFJSEsaNG4epU6di/fr1SElJgba2NiIjIzFp0iSldREEQVQW1WGVqQioBq2sptSqVUsYpl2/fj2uXLmC1atXa7ZRJeAdMnUX1eXSt3jL99HKkfBtXgaAp2K+a/BuzF/1LJ5LX9Eb8z+tU7a9rzxcZXzbjHg35lvp1OPS1xPzZbN/XKR82kAejcV8AR5iXudw6R31mnLpv8jn+xxVxsb8ng+UrxNRRb3a1mprH+fyBcQoL6rkPsSawsSJEzF27Fjo6OggKSkJmzdv1nSTZHhSwPfF91ZSh0sfpssXzSPu7T0uPQC8LeKLSmKpw+fUG+vz6XnDU4nA5xBzwb9pmfcaTzkdiokWX4ShPsyES897y1e0+L/W6jO+aDh/vnnBpX/I+bmzav9StagEvA7O7dJSLj0AoMH7OcTqMIdIDrECWb16dZXqERIEQWiKGrPKlCAIgiDeh+owh0gOkSAIgqhwCosoQTBBEARBUPongiAIggBoUQ1BEARBAKBFNQRBEAQBgIZMCYIgCAIAUESLagiCIAii+oREY2RkJW3ixInVWl8V21TV9FWxTXTPmteXtUwNMo03gKyKWVxcXLXWV8U2VTV9VWwT3bPm9WUtU1OsTOmfCIIgCKKmQQ6RIAiCIEAOkZDDli1bqrW+Mq5R3fWVcY2qpq+Ma1R3fVnL1BQoHyJBEARBgHqIBEEQBAGAHCJBEARBACCHSBAEQRAAyCESJdDT00OrVq003YwagUgkQpcuXbj0w4YNq8AWVS4GBgbcZUQiEWrXrl2u7RCLxZg+fTpXmXbt2nFfh+dvp1mzZtz1l4WyvAdEFdgMSaZ5GzBgAMvMzGRZWVkMALOzs2NHjx7lqsPPz0/ucRsbG9ajRw9Wq1YtmeN9+vSRq3d2dmZOTk4MAGvTpg2bMWMG69evn9rt2LFjB1e7XV1d2YwZM1ivXr0UasRiMTt16hRXvefOnePSR0REcL9vrq6uwnOvX78+s7S0lKubMWOGUlN2DXNzc9azZ08GgOnp6TFDQ0OF2i5durBLly6xGzduMADM1taWrV+/XqF+z549rHbt2szAwIBlZGSwO3fusFmzZinUT506ldWuXZsBYAEBASwhIUHp+waAhYWFcT3TqKgoduHCBTZp0iRmbGysUs/7txMfH69WO8r6nvG+BwCYpaUl09XVFV7r6ekxCwsL7s9jDTCNN4CsClh8fDwzMjJiiYmJwrGUlBSuOqR/gCXtyy+/ZJmZmezIkSMsOzubDRo0SDiXkJBQSj9//nwWExPD4uLi2JIlS9iZM2fY999/zyIiIth3331XSn/06FEZO3bsGMvNzRVey2vnhQsXhH9/8sknLCkpic2fP5+dPXuWzZkzR+H9HT16lBkZGan9PBYuXMg+/vhjtfXz5s1jM2fOZM2aNWN16tQRTJF+/vz57NixY+zy5csMAGvSpAk7e/asQu38+fPZnj172D///MNWrFjBVqxYwS5fvsz8/f0VXuOTTz5hsbGx7OrVqwwAa9myJTt9+rRC/fnz51mzZs1kPkdpaWkK9UlJSQwA8/X1ZStXrmTa2tpKP3fJyckMAOvduzc7evQos7W1lfs5Kmk//vgjW7duHevWrRtzcHAQTFmZli1bsiVLlrArV66wPXv2ME9Pz3L72/ntt9+EH3zKrKzvGe97ABRHp5FIJMJriUTCYmNj1f7s1iDTeAPIqoCdP3+eAVD5R52SkiLXUlNT2evXr0vpU1NThZ6hhYUFi4uLY1OnTi11rZJ6sVjM9PX12fPnz4XegJ6entz2JCQksF27djEPDw/m7u7OPDw82J07d5i7uztzd3eXe68lrxsbG8vq16/PADADAwOWmpqq8Bnt37+f3bhxgwUEBLA1a9YIpkj/4sULVlhYyN6+fcueP3/OXrx4wZ4/f65Qn5WVVcquXbumUC91Jjw/YkJCQmR6eIaGhuz48eNKryGRSGSuoewZyfscSZ2YPLt48SLT1tZmBw4cEN4vZXrp/a1evZoNGTJE4eeopIWGhpayM2fOqPybEIvF7OOPP2Y5OTksPT2dZWRksI8++qjMfztSu3TpEisoKGBXr14V/naU6XnfM973oORnqaSpKlMTjbJdEACAixcvwsfHB1paWmjZsiWmTp2Kc+fOldI1atQIffr0wdOnT2WOi0QiuXotLS38+++/AIAbN26ge/fuOHjwICwsLCASiUrpCwoKUFRUhFevXuHatWvIzc0FALx+/Vpu+hgnJydMmzYNc+fOxezZs5GSkoJXr14hMjJS4b2KxWKYmJhALBZDJBLh0aNHAIC8vDwUFBQoLPfXX3/hr7/+Unj+XYyMjNTWAkDz5s259G/fvgUAIfGqOnNG5ubmQjlpHZaWlgr1b968QX5+vvBaS0tLaaLXW7duoUuXLmCMQSKRYOrUqcjIyFCo37x5M65fv46UlBRERkbC3NwcL168UKhPSEhASEgIrKys8O2338LQ0FBlWqEePXooPf8uHTp0wLhx4/Dhhx/i1KlTGDhwIJKSktCkSRPExMTgyJEjMnp1/3ak9OvXj6s9vO8Z73sAAA8fPsTAgQPxxx9/AAAGDRok/F3819C4VybTvOnr67Mff/yRxcbGsri4OPbjjz/KzClILSAggLm6usqtY8+ePaWOnTlzhtnZ2ckc09LSYjt27GAFBQWl9OfPn2f6+voMABOJRMJxIyMjpUNjpqam7MCBA2zdunVyh25LWnZ2Nrt27ZrQA2vUqBEDwGrVqiX3l3JJ09PTY61atVL7uY4cOZLNmzePAWDNmjVjzs7OpTQffPABA8A++ugjuaao7pkzZ7JNmzaxa9eusU8++YSdO3eOTZkyRWl7vvvuO5acnMwWLFjA5s+fz5KSkti3336rUL9s2TL27bffsoyMDObp6ckOHz7MfvzxR4X6evXqsd27d7N79+6x+/fvs127drG6detyfRa1tLQUnhOJRMzBwUGY26tbty7r0KGDyjr79+/PZs+ezb7//nvBFGkjIiLYqFGjmJ6eXqlzo0aNUvq3ExsbyxYvXiz3b6ekqTv3W5b3rCzvQfPmzVlMTAy7ceMGu3nzJouOjmYtWrTget9qglGkGqJCMTU1RUFBAe7fv1/qXNeuXUv9ktbR0ZH5NSylXr16aNKkCS5evKj0ev3794erqyvmzp3L3VZ9fX00atQI169fl3t+wIABWLFiBXR0dNC8eXPY2dlh0aJFGDx4sFz9hg0bUFRUhB49eqBt27YwMTHByZMn4eLiIqNbuHAhFi5ciMDAwFJ1MMYwYcIEhW329PRE7969IRKJEBISgtOnT6u8TwcHB7i5uQEAIiMjkZycrFArEokwYcIEmWsEBASovIa66OjowMvLC5aWltDW/t+A1eLFixWWadq0KSwsLGT0UVFRCvUbN26EgYEBPvjgAwQEBGDo0KGIjY3FJ598Ilc/bdo0rFmzRubY1KlTsXbtWrl6e3t7pc/wXebPnw8nJyfY2NjAxsYGTZo0QXBwMLp166awTMeOHYXzqt6z96FWrVoQiUR4+fJlhdRfHdC4VybTnB07dqzUwpSSpqjc+PHjSx37+eefq42+LGXkLZ5QNp8m7dHyzOXwmIGBAROLxQwAa9WqFRs4cCDT1tZWWY6nd8JrJedWpbZo0SKZxVQl7fjx4ywoKIjNnj2bffXVV4Ipqn/p0qUsOzub/fXXX+zYsWPC51dZm6Tzc9L/16pVi4WEhKh830qasnnK0NBQlpGRwRYtWsTatm2r8hmVZe4XAGvQoAEzMzMTTJFu+/btMqtjTUxM2NatW5XWraOjw3x8fNi3336rVi+6phrNIf7HWbFiRZnKDR06FK9fv8bevXsBAOvXr4eurm610ZelTEFBQan5LWXzafn5+RCLxYKmfv36Sue7GjZsiCVLlqBp06bo378/2rRpgy5dusjtOQLFPQU3NzfUqVMHp0+fRnx8PLy9vTFq1CiF1yjZO9m+fTskEgl2795dqneSmpqq9N7s7OzkHtfT00Pr1q0RHBwMAPDy8sKlS5cwYcIEfPDBB5gxY4aMvlmzZlxzakOGDIGNjY3cUQRFvHr1CkDxHHGTJk3w+PFjWFlZldKNGDECvr6+sLKywtGjR4XjtWvXxuPHjxXW36NHDzRq1AjDhw/Hli1bYGRkhP379+Onn36Sq+ed+x04cCBWrlyJpk2b4sGDBzA3N0dmZibat28vV29ra4vnz58Lr589ewYHBwel1zh69CieP3+OhIQEvHnzRqm2pqNxr0xW/UxPT4+dPHmSjRgxgu3YsYOtWrWqWunLUiYgIID5+PiwlJQU1rJlS7Z27Vq2ceNGhXpfX1929OhRduvWLfbjjz+yzMxMNnToUIX6v//+mw0bNkzoRWppaanVA50yZQqbPXs2A1SvuFS3d2Jubq7UFNV/5swZmTlALS0tdubMGSYWi9mlS5dK6Tdv3szat2+v9ufu77//LrWfVZXNmzePGRsbs48//pjdvXuX3blzhy1atEjuPXt4eLBz584Jq5Td3d2Zg4OD0nnNkta+fXu2c+dO9ubNG4Ua3rnf5ORkVrduXeE96969O9u8ebNSvYmJifC6Tp06Sj9HgOptGf8h03gDyKqAtWzZkgUHB7NLly6xa9euCfauruT+OHNzc5aYmMjWrVuncM9cVdOXtQyg/sKjkmZjY8O++OILNnnyZNa6dWulWum+r5LOStkin8TERNa5c2cWExMjDNWp+uKT7sGUOlMDAwOVw3WNGjViAwcOZAMGDBAWICmyzMxMmb2aRkZGLCMjo9R9Se3SpUvszZs3LDMzU60tCAcPHmRXrlxhmzZtUmvry7umo6PDtZdUHWvdujVbsGABS0tLY+Hh4ezzzz9nDRo0UFrG09OTLV++nP3yyy9K9zgC/8tgn5ycLCw0K7mX9l0bPXo0S09PZ4sWLWKLFi1iGRkZchcDlTTeHyY11WhRDQGgeFHCggULsGrVKgwcOBDjxo2DSCTCwoULZXRZWVlgjEEkEgn/l8IYQ4sWLaq0vqxlyopYLEajRo1kFoDcunVLrjYsLAxeXl44deoUHB0d0alTJyxbtgzdu3eXq3d3d8fMmTMRHR2N5cuXw8rKCtOnT8e0adMUtmfmzJmwtrZGr1698PPPP2P8+PHYu3cvfvvtN7n6CRMmYP78+QgNDYVIJIKHhwcWLVqEbdu2ydWPHz8e8+bNQ3h4OEQiEdzd3bFkyRLs27cPCxcuxNdffy2jNzc3l1vPzZs35R4fM2aM3OM7d+4sdeyjjz6Sq5Xy7vaJqKgouLm54cWLFzLDxdLPibGxsdx6YmJisG/fPgQHB+Pu3btKr1kWTp06hSFDhmDp0qWoV68eHjx4AGdnZ7i6uios07ZtW3zwwQcQiUQ4c+aMym0Xly5dQsuWLZGdnY03b94I96xoaLymQg6RAADEx8fDyckJqampsLW1BVA8R+Xu7q7hllUdrK2tMWvWrFIrInv27ClXP2XKFCxYsAD3799HYWGhyi8ZBwcHrFu3Du3bt8fFixfRoEEDDB06FGlpaeV6HzwrUzMzM9G1a1c8efIEAFC3bl2cO3cOrVu3VlimSZMmGD16NDIzM1GrVi3k5OQoXQVqa2srrHqNiopCamqq0vZLJBIhbujly5cV7h1VNPcKQOXq3YrgXUf7LoocroGBAV69egWxWIyRI0fC2NgYe/bsEd4TefD8EAP4f5jUVGhRDQGgeOO7SCTClStXMHnyZNy+fRsNGzZUqNfW1sakSZMEhxkeHo7Nmzcr/HKqavqylAkODsamTZsQEBCAwsJChfVKmTZtGmxsbJR+cZUkKSkJHh4esLGxgUgkUvplDwChoaFyv2AVOWgpp0+fVmt7BgDk5OQIwREAIDc3V+kX64QJEzBt2jQ0a9YMycnJ6Ny5M2JiYhS2aerUqZg4cSIOHz4MANi9eze2bNmisMfq4eGBHTt24Pr16xCJRDAzM8PYsWPlOtzx48erdY8lEYlESE1NRYcOHVRq9+/fD29v71ILkBT98JEGavjhhx9w79497Nq1CyKRCCNHjlQa1DwvLw/m5uawtrbGzp07oa+vDy0tLYV63h9iAGTeY2XHajrUQyQAFEd8ycjIgImJCRYvXgwjIyMsX74csbGxcvX+/v6QSCTYsWMHAGD06NEoLCzExIkTq4W+LGWkvWh1CQ0NRa9evVQ6T96hPSkdO3YU/q2npwcvLy8UFBRgzpw5pbS8w4HS1aD29vbo0KEDjh49CsYYBg8ejNjYWEyaNElum1JTU+Hs7Izz58/DwcEBNjY2+OGHHzBixAi5+pSUFHTp0gV5eXkAintDMTExCr+84+Pj4evri3/++QdAca993759St+X77//Xu5xRXsdd+/ejW+//Vap4weAxo0b4969e9y9q/Pnz6Nz584qj0n55JNP8Omnn6Ju3bpo2bIlWrZsiU2bNsHT01Ou/sqVK+jUqZPaP8QAIDs7G2ZmZnj69ClEIhFMTExw9+5dPHjwABMnTkRiYqLadVVnqIdIACgeQtq1axcsLCwgkUgAFDsMRV9Mzs7OsLe3F16HhYUp3Sxc1fQ8ZerUqQMA+OOPPzBp0iQcOXJEZmn6u2HspM4kKysL4eHh+Ouvv2T0q1atktEPHDhQ5t/S8FlA8fuiyCG++yV17tw5hIeHy9VKhyTVDScn7bFcu3YN165dE46X3I4gj9evXwv3qqOjg8uXL8PGxkahXiQSyfxgkPZoFCGRSARnCBR/+Us/r4qQhg4Ein84DBgwQOmcWpMmTXDp0iXExsbKlH03AMO9e/cA8A8rFhYWwtfXF0FBQWCMwcfHR+mPpsmTJ8PFxQUXLlwAAFy9elXp6M2tW7dktl2ow4kTJ3DkyBGcPHkSANCrVy/07dsXBw4cwIYNGxQ665oGOUQCALBnzx7Mnj0baWlpKmNDAsV/1M2bN0dWVhYAwMrKSukfdVXT85RJSEiQWXwze/Zs4Zy8RThSZ3Lz5k3cvHkTOjo60NHREfTvUnJoLzExUe2hPqmjBornjBwdHdG4cWOFep7hwEWLFsm8NjQ0BGNMxkHIIycnB8bGxvj9999x6tQpPH36FHfu3FGo37ZtGy5cuCA4/SFDhmDr1q0K9fHx8QgICMCuXbsAACNHjkRCQoLSNv36668yr1esWIFjx44p1P/www9K65OiaE5Q1SIcX19frFmzBmvWrAFjDNHR0fD19VV4Hd54sur+ECuJk5OTTK//1KlTWLJkCWbOnKlyP29NghwiAaA4uG/JnokqZs+ejbCwMGRlZUEkEsHCwgLjxo2rNnqeMtKg27q6uqU2Lcv7spA6k6FDh+LgwYMy54YOHaq0Tcq+6N6lpKMuKChAdna20oUijDGkpKTAzMxM5XCglHbt2mHXrl2oW7cuAODRo0cYM2YM0tPT5eo//vhjAMVOJSwsDMbGxjhx4oTC+letWoXw8HB069YNIpEI48aNU9qznzRpEiZPnoypU6dCJBIhMjISGzZsUOtepBgYGCgNpK4sMHxJeIO3S7lx4waGDBmitj4iIgLffvst9PX14enpiS+++ELp36q8H2KqePLkCb7++msEBQUBALy9vfH06VOIxWK1fiDXFGgOkQBQHG3Dx8cHZ86ckfnSVzRcBxQPiUkXgGRmZqqMHlLV9LxlEhIS4OjoqPJYWfXqnH9fzpw5A2dnZ5XDgVKio6Mxd+5cYSjWw8MDS5YsUbrkXx1q166N3NxcmV5uSd4dhn4fSi560dLSQoMGDbBo0SKsX79err5Tp05Yt24d2rRpAx0dHSFji6IeX0kMDAwwZMgQ+Pr6YsCAAXI19evXx8SJE0utVlb0Y6ai48kCxbGCFyxYIEQsOnv2LBYtWoTnz5/D3NxcZti8JkM9RAIAMG7cOLRu3RoSiUT4Rahs/kpbWxufffYZ1yrQqqTnKdOoUSOYmppCX18f9vb2wtCpkZGR3LBbffv2Rf/+/WFqaioTJNrIyEhue44dOyZ8YTdv3rzUPJ0iZ1WWlbXqDgdKqVWrlsy8ZEREBGrVqsVVhzz27t2LgQMHCr1cKdLhxneHoRWt6JSibAVlScckDTSvbDj9t99+w4gRIxAcHAwnJyeMGTMG1tbWCvUSiQT9+/eHr68v+vbti0OHDmHTpk0K9UePHkVUVBROnz6t1mplxhgCAgJknODZs2cVBgOvX78+vv76a7Rr1w56enrCcWWrjx8/foypU6eiVq1apYbF/yvOEKAeIvH/lNx/qA5VbdVoRa4yHTNmDPz8/ODk5IT4+HjheG5uLrZv317qR4OtrS3s7e2xaNEizJ8/X0YfFhaGZ8+eyehV7fVUNITHe888c4hSDh8+jMTERGHObtSoUXByclK5Mra8KeuKTikODg7o1q0bGGM4e/as0mHZuLg4ODs7IyUlRXC00dHRpXrFnp6e8PHxQZ8+fRAWFob9+/dj3bp1cuOkliQpKUllbFFV3Lx5U+GzCAkJwf79+zFr1ix8/vnnGDt2LB4+fIhvvvlGYX1dunRBQEAADA0NYWFhAVtbW3z22WeYPHnye7WzOqLxcDlkmrctW7awNm3aqK2Xl7VBWSaHqqYvS5mPP/6Y65mqk3niXePJt1iWe969e7fSTAnvmomJCVuzZg1LSEhgiYmJbPXq1TJxMt/XTp8+rdYxqS1dulStYyXt+++/Z6mpqWzhwoVs4cKFLDk5mc2dO1ehPiIigkkkErZjxw62bNkyNn36dLnPtbCwkIWHh8tkC5EX7vBdW7x4MevXr997PTdlOT/j4+MZIBujNjw8XGl958+fZ82aNZMJr/dfjG9KQ6YEAKBbt24YO3as2qGbqtqq0YpcZSrlzJkzWLlypdCji4iIwKJFixRmeLe0tMTPP/+Mtm3bygxdKQoNx5tvsSz3rO6WAinPnj1TGgqurOjq6sLAwAD169eHiYmJzDB006ZNFZbr1atXqZ5Ov379lPZ+fHx84ODgIMyNL126FImJiQqzUYwePRpaWlqYMmUKZsyYATMzM3h5eZXSOTo6YsSIETh9+jSysrIQFBSkdMO8lGnTpuG7774TVo8qWpWqqBcuEomgr6+vsH7pitS7d++if//+uHPnDpo1a6ayXTk5OTKv1RnOrWmQQyQAFM978VByhSZQ/OWv7irQqqAvS5mtW7fi4sWLGD58OIDiL85t27bJ/bIEircUSOPDfvDBB0J8WEUsXLgQLi4uwpxdSkoKLC0t1Wq/uitr1Z1DDAwMVLjilTGmMLmuunz22WeYPn06mjZtioSEBOG5vHjxQu5il88//xxffPEFmjdvjpSUFOF47dq1ER0drfRa169fh56enuAQdXV1lc6LSYdfX79+XWr7SUmSk5ORnJyMb775Bl27doWPjw90dHTw999/48iRI/D395dbTt3VqSX3p77Ln3/+qfDcjz/+CCMjI8ycORPr1q2DkZFRqbRb73Lr1i106dIFjDFIJBJMnTpVZfzTmgjNIRJlQldXFzNnzhQm6k+dOoVVq1YpzKVW1fRlKSNv7kfZfBBvfFhptJLExEQhCk3JeSx5lGVlrTpIt0+UxNzcHNOnT4eWlhbMzMzK5TpTpkxRGKatJEZGRqhTpw5+/vlnmd5gbm6uyhWpR44cgbOzM06dOgXGGHr16oWzZ8/iwYMHACD0gMuaA7IkIpFImFtUtp/UxMQE1tbWMiMH8sLPiUQiDB06VMgvWVHUq1cPa9asgaenJ0QiEU6ePIlp06ZxRbupCZBDJMrE/v378eLFC+zZswdA8bBUnTp1hN5TVdeXpcy5c+cwe/ZsoUfStWtXrFixAl27dpWrP3v2LNzc3HDw4EGEhobi9u3bWLp0qcLA2AEBAThz5gy++eYbeHl5YerUqZBIJArDpOnq6uKLL74QFotERUVh06ZNSn8ElGVLgZWVFb777ju4u7tj1apV2Lp1q8xG8felXbt2pYaVpYt4FNGgQQMZvbJ9lYoyZEiRZspQtEhFiqKFO127dkVycjLy8vIwcuRIdOzYEWvWrFGo5433GhERAQ8PD6VtA4pHDH755ResXbtWrmOviKHvmojGJzLJqp9VtUUylbGoxs7OjiUnJ7Ps7Gx2/fp1lpiYyGxtbRXqnZycWK1atZipqSkLDAxkBw8eZJ06dVKoL5lvMTY2li1evFhpvsX9+/ezgIAA1r17dyFp7IEDB5Tec1xcHGvRogVLTExkYrGY+fn5sZ9++kmutnXr1mzXrl3s4sWLbOzYsWonyeWx+fPns9DQUHbv3j0WGBjI7t69y4KDgxXqBwwYwP755x/28uVLlpWVxQoLC9nFixdVXkcikbB27dqxdu3alWmxkzKTLl6xtbVlKSkpbOrUqUoXsaSmpjJdXV0h16WNjQ0LCgpSqJ83bx6bOXMma9asmUwuT3nPBgAbM2aMXJNX99q1a2XySr5r5f1+VwPTeAPIqqFt27ZN5svdxcWFrV+/vtroy1oGAKtduzarXbu2wvMhISHCv7/55hu1n6m9vT3Xe1CWHwHSZLMlVyBGR0eX0h04cIBlZWWxL774gtWvX1/mi1hZEmVeS01NZSKRSGh3w4YN2bFjx5TeM0/2eADMw8ODXb9+nYWHh7OIiAiWlZXF3NzcFOpfvHjBnj9/zp4/f85evXrFCgoK2PPnzxXqpcmWv//+ezZ+/HiZY/JMmgg6KSmJ6ejoCP9WpM/Kyipl6qxmVccUOU9lTrQmGy2qIbiQzrNIJBKMGTMGN2/eBGMMFhYWcsN5VTV9WcqoWpDwbozIBg0aCP8eNmwYli5dqrS8lF9//RVNmjRBcHAwgoKCFLZfSlJSEjp16iQEfXZxcVG5wCQvLw8SiQTJyclYtmwZ7t69K3ejvbOzMxhjmDVrFmbOnAkAwsKX8kyi/OrVKzDGUFBQgNq1a+PBgwdKw6rl5+fjyZMnEIvFEIlECA8Px7Jly5ReY+XKlejdu7faGTLeXfQyePBguLi4KKw/NzcX33zzDUaNGgV3d3eIxWKlAcd5470qex7yKBnoQcrz588RHx+PzZs3ywypy0us/F+GHCLBhaJwVNVFX5YyK1asQHJyMo4fPy5sSVEGTzzSkvTo0QONGjXC8OHDsWXLFhgZGWH//v0Ktwd06tRJcOhA8RxYRkaG4PDlLQJRd0uBqs3l5UV8fDyMjY3h7++PhIQEvHz5UmHKMaB4G0itWrUQGRmJPXv24MGDB0oj8wBly5BRkqNHjyrd1uHt7Q1fX19MmDAB9+/fh5mZGX755ReFet54rwDfPGtWVhYaNGiAffv2Ce27f/8+WrVqBX9/f5k5VXnOsySKtuPUVGhRDUGowM7ODiNGjEDfvn2RkJCAffv24cyZMwr1T58+RWRkJEQiEdzc3EpFmlHnS6Z9+/b4+uuv4e3trTDbQFkXgfAgb8HI6tWr1Q4OzoOFhQWMjIyQlpamUGNgYCAks1Y3e3xgYCCKiopkMmRoa2srXAVacv+fWCyGk5MTPDw8FC6eKgsmJiYwMzOTiWWalJQkVzt//nx0794dbdu2xd9//41+/frh7NmzGDZsmFy9vEU40mMXL15E+/btheNljZJUUyGHSBAcdOnSBT4+PvD09MScOXPkZh0o65dM69at4e3tjaFDh+Lx48cICgrCoUOH8PDhQxmdvr4+8vPzhZ5Rq1at0L9/f9y4cUNh7NmSe/fkoWhLgXTbh62tLXbt2oWtW7fi448/Rvfu3ZXWpwpVocsUOYeyoKOjg8mTJwsZNaQZMhRtUQkMDBT+XVBQgOvXr8Pf37/U+8CbdFnKokWL4Ofnh6ysLJm4wYpWmaampsLOzg5JSUmwt7dHw4YNERAQgEGDBsnVp6eno0+fPsKPFjMzM5w4cQLt2rWT2dKjjGbNmmHEiBFYsWKFSm1NgoZMCUJN6tevDwcHB3To0AE5OTnCPrZ3Kenw9PT0YG5uLjNkp4ht27Zh37596N27N+7evatQd+LECUyYMAFXr15FixYtEBMTgz179mDAgAFwdnbGd999V6pMUVERGGPYu3cv/vjjD7x69UqNO4bgdAcPHow1a9YgMDAQY8eOVausMlauXKnwnDznoMjpKHM+DRo0wHfffYeWLVsiLS0Nfn5+yM3NVdk2dfNR8iZdljJ8+HC0aNFC7a0rvPOsM2fOxNmzZ3Ht2jWIRCJYWVnhiy++gIGBgRD3Vh716tXDsGHD4OPjA1NTU6WZbmoyGl/ZQ0ZWlc3Pz48dP36chYWFscmTJ7MGDRqoVW7AgAEsMzOTZWVlMaB428bRo0ffuz2pqanCvxctWsR+++03BhRvLSh57l2zsbFhCxcuZAkJCWzXrl2sX79+KrdShIeHs2+++YZdvnyZNWrUiInFYqXXqEp2/Phx9uOPP7LevXuztWvXssDAQLXKWVlZsWPHjrEHDx6w+/fvs99//51ZWVmVW7sOHjyo9mcIAFu/fj0zNjZmn332Gfvnn39YYmKiynvR0dFhtra2zM7OTunWHUNDQzZ69Gh2/Phxdu3aNbZixQp269Ytjb93GjSNN4CMrEpbYWEhS05OZseOHWPHjh1jR48elTFF5eLj45mRkZFMwOSS2x2ktn//fgYUO7qUlBTBpK/f1Zc8dvbsWTZ48GDhtaptF1IbPnw4e/jwIZs1a5ZSXaNGjdiMGTNYt27dGABmZmbGRo8eXW7PVl9fn82dO1fYOtGyZUv24YcfKi3j6urK/Pz8GABWr149meDaJe3drQzKtkKUtJiYGDZq1CimpaXFtLS02MiRI9n58+fL7Z4dHR1ZTk4OO3HihNLP0eDBg0s5TgsLC9ahQwe1numWLVtUPtO8vDwWHh4uvL+AegHKa6rRkClBqOCDDz4oU7mCggKFgb9LIo0gou7q19TUVPzyyy+4ffs2WrZsiZMnTwKAygS2TZs2xYgRI/DRRx/h6dOnmDFjhsphsfv378tsK7l165bKKDI8bNu2DQkJCcKClZycHAQHB+Ovv/6Sq58/fz6cnJxgY2OD7du3Q0dHB7t375abG1AkEskEDtfS0pJ5rSjkm0gkwu7du4XXe/bswZQpU97rPkuyY8cOLFu2DGlpaUqz0Y8aNQrr169HXl4eoqOjce7cOURHRytddAT875l26dIFgPJn+t1332HEiBHYuHEj9u7di/3797/fzdUANO6Vyciqi/GkZwoICGA+Pj4sJSWFtWzZkq1du5Zt3LixXNowZ84ctnr1aplIOV26dGGjRo2SWyY8PJwlJCSwOXPmsBYtWqjcaB8VFcUA2U3qz58/F16X1/OUBgoo2YtW1suV9vpU9boBsOzsbHbt2jXuje0///wzmzNnDrOwsGDm5uZs9uzZbN68eeUWlEBVKqZ3zdLSkvn4+LB169ax+Ph49uDBA/bXX3+V2zMFioeJv/vuO5aamspevXrFvv76a2ZtbV1u73M1Mo03gIysWhjvnKC6odjedTo8zkddB52dnS3jDEo6Ck0OkUVHRzM9PT1hOLN58+bswoULCvXSc1K9gYGBQodYVpPnQMvzWa1cuZItWbKEde7cmTk4OAimrIyNjQ0bP348CwgIYBcvXmShoaHl8kylkXJKWvv27dlPP/3Erl69qrHPhQZN4w0gI6sWpu6coNR4Q7HxWkUt2qlM8/T0ZOHh4ezBgwds9+7dLDs7m3l4eCjUz5w5k23atIldu3aNffLJJywmJoZ9+eWXSq/RtWtXZmBgwACwkSNHspUrV3IlSS5vCw0NLWVnzpwppfv222/ZsWPHWExMDNu+fTv7/PPPmb29PROLxeX2TKVOc+fOnRr/LFQR03gDyMiqhUkXVqjrEENDQ1lGRgZbtGgRa9u2Lde1DAwMmK+vL/vzzz8VauQ5aFUrQKuacwDA6taty/r3788+/PBDVq9ePWZubq5U7+npyZYvX85++eUX5unpqbJ+3uDb2tra7Msvv2TBwcEsODiYTZ48udwDgqtjGRkZLC4uji1YsID16tWLGRkZlfmZKtKlpaWxMWPGsKtXr7KPPvqolGnyc6EJo0U1BKEmFy9ehI+PD7S0tNCyZUtMnToV586dU6jnDcUmkUjQv39/+Pr6om/fvjh06BA2bdqksH55i3ZUhY3buHGjsNH+66+/xtatW7Fr16733mhfFjp37gxTU1NERkbi77//RocOHbB27Vq4ubnJjcIjFotRp04dnD59GqdPn4ZEIoGfnx/S09PRtm1bhdfh3Uu5ceNGSCQSbNiwAUBxuLuNGzdi4sSJ73nHxRgbG2PMmDGwtLSUiVTzbnqmNm3aoE6dOujatSu6d++Ob775BoaGhkhJScG5c+ewfft2Gf27wQ6ke1nNzc1hbm4uN9jB559/jpEjR8LExKRUQmLG2H9yL6LGvTIZWXUw3vRMJa19+/Zs586d7M2bN6XOeXp6sq1bt7KcnBy2a9cuNmDAAJadna2yzrIs2uHNzFBRtnz5cpaens727t3LYmNj2fz589m9e/fY1KlT5T5Tb29v9uzZM3b79m0WHh7Ounfvzm7dusUOHz6scv6Ndy9lWbKI8Fh0dDRbuXIl8/PzUzuzhJaWFnNxcWGzZs1iV65cYQUFBaU08oZilQ3JljTpZ4FM8w0gI6sWxjsn2Lp1a7ZgwQKWlpbGwsPD2eeffy53Q3ZhYSELDw+X2U+nzuKNsjjoqrLR/tKlS0JbTUxMWF5eHmvZsqVCfVpaGmvRogUDwBwcHNjr16/ZkCFD1LoW717KhIQE1rx5c+G1lZVVuf5oULeugQMHsp9//plFRkayx48fs8jISLZ06VI2aNAgVr9+/XJrT4MGDdgPP/zAgoOD2YEDB9jChQu5AgfUJKNYpgShJqGhoVzpmWJiYrBv3z4EBwcrDcVmb2+PESNGYOjQocjKykJQUBDmz58PS0tLpfXb29sjOTmZ6x4aNWoEX19fxMXF4ezZszAzM0P37t3LdW+hOsTHx8ukX0pKSlIa3zQhIQGOjo7C64yMDLRp06ZC2tajRw9s27YNWVlZAABLS0uMGzcO4eHh5VL/9OnT8fLlS/z5558yqZje3Rd56NAhYe9hQkKCylBvJYOSy0Pe8GfXrl2xd+9ebN++HQkJCRCJROjYsSPGjh2LkSNHKp0SqImQQyQIDqRzgt7e3irnBMtC165d4ePjAy8vLyQnJ+PIkSPw9/eXq+V10FUJaUYQKe7u7jKv380IcuvWLfz666/C66+++krm9bs5KYGyB9/W1dXFzJkzhXiqp06dwqpVq2Sc1/vwxRdf4KeffsKzZ8+EdsnLMSn9EbBz506ZlE2KKBmU/F0YY5gwYUKp4zExMZg0aVKpH1Z2dnbYvHkzOnfurMYd1RzIIRJEGVCWnmn//v3w9vYW8hJKkX4RK8osURKRSARPT0/4+PgoDTatroMuq3OoKHgzgsyfP1+pftGiRe/dJin79+/HixcvsGfPHgCAj48P6tSpg+HDh5dL/VevXkWnTp3w+PFjpbq0tDT88ssvmD9/PmbPnl3qPM+Cl4YNG8oNRn/p0iW0a9dObhll52oyGh+3JSOrDqbunGDjxo0ZAGZubi7XFNUvb0uEqi0IUlO2aIeMzyp6Uc3Ro0eZvr6+Sp2rqyvbsGEDe/ToEQsMDJSxrVu3qixvZGTExo0bx06dOsVycnLkatLT05mJiUmp43Xq1GEZGRkafy8q22jbBUGoibrpme7duweAP0GvvC0RO3fuVLglQl7+xJkzZ3JdU1O823t+l3d70WvWrFFa37tbFt6HpKQkdOrUCRcuXAAAuLi4IDo6utzqLywsRHJyMsLCwmSGYd+9h+joaERHRyM+Pl7pcGhJ9PT0MGjQIPj6+qJjx46oXbs2hgwZojAH56pVq3Dy5EnMmjULiYmJAABHR0csW7ZM7jB0TYccIkGoiTRYsireHZaUomp4kne/nLoOuioiDWQ+efJkAJDJZp+Xl1dKn5CQAABwdXVF27ZthSDUw4YNE86VF506dcKYMWOEHzTm5ubIyMgQnLg6Q97K+P333/H777+rpW3QoAEsLCwQHBwMxhjS09Oxfv36UsmKAWD37t1wd3fHyZMn8dtvvyE0NBRXr15FRESEwvr9/f1x584dLF68WBgevXTpEn788Uf8+eefZbq/6gzNIRKECspjTlAdwsPDceLECYwbNw7u7u54+PAhkpOTYWtrWy71V0XOnj1bKlOFvGNSQkND0bt3b+HHg7a2Nk6ePIkePXqUW5vkBQUoCW/PXx4SiQStWrUCAFy+fFm4n5LwrgBNTk6GSCTCzp07sX//fuTk5ODatWulFusQytH4uC0ZWVW2ss4JljR1QrGpu1+ON39iVbakpCTm6uoqvO7SpUupPIYlLTMzUybjhImJCcvMzNT4ffCYh4cHu379OgsPD2cREREsKyuLubm5ldLFxMTI3ftqZ2enMD+jjY0N++GHH1hmZiaLjIxkDx48YI0aNVLZJmtra3b69GmWlpbGALAOHTqwuXPnavxZacA03gAyshppEomEDR48mO3fv589f/6cBQYGsgEDBrx3veXhoKuKdezYkSUnJ7Ps7GyWnZ3NkpKSlEae8fPzY9evX2fbtm1j27ZtY1lZWSqjvFQ1i4+Pl8lQYm1tzeLj40vpLl26pLAOZeek5uTkxFasWMFu3LjBoqOjlWrDw8OZs7OzTFxcqXP8j5nGG0BGVqWNNz0Tbyi2yso9WJWtdu3aagevbtSoERs0aBAbNGiQWr2fqmbyevHyjpXnClB3d3el52NjYxkgG7heWU+9photqiEIFRgZGXHpQ0JCEBUVhW7duuH69esAlK+SdHNz47pOWRftVFX69++Pdu3aQU9PTzi2ePFihfo3b97g7t270NPTQ6tWrdCqVStERUVVRlPLhfj4eAQEBAgLiUaNGiV3YVBZV4BaW1tj48aNaNSoETp06IAOHTrAzc1N4UpTAHj06BGaN28ufK68vLyq3UKt8kLjXpmMrLqZsjlBe3t7tnTpUnb16lV28uRJNn78eHb9+nWNt7kq2saNG9mOHTvYzZs32fz581lqaioLCAhQqJ8wYQJLTU1lT548YaGhoSwvL09l4OqqZjo6OmzGjBns0KFD7PDhw2zatGlyE/UCYB9++CGLiIhgjx49Yo8ePWIREREqh93LMvxpZWXFTp06xf7991+Wk5PDoqKimIWFhcaflQZM4w0gI6sWVpY5wa5du7J169axO3fusL///ptNnDixQtqmzqKdqmjSoULp/2vVqsVCQkIU6lNTU5murq4wnGdjY8OCgoI0fh/qWP369VmbNm1KHW/Xrl25But+n+FPAwMDZmhoqPFnpSkTgyAIpXh6emLr1q3Izs7G0KFDsWvXLjx58gTjx49XuVfr3Llz+PLLL2FqaopVq1apvZdRHSQSCQYPHoz9+/fj7t278PT0VJo/sSry6tUrAEBeXh6aNGmC/Px8WFlZKdS/fv1a2Myuo6ODy5cvw8bGplLa+r6sW7cODRo0KHXc1NRU6ZC6tbU1Tp8+jbS0NABAhw4dMHfuXIX6sgx/Tp06FbVr10ZeXh5WrVqFhIQE9OrVS53bqnFo3CuTkVVlK2t6pvcJxabMypo/sSravHnzmLGxMfv444/Z3bt32Z07d9iiRYsU6g8fPsyMjY3ZggULWEREBPv999/ZX3/9pfH7UMcuXryo8JyyIU3eIVB5w5+qPnfS0HS9e/dmR48eZba2thrJk1kFTOMNICOr0lbWOUHpMKCtrS1LSUlhU6dOZeHh4e/dnrI66KpuOjo6aq80BYpXTg4cOJBJJBKNt10du3z5ssJzyvZS8gyBisVitnz5cgbwDX9KP6urV68W8kyWvN5/xWjIlCBUkJycjG+++QYtW7bEwoUL4eDgAB0dHfz999+YOHGiwnLvhmJbu3Ytateu/d7tcXR0xPnz53H69GmcPHkS48ePh5aW1nvXqwn09fUxb948bNmyBW/fvkXDhg3x4YcfKi3j6uoKPz8/REZGIiYmBqamppXU2vfjypUr6NevX6njffv2FXIvyoNnCLSoqEjIG5mXl4eXL1+q1baEhASEhISgf//+CAkJgaGhIYqKitQqW9PQuFcmI6tuJhKJWK9evVhgYKBCTWVkp6+sRTsVZUFBQWz27NnCEKCenp7SBSDz589nx44dE3pbTZo0YWfPntX4fahj1tbW7PLly2zbtm1sypQpbMqUKWz79u3s8uXLzNraWmE53hWgK1asYEePHmWjRo1iH330kWDK2iYSiZiDgwMzNjZmAFjdunVZhw4dNP7MKtsolilBqEnXrl2RnJyMvLw8jBw5Eh07dsSaNWsUxraszOz06uZPrGrExcXB2dkZiYmJ6NixI4DiHrm9vb1cfVJSEhwcHGT0KSkp5RZPtqLR0dGBr68v2rdvD6A4kPbevXvVSj5sYGAAsVisstcnLzOGogTBUqR7Yd+lOu3vLC807pXJyKqDVdScYFmtohbtVKZFR0czPT09YQFH8+bN2YULFxTqpeekegMDg2oXv7WkqbNdZurUqax27doMAPP392cJCQmsV69e5dqOY8eOCXby5En27Nmzare/s5xM4w0gI6sWJv0S/v7779n48eNljpW0ygrFVtUcdFnM09OThYeHswcPHrDdu3ez7Oxs5uHhoVA/c+ZMtmnTJnbt2jX2ySefsHPnzrEpU6Zo/D54jHc/K+8KUFNTU3b48GF2//59du/ePXbw4EFmamrK1cZmzZqxvXv3avxZacA03gAysmphlTEnyGPqOuiqbnXr1mX9+/dnH374IatXr55KvaenJ1u+fDn75ZdfmKenp8bbr66VdbsM7wrQkydPMj8/P6alpcW0tLTY2LFj2cmTJ7nbq8nPtqaM5hAJQk0qc05QHap7/kQtLS3069cPrVu3BgBkZGTgxIkTKCwsVFjG2NgY1tbWAIB//vkHL168qJS2lgeFhYWIioqCn5+fEONWnXyFgYGBMDU1hZWVFezs7KClpYXw8HA4OTnJ1UvnWVUdK8natWuFVaxisRj29va4fv06Ro8ezXGH1R9yiARRTalqDpqHJk2aICwsDHfv3kVSUhJEIhEcHBzQuHFjfPDBB6W2FUgkEmzZsgVDhgxBVlYWxGIxLCwscOTIEXz++efIz8/X0J2oj729PUaMGIGhQ4ciKysLQUFBmD9/PiwtLZWWE4lEsLe3R1ZWFp4/f466devC1NRUiFzzLqdOncL27duxb98+AICPjw/GjRsHT09PhdcYM2aM8O+CggJcv369VALi/woa76aSkVVlo/RM5W/btm1j06ZNK3X8yy+/ZNu3by91/IcffmC7d++W2WhuaGjIduzYoTSyTVU1nu0ybm5uck2R3szMjB09epQ9ePCA3b9/nx05ckStxVYSiYS1a9eOtWvXjmlra2v8GWnINN4AMjIyDqsJDlpZPj95UVvS0tKYvr5+qeO1atWq1ols1dnPyrsCVFdXl7sdHh4e7Pr16yw8PJxFRESwrKwspU63BpvGG0BGRvYfM2WLQuSdU7a1orot/njf7TKqVoBeuXKFnT17lv3888+sX79+aoXDi4+PZ61atRJeW1tbs/j4eI0/q8o2ShBMEESlY2xsjI8++qjUcZFIJDdRMmMMJiYmEIlEpc5VtxBjGzduhJ2dHWxtbfH1119j69at2LlzJ7p3765W+ZycHGFjvzysra1hZmYGNzc3DBgwABs2bMCzZ8+ULqqRSCT4559/hNdXrlyBRCJR+55qCuQQCYKodCIiIjBw4EC55+Rldjc2NkZCQoJchyhdHVldeDfGbWBgIMaOHatQL28FaEpKikK9qakpXF1d4ebmBjs7O1y6dAlnz55V2qb4+HgEBAQIC7JGjhyJhIQE3lur9tAqU4IgiEqEd7sM7wrQwsJCxMXFYcmSJTh27JhabdLR0cHkyZPRrVs3iEQiREZGYsOGDXj79i3fzVVzyCESBKExjI2NMWbMGFhaWkJb+38DVtOmTZPRKRvuA4r32VUXyrJdRiKRoFWrVgCAy5cvC71Medja2qJbt25wd3eHubk5rly5goiICLkxTglZyCESBKExoqOjcf78eaSlpcnMBe7cuVNGFxoaqrAOxhh69uxZYW3UNB4eHtixYweuX78OkUgEMzMzjB07Vmng7Vq1aqFbt25wc3PDqFGjwBiDlZVVKV1qaqrSIefqEjS9vCCHSBCExkhISBDy99V0oqKi4ObmhhcvXsg4IZFIBMYYjI2N5ZaLj4+Hr6+vsOjF2toa+/btUxipJi4uDrq6ujh37hzOnj2LyMhIhRlZzM3NhTb89ddf6N+/v8x5ReVqKuQQCYLQGNOnT8fLly/x559/yqRAevr0qVy9trY2Jk2aBHd3dwDF83GbN29WOoRY3ZGX3kpZyqv69evj0aNH3Nf5L/04UYRY0w0gCOK/y9u3b/HLL78gJiYGCQkJSEhIQHx8vEL9xo0b4ejoiA0bNmDDhg1wdHTExo0bK7HFlY90BaiHhwc8PDywZcsWuStABwwYAHNzc8EZfv/990hOTsbRo0dVhocj/ofGN0OSkZH9N+3q1atqZbiQmjQVkqpjNcl0dHTYjBkz2KFDh9jhw4fZ9OnTmY6OTildSkqKEM3nww8/ZJcvX2YdO3ZkEyZMYCdOnJBbt4ODg2Dp6enM3t5e5pim772yjYZMCYLQGEePHsWIESPw6tUrtfQJCQkYNmwYsrKyAABWVlY4ePDgf36oDwCSk5Nhb28PANi6dSsuX76M5cuXA1A8HPpfXqwkD9qYTxCExigsLERycjLCwsJk5hDf3XYhZfbs2QgLC0NWVhZEIhEsLCwwbty4ympupcK7AlQkEqFWrVrIy8tDz549sWHDBuGcnp6e3Dp69OhRPo2tIZBDJAhCY/z+++/4/fff1daHhobC2toaNjY2EIlEyMzMrLGbxwcMGABA8QrQd1m9ejWSk5Px4sULZGRkCPOM9vb2pdJpyaNLly6l9oNWh1Ri5QkNmRIEUW34L64yBdRfAdq0aVM0bNgQKSkpQu+ycePGkEgkuHXrlsJyO3fuRIsWLZCcnCwkaGaMKeyp11TIIRIEoTGysrLkDgsqyiLv7+8PiUSCHTt2AABGjx6NwsJCTJw4sULbqWl4t0SMHDkSzZs3x+LFi2FmZobGjRsjLi5OoT49PR1t27Ytj6ZWa2jIlCAIjVFyc7menh6GDRuGunXrKtQ7OzsLC0cAICwsDMnJyRXYQs1RMlydvr4+7O3tZYKbKwpXt2HDBhQVFaFHjx5YvHgxcnNzcejQIbi4uCi81sWLF9G4cWPcu3ev/G6gGkIOkSAIjfHkyROZ12vWrEFUVBQWLFggV19YWIjmzZvLrDKVDvHVNFauXCn8+969e/j111+F18pWgHbq1AmOjo5ITEwEADx79gw6OjpKr1W/fn2kp6cjNjZWZnHT4MGD3+cWqh3kEAmC0Bgle0FisRhOTk6oXbu2Qn3JVaYAYGlpWWNXmZZ1BWh+fj7EYrEwFF2/fn2VOSMXLlxYpmvVNMghEgShMVauXCl8cUtTGw0bNkyhPjo6Gps3bxZ6R5s3b0ZMTEyltFWT8KwAXbt2LY4cOYKGDRvixx9/xNChQzFv3jyl9cvLQflfhBbVEAShMXR1deHl5SXzZc8Yw+LFi+Xq9+/fjxcvXmDPnj0AAB8fH9SpUwfDhw+vtDZXNmVZAWpjY4OePXtCJBLhzJkzyMzMlKsra8Dxmgo5RIIgNMbx48fx7NkzJCYmyswFlpwvK0nJaCzKjtUkeFeANm/eHDk5OXj79i08PDxga2uLnTt34vnz5xXYypqDxuPHkZGR/TctLS2NS79t2zbWqVMn4bWLiwtbv369xu+jIu3AgQOscePGauuTkpKYlpYWa9GiBbty5Qr79ddf2V9//aXx+6gORnOIBEFojHPnzqF9+/a4ePGiUp00jJlEIsGYMWNw8+ZNMMZgYWGB9PT0SmqtZuBdAVpUVITCwkJ8/PHHWLNmDX777TdhxSmhHHKIBEFojG7dusHPzw/Z2dl48+aNMHf1bpxOaRiz/yK8K0Dz8/MxYsQIjBkzBgMHDgQASCSSCmhZzYPmEAmC0BjSjO3v8l/L1F6etGnTBp9//jliYmIQFBQES0tLeHt7Y9myZZpuWpWHHCJBEEQVhFaAVj7kEAmCIGoQvPFhif9Bc4gEQRA1CN74sMT/oB4iQRBEDUc6/Eooh3qIBEEQNQje+LDE/yCHSBAEUYMomSVDGh+2Joe2K09oyJQgCIIgAIg13QCCIAii/DAyMsLKlSsRFxeHuLg4rFixAkZGRppuVrWAHCJBEEQNIjAwELm5uRg+fDiGDx+OFy9eYNu2bZpuVrWAhkwJgiBqEElJSTILaxQdI0pDPUSCIIgaxKtXr+Dq6iq87tq1K169eqXBFlUfqIdIEARRg7Czs8OOHTuE0G5Pnz7F2LFjkZaWpuGWVX3IIRIEQdQgdHR0MHToULRo0QImJiZ4/vw5GGNYvHixpptW5aF9iARBEDWIo0eP4tmzZ0hMTMTt27c13ZxqBfUQCYIgahBpaWno0KGDpptRLaFFNQRBEDWIc+fOoX379ppuRrWEeogEQRA1gNTUVDDGoK2tDWtra2RlZeHNmzdC/kQ7OztNN7HKQw6RIAiiBmBubq70/M2bNyupJdUXcogEQRAEAZpDJAiCIAgA5BAJgiAIAgA5RIIgCIIAQA6RIAiCIACQQyQIgiAIAMD/AV7Wri20SmfBAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Heatmap of numerical feature in the training dataset\n",
    "num_features = new.select_dtypes(['int', 'float'])\n",
    "corr_mat = num_features.corr()\n",
    "sb.heatmap(data=corr_mat)"
   ]
  },
  {
   "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>features</th>\n",
       "      <th>VIF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>isAliveFather</td>\n",
       "      <td>165.111250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>isAliveMother</td>\n",
       "      <td>93.203589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>isAliveHeir</td>\n",
       "      <td>77.020836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>isAliveSpouse</td>\n",
       "      <td>28.026185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>isMarried</td>\n",
       "      <td>15.455632</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>age</td>\n",
       "      <td>10.098704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>popularity</td>\n",
       "      <td>7.642320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>isPopular</td>\n",
       "      <td>4.347781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>houseSworeAllegiance</td>\n",
       "      <td>3.579165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>book4</td>\n",
       "      <td>3.242501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>numDeadRelations</td>\n",
       "      <td>3.237888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>boolDeadRelations</td>\n",
       "      <td>3.036974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>book3</td>\n",
       "      <td>2.898499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>male</td>\n",
       "      <td>2.783317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>book2</td>\n",
       "      <td>2.480204</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>isNoble</td>\n",
       "      <td>2.156332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>book5</td>\n",
       "      <td>2.125849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>inHouseOfWarLead</td>\n",
       "      <td>2.057726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>inHouseOfAllegnc</td>\n",
       "      <td>2.013128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>book1</td>\n",
       "      <td>1.713839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>isACommander</td>\n",
       "      <td>1.291378</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                features         VIF\n",
       "7          isAliveFather  165.111250\n",
       "6          isAliveMother   93.203589\n",
       "8            isAliveHeir   77.020836\n",
       "9          isAliveSpouse   28.026185\n",
       "10             isMarried   15.455632\n",
       "12                   age   10.098704\n",
       "16            popularity    7.642320\n",
       "15             isPopular    4.347781\n",
       "19  houseSworeAllegiance    3.579165\n",
       "4                  book4    3.242501\n",
       "13      numDeadRelations    3.237888\n",
       "14     boolDeadRelations    3.036974\n",
       "3                  book3    2.898499\n",
       "0                   male    2.783317\n",
       "2                  book2    2.480204\n",
       "11               isNoble    2.156332\n",
       "5                  book5    2.125849\n",
       "18      inHouseOfWarLead    2.057726\n",
       "20      inHouseOfAllegnc    2.013128\n",
       "1                  book1    1.713839\n",
       "17          isACommander    1.291378"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
    "x = num_features.drop(columns='isAlive')\n",
    "vif = pd.DataFrame()\n",
    "vif['features'] = x.columns\n",
    "\n",
    "vif['VIF'] = [variance_inflation_factor(x.values, i) for i in range(len(x.columns))]\n",
    "vif.sort_values(by='VIF', ascending=False)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are multicolinearity in the features. Let's drop some columns with multicolinearity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping columns with colinearity\n",
    "new = new.drop(['isAliveFather', 'isAliveMother', 'isAliveSpouse'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>male</th>\n",
       "      <th>mother</th>\n",
       "      <th>father</th>\n",
       "      <th>heir</th>\n",
       "      <th>book1</th>\n",
       "      <th>book4</th>\n",
       "      <th>age</th>\n",
       "      <th>numDeadRelations</th>\n",
       "      <th>boolDeadRelations</th>\n",
       "      <th>isPopular</th>\n",
       "      <th>popularity</th>\n",
       "      <th>isAlive</th>\n",
       "      <th>inHouseOfWarLead</th>\n",
       "      <th>houseSworeAllegiance</th>\n",
       "      <th>inHouseOfAllegnc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.605351</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.896321</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.267559</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.183946</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.043478</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1941</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.351171</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1942</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.096990</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1943</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.030100</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1944</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.130435</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1945</th>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>47.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.210702</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1946 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      male  mother  father  heir  book1  book4   age  numDeadRelations  \\\n",
       "0        1       0       0     0      0      0  37.0                11   \n",
       "1        1      -1      -1    -1      1      1  97.0                 1   \n",
       "2        1      -1      -1    -1      0      1  37.0                 0   \n",
       "3        0      -1      -1    -1      0      0  23.0                 0   \n",
       "4        0      -1      -1    -1      0      1  29.0                 0   \n",
       "...    ...     ...     ...   ...    ...    ...   ...               ...   \n",
       "1941     1      -1      -1    -1      1      1  37.0                 0   \n",
       "1942     1      -1      -1    -1      0      1  37.0                 0   \n",
       "1943     1      -1      -1    -1      1      1  37.0                 0   \n",
       "1944     1      -1      -1    -1      0      1  37.0                 0   \n",
       "1945     1      -1      -1    -1      0      1  47.0                 4   \n",
       "\n",
       "      boolDeadRelations  isPopular  popularity  isAlive  inHouseOfWarLead  \\\n",
       "0                     1          1    0.605351        0                 0   \n",
       "1                     1          1    0.896321        1                 0   \n",
       "2                     0          0    0.267559        1                 0   \n",
       "3                     0          0    0.183946        0                 0   \n",
       "4                     0          0    0.043478        1                 0   \n",
       "...                 ...        ...         ...      ...               ...   \n",
       "1941                  0          1    0.351171        0                 0   \n",
       "1942                  0          0    0.096990        0                 0   \n",
       "1943                  0          0    0.030100        1                 0   \n",
       "1944                  0          0    0.130435        1                 0   \n",
       "1945                  1          0    0.210702        0                 0   \n",
       "\n",
       "      houseSworeAllegiance  inHouseOfAllegnc  \n",
       "0                        0                 0  \n",
       "1                        0                 0  \n",
       "2                        0                 0  \n",
       "3                        0                 0  \n",
       "4                        0                 0  \n",
       "...                    ...               ...  \n",
       "1941                     0                 0  \n",
       "1942                     0                 0  \n",
       "1943                     0                 0  \n",
       "1944                     0                 0  \n",
       "1945                     0                 0  \n",
       "\n",
       "[1946 rows x 15 columns]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "character = char_pred.copy()\n",
    "def transform_data(data):\n",
    "    # Replacing the age\n",
    "    data.loc[1684, 'age'] = 25.0\n",
    "    data.loc[1868, 'age'] = 0.0\n",
    "    \n",
    "    # Replacing missing values\n",
    "    data[\"age\"].fillna(round(data[\"age\"].mean()), inplace=True)\n",
    "    data['dateOfBirth'].fillna(data['dateOfBirth'].median(), inplace=True)\n",
    "    data['DateoFdeath'].fillna(data['DateoFdeath'].median(), inplace=True)\n",
    "    \n",
    "    data.loc[:, \"culture\"] = [get_cult(x) for x in data.culture.fillna(\"\")]\n",
    "    \n",
    "    data.loc[:, \"mother\"] = pd.factorize(data[\"mother\"])[0]\n",
    "    data.loc[:, \"father\"] = pd.factorize(data[\"father\"])[0]\n",
    "    data.loc[:, \"heir\"] = pd.factorize(data[\"heir\"])[0]\n",
    "    data.loc[:, \"house\"] = pd.factorize(data[\"house\"])[0]\n",
    "    data.loc[:, \"spouse\"] = pd.factorize(data[\"spouse\"])[0]\n",
    "    data.loc[:, \"culture\"] = pd.factorize(data[\"culture\"])[0]\n",
    "    \n",
    "    data.loc[:, \"isAliveMother\"] = pd.factorize(data[\"isAliveMother\"])[0]\n",
    "    data.loc[:, \"isAliveFather\"] = pd.factorize(data[\"isAliveFather\"])[0]\n",
    "    data.loc[:, \"isAliveHeir\"] = pd.factorize(data[\"isAliveHeir\"])[0]\n",
    "    data.loc[:, \"isAliveSpouse\"] = pd.factorize(data[\"isAliveSpouse\"])[0]\n",
    "    # cat = ['title', 'mother', 'father', 'heir', 'house', 'spouse']\n",
    "    # for col in cat:\n",
    "    #     data[col].fillna('unknown', inplace=True)\n",
    "    # data.fillna(-1, inplace=True)\n",
    "    \n",
    "    data['isACommander'] = np.where(data['name'].isin(all_commanders), 1, 0)\n",
    "    data['inHouseOfWarLead'] = np.where(data['house'].isin(renam_warlead), 1, 0)\n",
    "    data['houseSworeAllegiance'] = np.where(data['house'].isin(hus_char_with_allegnc), 1, 0)\n",
    "    data['inHouseOfAllegnc'] = np.where(data['house'].isin(renam_husOfAllgnc), 1, 0)\n",
    "    \n",
    "    data = data.drop(['S.No', 'actual', 'pred', 'alive', 'plod', 'name', 'title',\n",
    "                   'house', 'isAliveFather', 'isAliveMother', 'isAliveSpouse', 'isAliveHeir'], axis=1)\n",
    "    \n",
    "    return data\n",
    "\n",
    "def select_features(data):\n",
    "    # Heatmap of numerical feature in the training dataset\n",
    "    num_features = data.select_dtypes(['int', 'float'])\n",
    "    corr_mat = num_features.corr()['isAlive'].abs().sort_values()\n",
    "    corr_less = corr_mat[corr_mat < 0.1]\n",
    "    data = data.drop(corr_less.index, axis=1)\n",
    "    \n",
    "    return data\n",
    "    \n",
    "da = transform_data(character)\n",
    "de = select_features(da)\n",
    "\n",
    "de"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "book3                   0.006693\n",
       "inHouseOfWarLead        0.029822\n",
       "book5                   0.032846\n",
       "isNoble                 0.042211\n",
       "isMarried               0.050037\n",
       "houseSworeAllegiance    0.052159\n",
       "isACommander            0.057968\n",
       "book2                   0.067200\n",
       "Name: isAlive, dtype: float64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Selecting features with coefficient greater than 0.1\n",
    "num_features = new.select_dtypes(['int', 'float'])\n",
    "corr_mat = num_features.corr()['isAlive'].abs().sort_values()\n",
    "corr_less = corr_mat[corr_mat < 0.1]\n",
    "corr_less"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "new = new.drop(corr_less.index, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8135095447870778"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = de.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "lgr = LogisticRegression(random_state=42, max_iter = 1000)\n",
    "lgr.fit(train1[features], train1['isAlive'])\n",
    "predlgr = lgr.predict(test1[features])\n",
    "lgr.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.684931506849315"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgr.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# List of categorical variable\n",
    "nominal_features = ['title', 'culture', 'mother', 'father', 'heir', 'spouse']\n",
    "\n",
    "# Counting unique values in each column\n",
    "unique_count = new[nominal_features].apply(lambda x: len(x.value_counts())).sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Arbitrary cutoff of 25 unique count\n",
    "unique_count = unique_count[unique_count > 25].index\n",
    "# Dropping \n",
    "new = new.drop(unique_count, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Converting the remaining text column to categorical type\n",
    "txt_col = new.select_dtypes(include=['object'])\n",
    "for col in txt_col:\n",
    "    new[col] = new[col].astype('category')\n",
    "    \n",
    "# Creating dummy columns from txt columns and add back to dataframe\n",
    "new = pd.concat([\n",
    "    new, pd.get_dummies(new.select_dtypes(include=['category']))\n",
    "], axis=1).drop(txt_col, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8127753303964758"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = new.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "lgr = LogisticRegression(random_state=42, max_iter = 1000)\n",
    "lgr.fit(train1[features], train1['isAlive'])\n",
    "predlgr = lgr.predict(test1[features])\n",
    "lgr.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6952054794520548"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgr.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier \n",
    "\n",
    "from sklearn.model_selection import cross_val_score, cross_val_predict, GridSearchCV, KFold, train_test_split \n",
    "from sklearn.metrics import roc_auc_score, accuracy_score, roc_curve \n",
    "from sklearn.metrics import confusion_matrix, classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8340675477239354"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = new.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "knc = KNeighborsClassifier(n_neighbors=6)\n",
    "knc.fit(train1[features], train1['isAlive'])\n",
    "predknc = knc.predict(test1[features])\n",
    "knc.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6815068493150684"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knc.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8281938325991189"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = de.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "knc = KNeighborsClassifier(n_neighbors=6)\n",
    "knc.fit(train1[features], train1['isAlive'])\n",
    "predknc = knc.predict(test1[features])\n",
    "knc.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6883561643835616"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knc.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 96 109]\n",
      " [ 77 302]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.55      0.47      0.51       205\n",
      "           1       0.73      0.80      0.76       379\n",
      "\n",
      "    accuracy                           0.68       584\n",
      "   macro avg       0.64      0.63      0.64       584\n",
      "weighted avg       0.67      0.68      0.67       584\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(confusion_matrix(test1['isAlive'], predknc))\n",
    "print(classification_report(test1['isAlive'], predknc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 93 112]\n",
      " [ 70 309]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.57      0.45      0.51       205\n",
      "           1       0.73      0.82      0.77       379\n",
      "\n",
      "    accuracy                           0.69       584\n",
      "   macro avg       0.65      0.63      0.64       584\n",
      "weighted avg       0.68      0.69      0.68       584\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(confusion_matrix(test1['isAlive'], predknc))\n",
    "print(classification_report(test1['isAlive'], predknc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 68  83]\n",
      " [ 68 365]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.50      0.45      0.47       151\n",
      "           1       0.81      0.84      0.83       433\n",
      "\n",
      "    accuracy                           0.74       584\n",
      "   macro avg       0.66      0.65      0.65       584\n",
      "weighted avg       0.73      0.74      0.74       584\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Creating array of values\n",
    "x = de[de.columns.drop('isAlive')].values\n",
    "y = de['isAlive'].values\n",
    "\n",
    "# Splitting data\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.3, shuffle=True, random_state=42)\n",
    "\n",
    "knn = KNeighborsClassifier(n_neighbors=6)\n",
    "\n",
    "# Import confusion matrix\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "# Fit the model to the training data\n",
    "knn.fit(x_train, y_train)\n",
    "\n",
    "# Predict the labels of the test data: y_pred\n",
    "y_pred = knn.predict(x_test)\n",
    "\n",
    "# Generate the confusion matrix and classification report\n",
    "print(confusion_matrix(y_test, y_pred))\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 59  92]\n",
      " [ 51 382]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.54      0.39      0.45       151\n",
      "           1       0.81      0.88      0.84       433\n",
      "\n",
      "    accuracy                           0.76       584\n",
      "   macro avg       0.67      0.64      0.65       584\n",
      "weighted avg       0.74      0.76      0.74       584\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Creating array of values\n",
    "x = new[new.columns.drop('isAlive')].values\n",
    "y = new['isAlive'].values\n",
    "\n",
    "# Splitting data\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.3, shuffle=True, random_state=42)\n",
    "\n",
    "knn = KNeighborsClassifier(n_neighbors=6)\n",
    "\n",
    "# Import confusion matrix\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "# Fit the model to the training data\n",
    "knn.fit(x_train, y_train)\n",
    "\n",
    "# Predict the labels of the test data: y_pred\n",
    "y_pred = knn.predict(x_test)\n",
    "\n",
    "# Generate the confusion matrix and classification report\n",
    "print(confusion_matrix(y_test, y_pred))\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8935389133627019"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = de.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "rdf = RandomForestClassifier(random_state=42)\n",
    "rdf.fit(train1[features], train1['isAlive'])\n",
    "predrdf = rdf.predict(test1[features])\n",
    "rdf.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9016152716593245"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = new.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "rdf = RandomForestClassifier(random_state=42)\n",
    "rdf.fit(train1[features], train1['isAlive'])\n",
    "predrdf = rdf.predict(test1[features])\n",
    "rdf.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.702054794520548"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rdf.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9016152716593245"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = new.copy()\n",
    "train1 = data1[:1362]\n",
    "test1 = data1[1362:]\n",
    "\n",
    "features = train1.columns.tolist()\n",
    "features.remove('isAlive')\n",
    "\n",
    "dt = DecisionTreeClassifier(random_state=42)\n",
    "dt.fit(train1[features], train1['isAlive'])\n",
    "pred_dt = dt.predict(test1[features])\n",
    "dt.score(train1[features], train1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6815068493150684"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.score(test1[features], test1['isAlive'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 1, ..., 1, 1, 0], dtype=int64)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separate our labels from our features\n",
    "x = new[new.columns.drop('isAlive')].values\n",
    "\n",
    "y = new['isAlive'].values\n",
    "y\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "seed = 42\n",
    "np.random.seed(seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data into 5 equal groups for validation\n",
    "kfold = KFold(n_splits=5, shuffle=True, random_state=seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's build different models to train with our data\n",
    "models = [LogisticRegression(max_iter=500, solver='liblinear'), RandomForestClassifier(n_estimators=400, random_state=seed), \n",
    "          DecisionTreeClassifier(random_state=seed), KNeighborsClassifier(n_neighbors=5)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Validate each model using K-fold cross validation\n",
    "mean=[]\n",
    "std=[]\n",
    "for model in models:\n",
    "    result = cross_val_score(model, x, y, cv=kfold, scoring=\"accuracy\", n_jobs=-1)\n",
    "    mean.append(result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "classifiers=['Logistic Regression', 'Random Forest', 'Decision Tree', 'SVM', 'KNN']\n",
    "\n",
    "plt.figure(figsize=(10, 10))\n",
    "for x in range(len(mean)):\n",
    "    sb.kdeplot(mean[x], shade=True)\n",
    "    \n",
    "plt.title(\"Distribution of each classifier's Accuracy\", fontsize=15)\n",
    "plt.legend(classifiers)\n",
    "plt.xlabel(\"Accuracy\", labelpad=20)\n",
    "plt.yticks([])\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    1451\n",
       "0     495\n",
       "Name: isAlive, dtype: int64"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "de.isAlive.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop = [\"S.No\", \"pred\", \"alive\", \"plod\", \"name\", \"isAlive\", \"DateoFdeath\"]\n",
    "# cols_dlt = ['actual', 'alive', 'name', 'plod', 'pred', 'isAlive', 'DateoFdeath', 'age', 'spouse', 'dateOfBirth', 'episodes', 'S.No']\n",
    "\n",
    "# preds_partial = preds[['S.No',\n",
    "#                       'male',\n",
    "#                       'dateOfBirth',\n",
    "#                       'book1',\n",
    "#                       'book4',\n",
    "#                       'numDeadRelations',\n",
    "#                       'boolDeadRelations',\n",
    "#                       'popularity',\n",
    "#                       'showOnly',\n",
    "#                       'houseDeathRate']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>S.No</th>\n",
       "      <th>actual</th>\n",
       "      <th>pred</th>\n",
       "      <th>alive</th>\n",
       "      <th>plod</th>\n",
       "      <th>name</th>\n",
       "      <th>title</th>\n",
       "      <th>male</th>\n",
       "      <th>culture</th>\n",
       "      <th>dateOfBirth</th>\n",
       "      <th>DateoFdeath</th>\n",
       "      <th>mother</th>\n",
       "      <th>father</th>\n",
       "      <th>heir</th>\n",
       "      <th>house</th>\n",
       "      <th>spouse</th>\n",
       "      <th>book1</th>\n",
       "      <th>book2</th>\n",
       "      <th>book3</th>\n",
       "      <th>book4</th>\n",
       "      <th>book5</th>\n",
       "      <th>isAliveMother</th>\n",
       "      <th>isAliveFather</th>\n",
       "      <th>isAliveHeir</th>\n",
       "      <th>isAliveSpouse</th>\n",
       "      <th>isMarried</th>\n",
       "      <th>isNoble</th>\n",
       "      <th>age</th>\n",
       "      <th>numDeadRelations</th>\n",
       "      <th>boolDeadRelations</th>\n",
       "      <th>isPopular</th>\n",
       "      <th>popularity</th>\n",
       "      <th>isAlive</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.946</td>\n",
       "      <td>Viserys II Targaryen</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Rhaenyra Targaryen</td>\n",
       "      <td>Daemon Targaryen</td>\n",
       "      <td>Aegon IV Targaryen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.605351</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.387</td>\n",
       "      <td>0.613</td>\n",
       "      <td>Walder Frey</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Rivermen</td>\n",
       "      <td>208.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Frey</td>\n",
       "      <td>Perra Royce</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>97.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.896321</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.493</td>\n",
       "      <td>0.507</td>\n",
       "      <td>Addison Hill</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Swyft</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.267559</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.924</td>\n",
       "      <td>Aemma Arryn</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>82.0</td>\n",
       "      <td>105.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Arryn</td>\n",
       "      <td>Viserys I Targaryen</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.183946</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.617</td>\n",
       "      <td>0.383</td>\n",
       "      <td>Sylva Santagar</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>Dornish</td>\n",
       "      <td>276.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Santagar</td>\n",
       "      <td>Eldon Estermont</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.043478</td>\n",
       "      <td>1</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",
       "      <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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1941</th>\n",
       "      <td>1942</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.803</td>\n",
       "      <td>0.197</td>\n",
       "      <td>Luwin</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>Westeros</td>\n",
       "      <td>NaN</td>\n",
       "      <td>299.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Stark</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.351171</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1942</th>\n",
       "      <td>1943</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.987</td>\n",
       "      <td>0.013</td>\n",
       "      <td>Reek</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>299.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Bolton</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.096990</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1943</th>\n",
       "      <td>1944</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.520</td>\n",
       "      <td>0.480</td>\n",
       "      <td>Symeon Star-Eyes</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.030100</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1944</th>\n",
       "      <td>1945</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.637</td>\n",
       "      <td>0.363</td>\n",
       "      <td>Coldhands</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Three-eyed crow</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.130435</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1945</th>\n",
       "      <td>1946</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.093</td>\n",
       "      <td>0.907</td>\n",
       "      <td>Tytos Lannister</td>\n",
       "      <td>37</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>220.0</td>\n",
       "      <td>267.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>House Lannister</td>\n",
       "      <td>Jeyne Marbrand</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>47.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.210702</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1946 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      S.No  actual  pred  alive   plod                  name  title  male  \\\n",
       "0        1       0     0  0.054  0.946  Viserys II Targaryen     -1     1   \n",
       "1        2       1     0  0.387  0.613           Walder Frey      0     1   \n",
       "2        3       1     0  0.493  0.507          Addison Hill      1     1   \n",
       "3        4       0     0  0.076  0.924           Aemma Arryn      2     0   \n",
       "4        5       1     1  0.617  0.383        Sylva Santagar      3     0   \n",
       "...    ...     ...   ...    ...    ...                   ...    ...   ...   \n",
       "1941  1942       0     1  0.803  0.197                 Luwin     -1     1   \n",
       "1942  1943       0     1  0.987  0.013                  Reek     -1     1   \n",
       "1943  1944       1     1  0.520  0.480      Symeon Star-Eyes     -1     1   \n",
       "1944  1945       1     1  0.637  0.363             Coldhands     -1     1   \n",
       "1945  1946       0     0  0.093  0.907       Tytos Lannister     37     1   \n",
       "\n",
       "       culture  dateOfBirth  DateoFdeath              mother  \\\n",
       "0          NaN          NaN          NaN  Rhaenyra Targaryen   \n",
       "1     Rivermen        208.0          NaN                 NaN   \n",
       "2          NaN          NaN          NaN                 NaN   \n",
       "3          NaN         82.0        105.0                 NaN   \n",
       "4      Dornish        276.0          NaN                 NaN   \n",
       "...        ...          ...          ...                 ...   \n",
       "1941  Westeros          NaN        299.0                 NaN   \n",
       "1942       NaN          NaN        299.0                 NaN   \n",
       "1943       NaN          NaN          NaN                 NaN   \n",
       "1944       NaN          NaN          NaN                 NaN   \n",
       "1945       NaN        220.0        267.0                 NaN   \n",
       "\n",
       "                father                heir            house  \\\n",
       "0     Daemon Targaryen  Aegon IV Targaryen              NaN   \n",
       "1                  NaN                 NaN       House Frey   \n",
       "2                  NaN                 NaN      House Swyft   \n",
       "3                  NaN                 NaN      House Arryn   \n",
       "4                  NaN                 NaN   House Santagar   \n",
       "...                ...                 ...              ...   \n",
       "1941               NaN                 NaN      House Stark   \n",
       "1942               NaN                 NaN     House Bolton   \n",
       "1943               NaN                 NaN              NaN   \n",
       "1944               NaN                 NaN  Three-eyed crow   \n",
       "1945               NaN                 NaN  House Lannister   \n",
       "\n",
       "                   spouse  book1  book2  book3  book4  book5  isAliveMother  \\\n",
       "0                     NaN      0      0      0      0      0            1.0   \n",
       "1             Perra Royce      1      1      1      1      1            NaN   \n",
       "2                     NaN      0      0      0      1      0            NaN   \n",
       "3     Viserys I Targaryen      0      0      0      0      0            NaN   \n",
       "4         Eldon Estermont      0      0      0      1      0            NaN   \n",
       "...                   ...    ...    ...    ...    ...    ...            ...   \n",
       "1941                  NaN      1      1      1      1      1            NaN   \n",
       "1942                  NaN      0      1      0      1      1            NaN   \n",
       "1943                  NaN      1      1      1      1      1            NaN   \n",
       "1944                  NaN      0      0      1      1      1            NaN   \n",
       "1945       Jeyne Marbrand      0      0      1      1      1            NaN   \n",
       "\n",
       "      isAliveFather  isAliveHeir  isAliveSpouse  isMarried  isNoble   age  \\\n",
       "0               0.0          0.0            NaN          0        0   NaN   \n",
       "1               NaN          NaN            1.0          1        1  97.0   \n",
       "2               NaN          NaN            NaN          0        1   NaN   \n",
       "3               NaN          NaN            0.0          1        1  23.0   \n",
       "4               NaN          NaN            1.0          1        1  29.0   \n",
       "...             ...          ...            ...        ...      ...   ...   \n",
       "1941            NaN          NaN            NaN          0        0   NaN   \n",
       "1942            NaN          NaN            NaN          0        0   NaN   \n",
       "1943            NaN          NaN            NaN          0        0   NaN   \n",
       "1944            NaN          NaN            NaN          0        0   NaN   \n",
       "1945            NaN          NaN            1.0          1        1  47.0   \n",
       "\n",
       "      numDeadRelations  boolDeadRelations  isPopular  popularity  isAlive  \n",
       "0                   11                  1          1    0.605351        0  \n",
       "1                    1                  1          1    0.896321        1  \n",
       "2                    0                  0          0    0.267559        1  \n",
       "3                    0                  0          0    0.183946        0  \n",
       "4                    0                  0          0    0.043478        1  \n",
       "...                ...                ...        ...         ...      ...  \n",
       "1941                 0                  0          1    0.351171        0  \n",
       "1942                 0                  0          0    0.096990        0  \n",
       "1943                 0                  0          0    0.030100        1  \n",
       "1944                 0                  0          0    0.130435        1  \n",
       "1945                 4                  1          0    0.210702        0  \n",
       "\n",
       "[1946 rows x 33 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "char1 = char_pred.copy()\n",
    "# Factorize our categorical features\n",
    "char1.loc[:, \"title\"] = pd.factorize(char1[\"title\"])[0]\n",
    "\n",
    "# char1.loc[:, \"culture\"] = pd.factorize(char1[\"culture\"])[0]\n",
    "# char1.loc[:, \"mother\"] = pd.factorize(char1[\"mother\"])[0]\n",
    "# char1.loc[:, \"father\"] = pd.factorize(char1[\"father\"])[0]\n",
    "# char1.loc[:, \"heir\"] = pd.factorize(char1[\"heir\"])[0]\n",
    "# char1.loc[:, \"house\"] = pd.factorize(char1[\"house\"])[0]\n",
    "# char1.loc[:, \"spouse\"] = pd.factorize(char1[\"spouse\"])[0]\n",
    "char1"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "b68a3c467d07aa8b20c71071b0b7606f948be72d30c76b97a14d11803ebfb6c4"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}