HR Analytics Job Change Classification - CART.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Table of contents\n",
    "1. [Import Data and Libraries](#Import)\n",
    "2. [EDA](#EDA)\n",
    "3. [Model Training](#Model)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "# Import Data and Libraries<a name=\"Import\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# hide warnings\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# print all columns\n",
    "pd.set_option('display.max_columns', None)\n",
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ref - https://www.kaggle.com/arashnic/hr-analytics-job-change-of-data-scientists?select=aug_train.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>enrollee_id</th>\n",
       "      <th>city</th>\n",
       "      <th>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>company_size</th>\n",
       "      <th>company_type</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>8949</td>\n",
       "      <td>city_103</td>\n",
       "      <td>0.920</td>\n",
       "      <td>Male</td>\n",
       "      <td>Has relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>&gt;20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>36</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29725</td>\n",
       "      <td>city_40</td>\n",
       "      <td>0.776</td>\n",
       "      <td>Male</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>15</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>&gt;4</td>\n",
       "      <td>47</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11561</td>\n",
       "      <td>city_21</td>\n",
       "      <td>0.624</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>Full time course</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never</td>\n",
       "      <td>83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>33241</td>\n",
       "      <td>city_115</td>\n",
       "      <td>0.789</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>Business Degree</td>\n",
       "      <td>&lt;1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>never</td>\n",
       "      <td>52</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>666</td>\n",
       "      <td>city_162</td>\n",
       "      <td>0.767</td>\n",
       "      <td>Male</td>\n",
       "      <td>Has relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Masters</td>\n",
       "      <td>STEM</td>\n",
       "      <td>&gt;20</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Funded Startup</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   enrollee_id      city  city_development_index gender  \\\n",
       "0         8949  city_103                   0.920   Male   \n",
       "1        29725   city_40                   0.776   Male   \n",
       "2        11561   city_21                   0.624    NaN   \n",
       "3        33241  city_115                   0.789    NaN   \n",
       "4          666  city_162                   0.767   Male   \n",
       "\n",
       "       relevent_experience enrolled_university education_level  \\\n",
       "0  Has relevent experience       no_enrollment        Graduate   \n",
       "1   No relevent experience       no_enrollment        Graduate   \n",
       "2   No relevent experience    Full time course        Graduate   \n",
       "3   No relevent experience                 NaN        Graduate   \n",
       "4  Has relevent experience       no_enrollment         Masters   \n",
       "\n",
       "  major_discipline experience company_size    company_type last_new_job  \\\n",
       "0             STEM        >20          NaN             NaN            1   \n",
       "1             STEM         15        50-99         Pvt Ltd           >4   \n",
       "2             STEM          5          NaN             NaN        never   \n",
       "3  Business Degree         <1          NaN         Pvt Ltd        never   \n",
       "4             STEM        >20        50-99  Funded Startup            4   \n",
       "\n",
       "   training_hours  target  \n",
       "0              36     1.0  \n",
       "1              47     0.0  \n",
       "2              83     0.0  \n",
       "3              52     1.0  \n",
       "4               8     0.0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# import data\n",
    "df = pd.read_csv(\"aug_train.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "#### Features\n",
    "- enrollee_id : Unique ID for candidate\n",
    "- city: City code\n",
    "- city_ development _index : Developement index of the city (scaled)\n",
    "- gender: Gender of candidate\n",
    "- relevent_experience: Relevant experience of candidate\n",
    "- enrolled_university: Type of University course enrolled if any\n",
    "- education_level: Education level of candidate\n",
    "- major_discipline :Education major discipline of candidate\n",
    "- experience: Candidate total experience in years\n",
    "- company_size: No of employees in current employer's company\n",
    "- company_type : Type of current employer\n",
    "- lastnewjob: Difference in years between previous job and current job\n",
    "- training_hours: training hours completed\n",
    "- target: 0 – Not looking for job change, 1 – Looking for a job change"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "# EDA<a name=\"EDA\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(19158, 14)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "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>enrollee_id</th>\n",
       "      <th>city</th>\n",
       "      <th>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>company_size</th>\n",
       "      <th>company_type</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>19158.000000</td>\n",
       "      <td>19158</td>\n",
       "      <td>19158.000000</td>\n",
       "      <td>14650</td>\n",
       "      <td>19158</td>\n",
       "      <td>18772</td>\n",
       "      <td>18698</td>\n",
       "      <td>16345</td>\n",
       "      <td>19093</td>\n",
       "      <td>13220</td>\n",
       "      <td>13018</td>\n",
       "      <td>18735</td>\n",
       "      <td>19158.000000</td>\n",
       "      <td>19158.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>123</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>22</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>city_103</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Male</td>\n",
       "      <td>Has relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>&gt;20</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4355</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13221</td>\n",
       "      <td>13792</td>\n",
       "      <td>13817</td>\n",
       "      <td>11598</td>\n",
       "      <td>14492</td>\n",
       "      <td>3286</td>\n",
       "      <td>3083</td>\n",
       "      <td>9817</td>\n",
       "      <td>8040</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>16875.358179</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.828848</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>NaN</td>\n",
       "      <td>65.366896</td>\n",
       "      <td>0.249348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9616.292592</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.123362</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>NaN</td>\n",
       "      <td>60.058462</td>\n",
       "      <td>0.432647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.448000</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>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>8554.250000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.740000</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>NaN</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>16982.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.903000</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>NaN</td>\n",
       "      <td>47.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>25169.750000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.920000</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>NaN</td>\n",
       "      <td>88.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>33380.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.949000</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>NaN</td>\n",
       "      <td>336.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         enrollee_id      city  city_development_index gender  \\\n",
       "count   19158.000000     19158            19158.000000  14650   \n",
       "unique           NaN       123                     NaN      3   \n",
       "top              NaN  city_103                     NaN   Male   \n",
       "freq             NaN      4355                     NaN  13221   \n",
       "mean    16875.358179       NaN                0.828848    NaN   \n",
       "std      9616.292592       NaN                0.123362    NaN   \n",
       "min         1.000000       NaN                0.448000    NaN   \n",
       "25%      8554.250000       NaN                0.740000    NaN   \n",
       "50%     16982.500000       NaN                0.903000    NaN   \n",
       "75%     25169.750000       NaN                0.920000    NaN   \n",
       "max     33380.000000       NaN                0.949000    NaN   \n",
       "\n",
       "            relevent_experience enrolled_university education_level  \\\n",
       "count                     19158               18772           18698   \n",
       "unique                        2                   3               5   \n",
       "top     Has relevent experience       no_enrollment        Graduate   \n",
       "freq                      13792               13817           11598   \n",
       "mean                        NaN                 NaN             NaN   \n",
       "std                         NaN                 NaN             NaN   \n",
       "min                         NaN                 NaN             NaN   \n",
       "25%                         NaN                 NaN             NaN   \n",
       "50%                         NaN                 NaN             NaN   \n",
       "75%                         NaN                 NaN             NaN   \n",
       "max                         NaN                 NaN             NaN   \n",
       "\n",
       "       major_discipline experience company_size company_type last_new_job  \\\n",
       "count             16345      19093        13220        13018        18735   \n",
       "unique                6         22            8            6            6   \n",
       "top                STEM        >20        50-99      Pvt Ltd            1   \n",
       "freq              14492       3286         3083         9817         8040   \n",
       "mean                NaN        NaN          NaN          NaN          NaN   \n",
       "std                 NaN        NaN          NaN          NaN          NaN   \n",
       "min                 NaN        NaN          NaN          NaN          NaN   \n",
       "25%                 NaN        NaN          NaN          NaN          NaN   \n",
       "50%                 NaN        NaN          NaN          NaN          NaN   \n",
       "75%                 NaN        NaN          NaN          NaN          NaN   \n",
       "max                 NaN        NaN          NaN          NaN          NaN   \n",
       "\n",
       "        training_hours        target  \n",
       "count     19158.000000  19158.000000  \n",
       "unique             NaN           NaN  \n",
       "top                NaN           NaN  \n",
       "freq               NaN           NaN  \n",
       "mean         65.366896      0.249348  \n",
       "std          60.058462      0.432647  \n",
       "min           1.000000      0.000000  \n",
       "25%          23.000000      0.000000  \n",
       "50%          47.000000      0.000000  \n",
       "75%          88.000000      0.000000  \n",
       "max         336.000000      1.000000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe(include='all')"
   ]
  },
  {
   "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>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>company_size</th>\n",
       "      <th>company_type</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>Male</td>\n",
       "      <td>Has relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>&gt;20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>36</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>Male</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>15</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>&gt;4</td>\n",
       "      <td>47</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>Full time course</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>STEM</td>\n",
       "      <td>5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>never</td>\n",
       "      <td>83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>NaN</td>\n",
       "      <td>No relevent experience</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Graduate</td>\n",
       "      <td>Business Degree</td>\n",
       "      <td>&lt;1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>never</td>\n",
       "      <td>52</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>Male</td>\n",
       "      <td>Has relevent experience</td>\n",
       "      <td>no_enrollment</td>\n",
       "      <td>Masters</td>\n",
       "      <td>STEM</td>\n",
       "      <td>&gt;20</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Funded Startup</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index gender      relevent_experience enrolled_university  \\\n",
       "0                   0.920   Male  Has relevent experience       no_enrollment   \n",
       "1                   0.776   Male   No relevent experience       no_enrollment   \n",
       "2                   0.624    NaN   No relevent experience    Full time course   \n",
       "3                   0.789    NaN   No relevent experience                 NaN   \n",
       "4                   0.767   Male  Has relevent experience       no_enrollment   \n",
       "\n",
       "  education_level major_discipline experience company_size    company_type  \\\n",
       "0        Graduate             STEM        >20          NaN             NaN   \n",
       "1        Graduate             STEM         15        50-99         Pvt Ltd   \n",
       "2        Graduate             STEM          5          NaN             NaN   \n",
       "3        Graduate  Business Degree         <1          NaN         Pvt Ltd   \n",
       "4         Masters             STEM        >20        50-99  Funded Startup   \n",
       "\n",
       "  last_new_job  training_hours  target  \n",
       "0            1              36     1.0  \n",
       "1           >4              47     0.0  \n",
       "2        never              83     0.0  \n",
       "3        never              52     1.0  \n",
       "4            4               8     0.0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# drop unnecessary columns\n",
    "df.drop(['enrollee_id', 'city'], axis = 1, inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Investigating Categorical Variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### `gender`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Male      13221\n",
       "NaN        4508\n",
       "Female     1238\n",
       "Other       191\n",
       "Name: gender, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['gender'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since `number of females << number of males`, we will convert this into a binary feature - **`Male_Yes`** and **`Male_No`**` and binarize them to 1 and 0, respectively"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    13221\n",
       "0     5937\n",
       "Name: gender, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['gender'] = df['gender'].map(lambda x : 1 if x == 'Male' else 0)\n",
    "df['gender'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### `relevant_experience`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Has relevent experience    13792\n",
       "No relevent experience      5366\n",
       "Name: relevent_experience, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['relevent_experience'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    13792\n",
       "0     5366\n",
       "Name: relevent_experience, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# minimiozing values for relevent_experience\n",
    "df['relevent_experience'] = df['relevent_experience'].map(lambda x : 1 if x == 'Has relevent experience' else 0)\n",
    "df['relevent_experience'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Academics - `enrolled_university`, `education_level`, `major_discipline`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "city_development_index       0\n",
       "gender                       0\n",
       "relevent_experience          0\n",
       "enrolled_university        386\n",
       "education_level            460\n",
       "major_discipline          2813\n",
       "experience                  65\n",
       "company_size              5938\n",
       "company_type              6140\n",
       "last_new_job               423\n",
       "training_hours               0\n",
       "target                       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "no_enrollment       13817\n",
       "Full time course     3757\n",
       "Part time course     1198\n",
       "NaN                   386\n",
       "Name: enrolled_university, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['enrolled_university'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Graduate          11598\n",
       "Masters            4361\n",
       "High School        2017\n",
       "NaN                 460\n",
       "Phd                 414\n",
       "Primary School      308\n",
       "Name: education_level, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['education_level'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "STEM               14492\n",
       "NaN                 2813\n",
       "Humanities           669\n",
       "Other                381\n",
       "Business Degree      327\n",
       "Arts                 253\n",
       "No Major             223\n",
       "Name: major_discipline, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['major_discipline'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Let's check distribution of values when one of the three variables is NULL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "when enrolled_university is null\n",
      "Graduate          171\n",
      "NaN               100\n",
      "Masters            53\n",
      "High School        47\n",
      "Primary School      9\n",
      "Phd                 6\n",
      "Name: education_level, dtype: int64\n",
      "\n",
      "when enrolled_university is null\n",
      "STEM               208\n",
      "NaN                157\n",
      "Humanities           9\n",
      "No Major             4\n",
      "Other                4\n",
      "Business Degree      2\n",
      "Arts                 2\n",
      "Name: major_discipline, dtype: int64\n",
      "\n",
      "when education_level is null\n",
      "no_enrollment       270\n",
      "NaN                 100\n",
      "Full time course     69\n",
      "Part time course     21\n",
      "Name: enrolled_university, dtype: int64\n",
      "\n",
      "when education_level is null\n",
      "NaN    460\n",
      "Name: major_discipline, dtype: int64\n",
      "\n",
      "when major_discipline is null\n",
      "no_enrollment       1481\n",
      "Full time course     976\n",
      "Part time course     199\n",
      "NaN                  157\n",
      "Name: enrolled_university, dtype: int64\n",
      "\n",
      "when major_discipline is null\n",
      "High School       2017\n",
      "NaN                460\n",
      "Primary School     308\n",
      "Graduate            22\n",
      "Masters              6\n",
      "Name: education_level, dtype: int64\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in ['enrolled_university', 'education_level', 'major_discipline']:\n",
    "    for j in ['enrolled_university', 'education_level', 'major_discipline']:\n",
    "        if i != j:\n",
    "            print(\"when %s is null\" % (i))\n",
    "            print(df.loc[df[i].isna(), j].value_counts(dropna = False))\n",
    "            print()\n",
    "        else:\n",
    "            continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "when `enrolled_university` is NULL, `education_level` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "when `enrolled_university` is NULL, `major_discipline` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "when `education_level` is NULL, `enrolled_university` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "when `education_level` is NULL, `major_discipline` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "when `major_discipline` is NULL, `enrolled_university` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "when `major_discipline` is NULL, `education_level` -\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "for i in ['enrolled_university', 'education_level', 'major_discipline']:\n",
    "    for j in ['enrolled_university', 'education_level', 'major_discipline']:\n",
    "        if i != j:\n",
    "            print(\"when `%s` is NULL, `%s` -\" % (i, j))\n",
    "            sns.countplot(df.loc[df[i].isna(), j].fillna(\"N/A\"))\n",
    "            plt.show()\n",
    "            print()\n",
    "        else:\n",
    "            continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "tags": []
   },
   "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>enrolled_university</th>\n",
       "      <th>Full time course</th>\n",
       "      <th>N/A</th>\n",
       "      <th>Part time course</th>\n",
       "      <th>no_enrollment</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>education_level</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Graduate</th>\n",
       "      <td>2209</td>\n",
       "      <td>171</td>\n",
       "      <td>784</td>\n",
       "      <td>8434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>High School</th>\n",
       "      <td>864</td>\n",
       "      <td>47</td>\n",
       "      <td>166</td>\n",
       "      <td>940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Masters</th>\n",
       "      <td>557</td>\n",
       "      <td>53</td>\n",
       "      <td>204</td>\n",
       "      <td>3547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>N/A</th>\n",
       "      <td>69</td>\n",
       "      <td>100</td>\n",
       "      <td>21</td>\n",
       "      <td>270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Phd</th>\n",
       "      <td>25</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Primary School</th>\n",
       "      <td>33</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>255</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "enrolled_university  Full time course  N/A  Part time course  no_enrollment\n",
       "education_level                                                            \n",
       "Graduate                         2209  171               784           8434\n",
       "High School                       864   47               166            940\n",
       "Masters                           557   53               204           3547\n",
       "N/A                                69  100                21            270\n",
       "Phd                                25    6                12            371\n",
       "Primary School                     33    9                11            255"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['education_level'].fillna(\"N/A\"), df['enrolled_university'].fillna(\"N/A\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "#### Observations - \n",
    "1. ***`enrolled_university`*** -  \n",
    "- Majority of NULLS have **`education_level` == *graduation or higher*** , so we can assume that they did not enroll to any university course. Remaining values can be assumed as \"not enrolled\" too.  \n",
    "- To further solidify our assumption, we will take a look at the crosstab above as well. We observe that `education_level` == NULL has maximum values against `enrolled_university` = \"no enrollment\". While, for `enrolled_university` == NULL has maximum values in `education_level` = \"Graduate\".\n",
    "\n",
    "2. ***`education_level`*** - We plan to make this variable have 3 categories - \"High School or Lower\", \"Graduation\" and \"Post Graduation\". We can then club NULLS with \"High School or Lower\" category since their `major_discipline` is also NULL (heavy majority of the cases where `major_discipline` is NULL have `education_level` up to High School only). \n",
    "\n",
    "3. ***`major_discipline`*** - We plan to make this a binary variable - \"STEM\" and \"Non-STEM\". Now, heavy majority of the cases where Major Discipline is NULL have education up to High School only. So, we will simply put NULLS into \"Non-STEM\" category."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='enrolled_university', ylabel='count'>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# enrolled_university\n",
    "df['enrolled_university'] = df['enrolled_university'].fillna(\"no_enrollment\")\n",
    "sns.countplot(df['enrolled_university'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- The variable `enrolled_university` is also extremely skewed which is why we will merge \"Full Time\" and \"Part Time\" into one category i.e. **enrolled**.\n",
    "- We will also binarize the variable into 1 and 0 meaning \"Enrolled\" and \"Not Enrolled\" respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='enrolled_university', ylabel='count'>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['enrolled_university'] = df['enrolled_university'].map(lambda x : 0 if x == \"no_enrollment\" else 1)\n",
    "sns.countplot(df['enrolled_university'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Graduate          11598\n",
       "Masters            4361\n",
       "High School        2017\n",
       "Phd                 414\n",
       "Primary School      308\n",
       "Name: education_level, dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# education_level\n",
    "df['education_level'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "def education(x):\n",
    "    if x in ['Masters', 'Phd']:\n",
    "        return 2\n",
    "        # \"Post Graduate\"\n",
    "    elif x == \"Graduate\":\n",
    "        return 1\n",
    "        # \"Graduate\"\n",
    "    else:\n",
    "        return 0 \n",
    "        # \"Non-Graduate\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='education_level', ylabel='count'>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['education_level'] = df['education_level'].map(education)\n",
    "sns.countplot(df['education_level'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='major_discipline', ylabel='count'>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# major_discipline\n",
    "df['major_discipline'] = df['major_discipline'].map(lambda x : 1 if x == \"STEM\" else 0)\n",
    "sns.countplot(df['major_discipline'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### `experience`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       ">20    3286\n",
       "5      1430\n",
       "4      1403\n",
       "3      1354\n",
       "6      1216\n",
       "2      1127\n",
       "7      1028\n",
       "10      985\n",
       "9       980\n",
       "8       802\n",
       "15      686\n",
       "11      664\n",
       "14      586\n",
       "1       549\n",
       "<1      522\n",
       "16      508\n",
       "12      494\n",
       "13      399\n",
       "17      342\n",
       "19      304\n",
       "18      280\n",
       "20      148\n",
       "NaN      65\n",
       "Name: experience, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['experience'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Observations - \n",
    "1. NULLS - There are only 65 NULLS, we can equate them to 0 because this is the information which companies usually have or we can simply remove the rows. We will do the former.\n",
    "2. We can probably use this as a categorical variable by using ranges of experience, which is also commonly seen in job requirements. We can use the below groups - \n",
    "        - 0 to 2\n",
    "        - 3 to 6\n",
    "        - 7 to 10\n",
    "        - 10 to 20\n",
    "        - 20 and above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def experience_groups(x):\n",
    "    if (x in [np.nan, '1', '2']):\n",
    "        return 0 \n",
    "        # \"0 to 2\"\n",
    "    elif x in ['3', '4', '5', '6']:\n",
    "        return 1\n",
    "        # \"3 to 6\"\n",
    "    elif x in ['7', '8', '9', '10']:\n",
    "        return 2\n",
    "        # \"7 to 10\"\n",
    "    elif x == \">20\":\n",
    "        return 3\n",
    "        # \"20+\"\n",
    "    else:\n",
    "        return 4\n",
    "        # \"10 to 20\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    5403\n",
       "4    4933\n",
       "2    3795\n",
       "3    3286\n",
       "0    1741\n",
       "Name: experience, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['experience'] = df['experience'].map(experience_groups)\n",
    "df['experience'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  `last_new_job`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1        8040\n",
       ">4       3290\n",
       "2        2900\n",
       "never    2452\n",
       "4        1029\n",
       "3        1024\n",
       "NaN       423\n",
       "Name: last_new_job, dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['last_new_job'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='last_new_job', ylabel='count'>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEHCAYAAABfkmooAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAcpklEQVR4nO3de5RV5Z3m8e+jYCoYuYq2UmgRYeLCO5RKh2476kSQNpBJ1MZJBJUMk0gSzbTpEFeWd7vNOOnEtJcsElCIRry1QmjEMKjRXFRAURHbAW+hiJeSWzQEBPzNH/stPGIV+1R5dp0q6vmsdVad/e537/07LOCpd+993q2IwMzMbFf2qHYBZmbW8TkszMwsl8PCzMxyOSzMzCyXw8LMzHJ1q3YBRdh3332jrq6u2mWYmXUqS5cufSsi+je3brcMi7q6OpYsWVLtMszMOhVJr7a0zqehzMwsl8PCzMxyOSzMzCxXodcsJPUGfgYcDgRwHvACcAdQB7wCnBkR6yUJuA4YA2wCzomIJ9N+JgLfS7u9KiJmFlm3mdnOtm7dSkNDA5s3b652KR9ZTU0NtbW1dO/evextir7AfR2wICJOl7QX0AO4GFgUEddImgpMBb4DnAoMSa/jgZuA4yX1BS4F6skCZ6mkuRGxvuDazcx2aGhoYJ999qGuro7sd9vOKSJYu3YtDQ0NDBo0qOztCjsNJakXcAIwHSAi3o2IDcA4oGlkMBP4fHo/DpgVmceA3pIOAEYBCyNiXQqIhcDoouo2M2vO5s2b6devX6cOCgBJ9OvXr9UjpCKvWQwCGoGbJT0l6WeS9gb2j4jXUp/Xgf3T+wHA6pLtG1JbS+0fIGmypCWSljQ2Nlb4o5iZ0emDoklbPkeRYdENGAbcFBHHAH8mO+W0Q2Tzo1dkjvSImBYR9RFR379/s98pMTOzNioyLBqAhoh4PC3fTRYeb6TTS6Sfb6b1a4CBJdvXpraW2s3MOqQNGzZw4403Fn6c++67jxUrVhR+HCjwAndEvC5ptaRPRcQLwMnAivSaCFyTfs5Jm8wFvi5pNtkF7o0R8ZqkB4B/ltQn9TsF+G5r6xn+7Vkf7QOVWHrthIrty8x2P01hcf7555fVPyKICPbYo3W/v993332cdtppDB06tC1ltkrRd0N9A7gt3Qn1EnAu2WjmTkmTgFeBM1Pf+WS3za4iu3X2XICIWCfpSmBx6ndFRKwruG4zszabOnUqL774IkcffTQnnngizzzzDOvXr2fr1q1cddVVjBs3jldeeYVRo0Zx/PHHs3TpUubPn8+sWbO49dZb6d+/PwMHDmT48OFcdNFFvPjii0yZMoXGxkZ69OjBT3/6U9atW8fcuXP59a9/zVVXXcU999zDIYccUthnKjQsImIZ2S2vOzu5mb4BTGlhPzOAGRUtzsysINdccw3Lly9n2bJlbNu2jU2bNtGzZ0/eeustRowYwdixYwFYuXIlM2fOZMSIESxevJh77rmHp59+mq1btzJs2DCGDx8OwOTJk/nJT37CkCFDePzxxzn//PN58MEHGTt2LKeddhqnn3564Z9pt5xI0Myso4gILr74Yh555BH22GMP1qxZwxtvvAHAwQcfzIgRIwD47W9/y7hx46ipqaGmpobPfe5zALzzzjv87ne/44wzztixzy1btrT753BYmJkV6LbbbqOxsZGlS5fSvXt36urqdnzHYe+9987d/r333qN3794sW7as4Ep3zXNDmZlV2D777MPbb78NwMaNG9lvv/3o3r07Dz30EK++2vws4CNHjuSXv/wlmzdv5p133mHevHkA9OzZk0GDBnHXXXcB2Ujl6aef/tBxiuawMDOrsH79+jFy5EgOP/xwli1bxpIlSzjiiCOYNWsWhx56aLPbHHvssYwdO5YjjzySU089lSOOOIJevXoB2ehk+vTpHHXUURx22GHMmZPdRDp+/HiuvfZajjnmGF588cVCP5NPQ5mZFeAXv/hFbp/ly5d/YPmiiy7isssuY9OmTZxwwgk7LnAPGjSIBQsWfGj7kSNHdv7vWZiZWetMnjyZFStWsHnzZiZOnMiwYcOqXdIODgszsw6inNFItfiahZmZ5XJYmJlZLoeFmZnlcliYmVkuX+A2M2uDSs5kDeXPZr1gwQIuuOACtm/fzle+8hWmTv3AY4LYsmULEyZMYOnSpfTr14877riDurq6j1yfRxZmZp3E9u3bmTJlCvfffz8rVqzg9ttv/9D3LKZPn06fPn1YtWoV3/rWt/jOd75TkWM7LMzMOoknnniCwYMH88lPfpK99tqL8ePH7/g2d5M5c+YwceJEAE4//XQWLVpENqn3R+OwMDPrJNasWcPAge8/OLS2tpY1a9a02Kdbt2706tWLtWvXfuRjOyzMzCyXw8LMrJMYMGAAq1ev3rHc0NDAgAEDWuyzbds2Nm7cSL9+/T7ysR0WZmadxLHHHsvKlSt5+eWXeffdd5k9e/aOp+41GTt2LDNnzgTg7rvv5qSTTkLSRz62b501M2uDcm91raRu3bpx/fXXM2rUKLZv3855553HYYcdxiWXXEJ9fT1jx45l0qRJnH322QwePJi+ffsye/bsyhy7InsxM7N2MWbMGMaMGfOBtiuuuGLH+5qamh0PSqokn4YyM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPLVeits5JeAd4GtgPbIqJeUl/gDqAOeAU4MyLWK/vWyHXAGGATcE5EPJn2MxH4XtrtVRExs8i6zczy/OGKIyq6v4MueTa3z3nnnce8efPYb7/9WL58+YfWRwQXXHAB8+fPp0ePHtxyyy0MGzasIvW1x8jixIg4OiLq0/JUYFFEDAEWpWWAU4Eh6TUZuAkghculwPHAccClkvq0Q91mZh3KOeecw4IFC1pcf//997Ny5UpWrlzJtGnT+NrXvlaxY1fjNNQ4oGlkMBP4fEn7rMg8BvSWdAAwClgYEesiYj2wEBjdzjWbmVXdCSecQN++fVtcP2fOHCZMmIAkRowYwYYNG3jttdcqcuyiwyKAX0laKmlyats/Ipqqfx3YP70fAKwu2bYhtbXU/gGSJktaImlJY2NjJT+DmVmnUM4U5m1V9HQffxMRayTtByyU9J+lKyMiJH30p3Jk+5oGTAOor6+vyD7NzCxT6MgiItakn28C95Jdc3gjnV4i/XwzdV8DDCzZvDa1tdRuZmYlypnCvK0KCwtJe0vap+k9cAqwHJgLTEzdJgJNzwScC0xQZgSwMZ2uegA4RVKfdGH7lNRmZmYlxo4dy6xZs4gIHnvsMXr16sUBBxxQkX0XeRpqf+DeNI96N+AXEbFA0mLgTkmTgFeBM1P/+WS3za4iu3X2XICIWCfpSmBx6ndFRKwrsG4zs1zl3OpaaWeddRYPP/wwb731FrW1tVx++eVs3boVgK9+9auMGTOG+fPnM3jwYHr06MHNN99csWMXFhYR8RJwVDPta4GTm2kPYEoL+5oBzKh0jWZmncntt9++y/WSuOGGGwo5tr/BbWZmuRwWZmaWy2FhZlam7Gx559eWz+GwMDMrQ01NDWvXru30gRERrF27lpqamlZt52dwm5mVoba2loaGBnaHGSJqamqora1t1TYOCzOzMnTv3p1BgwZVu4yq8WkoMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NchYeFpD0lPSVpXloeJOlxSask3SFpr9T+sbS8Kq2vK9nHd1P7C5JGFV2zmZl9UHuMLC4Ani9Z/j7ww4gYDKwHJqX2ScD61P7D1A9JQ4HxwGHAaOBGSXu2Q91mZpYUGhaSaoG/B36WlgWcBNyduswEPp/ej0vLpPUnp/7jgNkRsSUiXgZWAccVWbeZmX1Q0SOLHwH/BLyXlvsBGyJiW1puAAak9wOA1QBp/cbUf0d7M9vsIGmypCWSljQ2Nlb4Y5iZdW2FhYWk04A3I2JpUccoFRHTIqI+Iur79+/fHoc0M+syuhW475HAWEljgBqgJ3Ad0FtStzR6qAXWpP5rgIFAg6RuQC9gbUl7k9JtzMysHRQ2soiI70ZEbUTUkV2gfjAivgQ8BJyeuk0E5qT3c9Myaf2DERGpfXy6W2oQMAR4oqi6zczsw4ocWbTkO8BsSVcBTwHTU/t04OeSVgHryAKGiHhO0p3ACmAbMCUitrd/2WZmXVe7hEVEPAw8nN6/RDN3M0XEZuCMFra/Gri6uArNzGxX/A1uMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NcZYWFpEXltJmZ2e5pl8+zkFQD9AD2ldQHUFrVExhQcG1mZtZB5D386H8CFwIHAkt5Pyz+BFxfXFlmZtaR7DIsIuI64DpJ34iIf2unmszMrIMp67GqEfFvkj4N1JVuExGzCqrLzMw6kLLCQtLPgUOAZcD21ByAw8LMrAsoKyyAemBoRESRxZiZWcdU7vcslgN/VWQhZmbWcZU7stgXWCHpCWBLU2NEjC2kKjMz61DKDYvLiizCzMw6tnLvhvp10YWYmVnHVe7dUG+T3f0EsBfQHfhzRPQsqjAzM+s4yrrAHRH7RETPFA4fB74I3LirbSTVSHpC0tOSnpN0eWofJOlxSask3SFpr9T+sbS8Kq2vK9nXd1P7C5JGtfXDmplZ27R61tnI3Afk/ae9BTgpIo4CjgZGSxoBfB/4YUQMBtYDk1L/ScD61P7D1A9JQ4HxwGHAaOBGSXu2tm4zM2u7ck9DfaFkcQ+y711s3tU26TsZ76TF7ukVwEnAf0/tM8kunt8EjOP9C+l3A9dLUmqfHRFbgJclrQKOA35fTu1mZvbRlXs31OdK3m8DXiH7T3yX0ghgKTAYuAF4EdgQEdtSlwben712ALAaICK2SdoI9Evtj5XstnSb0mNNBiYDHHTQQWV+LDMzK0e5d0Od25adR8R24GhJvYF7gUPbsp8yjzUNmAZQX1/vb5qbmVVQuQ8/qpV0r6Q30+seSbXlHiQiNgAPAX8N9JbUFFK1wJr0fg0wMB2vG9ALWFva3sw2ZmbWDsq9wH0zMJfsuRYHAr9MbS2S1D+NKJD0ceCzwPNkoXF66jYRmJPez03LpPUPpusec4Hx6W6pQcAQ4Iky6zYzswoo95pF/4goDYdbJF2Ys80BwMx03WIP4M6ImCdpBTBb0lXAU8D01H868PN0AXsd2R1QRMRzku4EVpBdL5mSTm+ZmVk7KTcs1kr6MnB7Wj6L7BRRiyLiGeCYZtpfIrubaef2zcAZLezrauDqMms1M7MKK/c01HnAmcDrwGtkp4nOKagmMzPrYModWVwBTIyI9QCS+gL/hyxEzMxsN1fuyOLIpqAAiIh1NHOKyczMdk/lhsUekvo0LaSRRbmjEjMz6+TK/Q//B8DvJd2Vls/AF5zNzLqMcr/BPUvSErJ5nQC+EBEriivLdjb827Mqtq+l106o2L7MrGso+1RSCgcHhJlZF9TqKcrNzKzrcViYmVkuh4WZmeVyWJiZWS6HhZmZ5XJYmJlZLoeFmZnlcliYmVkuh4WZmeVyWJiZWS6HhZmZ5XJYmJlZLoeFmZnlcliYmVkuh4WZmeXyo1HNrEPzg786Bo8szMwsl0cWbfCHK46o2L4OuuTZiu3LzKwoHlmYmVmuwsJC0kBJD0laIek5SRek9r6SFkpamX72Se2S9GNJqyQ9I2lYyb4mpv4rJU0sqmYzM2tekSOLbcA/RsRQYAQwRdJQYCqwKCKGAIvSMsCpwJD0mgzcBFm4AJcCxwPHAZc2BYyZmbWPwsIiIl6LiCfT+7eB54EBwDhgZuo2E/h8ej8OmBWZx4Dekg4ARgELI2JdRKwHFgKji6rbzMw+rF2uWUiqA44BHgf2j4jX0qrXgf3T+wHA6pLNGlJbS+07H2OypCWSljQ2Nlb2A5iZdXGFh4WkTwD3ABdGxJ9K10VEAFGJ40TEtIioj4j6/v37V2KXZmaWFBoWkrqTBcVtEfHvqfmNdHqJ9PPN1L4GGFiyeW1qa6ndzMzaSZF3QwmYDjwfEf9asmou0HRH00RgTkn7hHRX1AhgYzpd9QBwiqQ+6cL2KanNzMzaSZFfyhsJnA08K2lZarsYuAa4U9Ik4FXgzLRuPjAGWAVsAs4FiIh1kq4EFqd+V0TEugLrNjOznRQWFhHxG0AtrD65mf4BTGlhXzOAGZWrzszMWsPf4DYzs1wOCzMzy+WwMDOzXA4LMzPL5SnKrV34ATZmnZtHFmZmlsthYWZmuRwWZmaWy2FhZma5HBZmZpbLYWFmZrkcFmZmlsthYWZmuRwWZmaWy2FhZma5HBZmZpbLYWFmZrkcFmZmlsthYWZmuRwWZmaWy2FhZma5HBZmZpbLYWFmZrkcFmZmlsthYWZmuRwWZmaWq7CwkDRD0puSlpe09ZW0UNLK9LNPapekH0taJekZScNKtpmY+q+UNLGoes3MrGVFjixuAUbv1DYVWBQRQ4BFaRngVGBIek0GboIsXIBLgeOB44BLmwLGzMzaT2FhERGPAOt2ah4HzEzvZwKfL2mfFZnHgN6SDgBGAQsjYl1ErAcW8uEAMjOzgrX3NYv9I+K19P51YP/0fgCwuqRfQ2prqf1DJE2WtETSksbGxspWbWbWxVXtAndEBBAV3N+0iKiPiPr+/ftXardmZkb7h8Ub6fQS6eebqX0NMLCkX21qa6ndzMzaUXuHxVyg6Y6micCckvYJ6a6oEcDGdLrqAeAUSX3She1TUpuZmbWjbkXtWNLtwGeAfSU1kN3VdA1wp6RJwKvAman7fGAMsArYBJwLEBHrJF0JLE79roiInS+am5lZwQoLi4g4q4VVJzfTN4ApLexnBjCjgqWZmVkr+RvcZmaWq7CRhdnuZPi3Z1VsX0uvnVCxfZWjkrVD+9dvHYNHFmZmlssjC+t0/nDFERXb10GXPFuxfZntzhwWZtZlVPIXDehav2z4NJSZmeVyWJiZWS6HhZmZ5fI1iy7IF4jNrLU8sjAzs1wOCzMzy+WwMDOzXA4LMzPL5bAwM7NcDgszM8vlsDAzs1wOCzMzy+WwMDOzXA4LMzPL5ek+zMwK1JmfsljKIwszM8vlkYVZO+vsEzl29vqtbTyyMDOzXA4LMzPL5dNQZmadRDVPAXpkYWZmuTpNWEgaLekFSaskTa12PWZmXUmnCAtJewI3AKcCQ4GzJA2tblVmZl1HpwgL4DhgVUS8FBHvArOBcVWuycysy1BEVLuGXJJOB0ZHxFfS8tnA8RHx9ZI+k4HJafFTwAsFlrQv8FaB+y+a668u1189nbl2KL7+gyOif3Mrdpu7oSJiGjCtPY4laUlE1LfHsYrg+qvL9VdPZ64dqlt/ZzkNtQYYWLJcm9rMzKwddJawWAwMkTRI0l7AeGBulWsyM+syOsVpqIjYJunrwAPAnsCMiHiuiiW1y+muArn+6nL91dOZa4cq1t8pLnCbmVl1dZbTUGZmVkUOCzMzy+WwaAVJMyS9KWl5tWupJEk/lvROtevoCiTtKekpSfOqXUtrSKqR9ISkpyU9J+nyatfUGpIGSnpI0opU/wXVrqk5kkLSD0qWL5J0WcnyAZJ+VbJ8oaTNknoVXZvDonVuAUZXu4hySepTRp96ILdfZyepo9zMcQHwfLWLaIMtwEkRcRRwNDBa0ojqltQq24B/jIihwAhgSgedMmgL8AVJ+7awfjTZjT5NziK7W/QLRRfmsGiFiHgEWFftOlphiaTbJJ0kSTuvTHNuXQv8U/uXBpLqJD0v6afpt71fSfq4pEMkLZC0VNKjkg6V1EvSq5L2SNvuLWm1pO7N9U99bpH0E0mPA/+7Gp+xlKRa4O+Bn1W7ltaKTNPos3t6dZq7YyLitYh4Mr1/myywB1S3qmZtI7vj6VstrB8N3A8g6RDgE8D3yEKjUA6L3dt/AW4Hvg6skHSxpANL1n8dmBsRr1WluswQ4IaIOAzYAHyR7B/LNyJiOHARcGNEbASWAX+XtjsNeCAitjbXv2T/tcCnI+J/tcNnyfMjsmB+r8p1tEk6hbYMeBNYGBGPV7mkNpFUBxwDdNT6bwC+tPOppfTL3aciYkVqGk82T96jwKck7V9kUR1laG4FiIjtwDxgnqT+wL8Af5D0aaABOAP4TPUqBODliFiW3i8F6oBPA3eVDIY+ln7eAfwD8BDZP5QbJX1iF/0B7kp/DlUl6TTgzYhYKukzVS6nTdKf49GSegP3Sjo8IjrV9bv09+Ue4MKI+FO162lORPxJ0izgm8BfSlYdzwcD7izgv0XEe5LuIfv3fH1RdTksdnPpt5PxwDnAu8B5wDPAycBgYFX6T7aHpFURMbidS9xS8n47sD+wISKObqbvXOCfJfUFhgMPAnvvoj/AnytX6kcyEhgraQxQA/SUdGtEfLnKdbVaRGyQ9BDZKZFOExaSupMFxW0R8e/VrifHj4AngZtL2k4FFgBIOoJsVL4w/fvdC3iZAsPCp6F2Y5JuJfsLNwiYEBF/FxGzImJzRPxHRPxVRNRFRB2wqQpB0Zw/AS9LOgNAmaMA0jnzxcB1wLyI2J5+O2y2f0cSEd+NiNr0Zz0eeLAzBYWk/mlEgaSPA58F/rOqRbVCumY3HXg+Iv612vXkiYh1wJ3ApJLmk4H/m96fBVzW9O83Ig4EDpR0cFE1OSxaQdLtwO/Jzg82SJqUt02V3Ul2jnNqRKysdjGt8CVgkqSngef44LNL7gC+nH6W098q4wDgIUnPkAX2wojoTLf/jgTOBk6StCy9xlS7qBw/IJuSnHQaeXO6OA/ZLxz37tT/3tReCE/3YWbWwUn6MlAbEddUrQaHhZmZ5fFpKDMzy+WwMDOzXA4LMzPL5bAwM7NcDgszM8vlsLAuqa1TsqcpoXtUup421FEv6cc5fS6TdFF71WS7N4eFWetcCFQ9LCJiSUR8s9p1WNfhsLAuTdInJC2S9KSkZyWNS+17S/oPZQ/7WS7pHyR9EziQ7JvMD+1in+9Iujpt+1jTbKBpyox7JC1Or5Gp/VlJvdNUJWslTUjtsyR9toVjfEbpAUqS+kq6T9Iz6XhHlnQ9StLvJa2U9D8q8odmXZLDwrq6zWQzdw4DTgR+kOYRGg38MSKOiojDgQUR8WPgj8CJEXHiLva5N/BYelDQI0DTf9LXAT+MiGPJpmJveq7Fb8mmozgMeAn429T+18DvyvgMlwNPRcSRwMXArJJ1RwInpX1dstMU9WZl86yz1tWJbCbbE8ieMzGAbObbZ8mC4/tkkxY+2op9vks2NTxk0643jQ7+KzC0ZCr1nmnK7EeBE4BXgZuAyZIGAOsjopxZc/+GLHyIiAcl9ZPUM62bExF/Af6SRkPHAfe14rOYAR5ZmH0J6A8MT9OcvwHURMT/A4aRhcZVki5pxT63xvvz6Gzn/V/K9gBGRMTR6TUgzaT7CNlo4m+Bh4FG4HSyEPmodp7Px/P7WJs4LKyr60X2UKKtkk4EDgZIp2s2RcStZI+eHZb6vw3s08Zj/Qr4RtOCpKMBImI12eyiQyLiJeA3ZE/8e6TM/T5KFnqkByu9VfJgn3GSaiT1I3vQ1eI21m5dnE9DWVd3G/BLSc8CS3j/GQ1HANdKeg/YCnwttU8DFkj6Y851i+Z8E7ghTfPdjSwMvprWPQ7smd4/SvZUw9/k7K9plHAZMCPtdxMwsaTPM2RPFtwXuDIi/tjKms0Azzpr1ilJ+iIwNiIm5nY2qwCPLMw6GUljgavJHpFr1i4cFmZtJOlx4GM7NZ8dEc9W8BijgO/v1PxyRBxaqWOYlcOnoczMLJfvhjIzs1wOCzMzy+WwMDOzXA4LMzPL9f8Bc1rDXUjL3p8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(df['last_new_job'].fillna(\"N/A\"), hue = df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- There are 423 NULLs in this variable and these can be equated to \"never\" since this is the information the company usually has.\n",
    "- Based on the frequency distribution above, we can create the below groups - \n",
    "        - \"1\"\n",
    "        - \"2 or more\"\n",
    "        - \"never\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def last_job_groups(x):\n",
    "    if x in [np.nan, 'never']:\n",
    "        return 0 \n",
    "        # \"never\"\n",
    "    elif x == \"1\":\n",
    "        return 1\n",
    "    else:\n",
    "        return 2 \n",
    "        # \"2 or more\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    8243\n",
       "1    8040\n",
       "0    2875\n",
       "Name: last_new_job, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['last_new_job'] = df['last_new_job'].map(last_job_groups)\n",
    "df['last_new_job'].value_counts(dropna = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>company_size</th>\n",
       "      <th>company_type</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>36</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>2</td>\n",
       "      <td>47</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>0</td>\n",
       "      <td>0</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>0</td>\n",
       "      <td>83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Pvt Ltd</td>\n",
       "      <td>0</td>\n",
       "      <td>52</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>50-99</td>\n",
       "      <td>Funded Startup</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index  gender  relevent_experience  enrolled_university  \\\n",
       "0                   0.920       1                    1                    0   \n",
       "1                   0.776       1                    0                    0   \n",
       "2                   0.624       0                    0                    1   \n",
       "3                   0.789       0                    0                    0   \n",
       "4                   0.767       1                    1                    0   \n",
       "\n",
       "   education_level  major_discipline  experience company_size    company_type  \\\n",
       "0                1                 1           3          NaN             NaN   \n",
       "1                1                 1           4        50-99         Pvt Ltd   \n",
       "2                1                 1           1          NaN             NaN   \n",
       "3                1                 0           4          NaN         Pvt Ltd   \n",
       "4                2                 1           3        50-99  Funded Startup   \n",
       "\n",
       "   last_new_job  training_hours  target  \n",
       "0             1              36     1.0  \n",
       "1             2              47     0.0  \n",
       "2             0              83     0.0  \n",
       "3             0              52     1.0  \n",
       "4             2               8     0.0  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### `company_size` and `company_type`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NaN          0.309949\n",
       "50-99        0.160925\n",
       "100-500      0.134200\n",
       "10000+       0.105387\n",
       "10/49        0.076783\n",
       "1000-4999    0.069318\n",
       "<10          0.068274\n",
       "500-999      0.045777\n",
       "5000-9999    0.029387\n",
       "Name: company_size, dtype: float64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['company_size'].value_counts(dropna = False, normalize = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pvt Ltd                0.512423\n",
       "NaN                    0.320493\n",
       "Funded Startup         0.052250\n",
       "Public Sector          0.049849\n",
       "Early Stage Startup    0.031475\n",
       "NGO                    0.027195\n",
       "Other                  0.006316\n",
       "Name: company_type, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# value_counts\n",
    "df['company_type'].value_counts(dropna = False, normalize = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Over 30% of data is missing for both the variables, so we will drop these features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(['company_size', 'company_type'], axis = 1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>36</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>47</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>83</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>52</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index  gender  relevent_experience  enrolled_university  \\\n",
       "0                   0.920       1                    1                    0   \n",
       "1                   0.776       1                    0                    0   \n",
       "2                   0.624       0                    0                    1   \n",
       "3                   0.789       0                    0                    0   \n",
       "4                   0.767       1                    1                    0   \n",
       "\n",
       "   education_level  major_discipline  experience  last_new_job  \\\n",
       "0                1                 1           3             1   \n",
       "1                1                 1           4             2   \n",
       "2                1                 1           1             0   \n",
       "3                1                 0           4             0   \n",
       "4                2                 1           3             2   \n",
       "\n",
       "   training_hours  target  \n",
       "0              36     1.0  \n",
       "1              47     0.0  \n",
       "2              83     0.0  \n",
       "3              52     1.0  \n",
       "4               8     0.0  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Investigating Numerical Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to get normality plots\n",
    "def prob_plots(x):\n",
    "    from scipy.stats import probplot\n",
    "    fig, ax = plt.subplots(1, 3, figsize = (18, 6))\n",
    "\n",
    "    res = probplot(x, plot = ax[0])\n",
    "    ax[0].set_title(\"Q-Q Plot\")\n",
    "\n",
    "    ax[1] = sns.kdeplot(x, fill = True, ax = ax[1])\n",
    "    ax[1].set_title(\"KDE (PDF) Plot\")\n",
    "\n",
    "    ax[2] = sns.ecdfplot(x, ax = ax[2])\n",
    "    ax[2].set_title(\"KDE (PDF) Plot\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### `city_development_index`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "prob_plots(df['city_development_index'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='target', ylabel='city_development_index'>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = df['city_development_index'], x = df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It seems that a higher median `city_development_index` relates to employees being more probable to stay in their current jobs.\n",
    "- We will keep the variable as is for now."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='city_development_index'>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check for outliers\n",
    "sns.boxplot(y = df['city_development_index'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### `training_hours`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "prob_plots(df['training_hours'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='target', ylabel='training_hours'>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = df['training_hours'], x = df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It seems the feature has too many outliers, or rather extreme values, in it. Hence, we will try applying a log transformation to see if that effect is reduced. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='target', ylabel='training_hours'>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check for outliers\n",
    "sns.boxplot(y = np.log(df['training_hours']), x = df['target'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- This looks much better. We will move ahead with the transformation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>target</th>\n",
       "      <th>log_training_hours</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.583519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.850148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.418841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.951244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.079442</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index  gender  relevent_experience  enrolled_university  \\\n",
       "0                   0.920       1                    1                    0   \n",
       "1                   0.776       1                    0                    0   \n",
       "2                   0.624       0                    0                    1   \n",
       "3                   0.789       0                    0                    0   \n",
       "4                   0.767       1                    1                    0   \n",
       "\n",
       "   education_level  major_discipline  experience  last_new_job  target  \\\n",
       "0                1                 1           3             1     1.0   \n",
       "1                1                 1           4             2     0.0   \n",
       "2                1                 1           1             0     0.0   \n",
       "3                1                 0           4             0     1.0   \n",
       "4                2                 1           3             2     0.0   \n",
       "\n",
       "   log_training_hours  \n",
       "0            3.583519  \n",
       "1            3.850148  \n",
       "2            4.418841  \n",
       "3            3.951244  \n",
       "4            2.079442  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['log_training_hours'] = np.log(df['training_hours'])\n",
    "df.drop('training_hours', axis = 1, inplace = True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Check for multicollinearity between two numeric variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# correlation plot\n",
    "plt.figure(figsize=(8,6))\n",
    "sns.set_context('paper', font_scale=1.5)\n",
    "\n",
    "corr = df[[\"city_development_index\", \"log_training_hours\"]].corr()\n",
    "\n",
    "hm = sns.heatmap(corr, cmap=\"coolwarm\", annot=True, vmin = -1, vmax = 1, fmt='.3f')\n",
    "hm.set_xticklabels(labels=corr.columns.values, rotation=90)\n",
    "hm.set_yticklabels(labels=corr.columns.values, rotation=0);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Model Training<a name=\"Model\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Data Preparation for sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train = (13410, 9)\n",
      "X_test = (5748, 9)\n",
      "y_train = (13410,)\n",
      "y_test = (5748,)\n"
     ]
    }
   ],
   "source": [
    "# train test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X = df.drop('target', axis = 1)\n",
    "y = df['target']\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)\n",
    "\n",
    "print('X_train =', X_train.shape)\n",
    "print('X_test =', X_test.shape)\n",
    "print('y_train =', y_train.shape)\n",
    "print('y_test =', y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 1. Decision Trees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# build tree\n",
    "from sklearn import tree\n",
    "clf = tree.DecisionTreeClassifier(random_state = 42).fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9803877703206563"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check score on training set\n",
    "clf.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6833681280445373"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check score on testing set\n",
    "clf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see that the model has overfit. We will now try to regularize this Decision Tree to eliminate overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=5, estimator=DecisionTreeClassifier(random_state=42),\n",
       "                   n_iter=100,\n",
       "                   param_distributions={'criterion': ['gini', 'entropy'],\n",
       "                                        'max_depth': [2, 5, 10, 15, 25],\n",
       "                                        'max_leaf_nodes': [5, 10, 20, 30, 40],\n",
       "                                        'min_samples_leaf': [1, 5, 10, 15, 20],\n",
       "                                        'min_samples_split': [2, 10, 20, 30,\n",
       "                                                              40]},\n",
       "                   random_state=42, return_train_score='True',\n",
       "                   scoring='accuracy')"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "cv_parameters = {\"criterion\": [\"gini\", \"entropy\"],\n",
    "              \"min_samples_split\": [2, 10, 20, 30, 40],\n",
    "              \"max_depth\": [2, 5, 10, 15, 25],\n",
    "              \"min_samples_leaf\": [1, 5, 10, 15, 20],\n",
    "              \"max_leaf_nodes\": [5, 10, 20, 30, 40],\n",
    "              }\n",
    "\n",
    "dt_classifier = RandomizedSearchCV(estimator=tree.DecisionTreeClassifier(random_state = 42), \n",
    "                                   param_distributions=cv_parameters,\n",
    "                                   cv = 5,\n",
    "                                   scoring = 'accuracy',\n",
    "                                   return_train_score = 'True',\n",
    "                                   n_iter=100,\n",
    "                                   random_state=42\n",
    "                                  )\n",
    "dt_classifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7868754660700968"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt_classifier.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "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>param_min_samples_split</th>\n",
       "      <th>param_min_samples_leaf</th>\n",
       "      <th>param_max_leaf_nodes</th>\n",
       "      <th>param_max_depth</th>\n",
       "      <th>param_criterion</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>mean_train_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>std_train_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>entropy</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.005427</td>\n",
       "      <td>0.001357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>gini</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.005427</td>\n",
       "      <td>0.001357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>entropy</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.005427</td>\n",
       "      <td>0.001357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>gini</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.005427</td>\n",
       "      <td>0.001357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>30</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>25</td>\n",
       "      <td>entropy</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.786875</td>\n",
       "      <td>0.005427</td>\n",
       "      <td>0.001357</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   param_min_samples_split param_min_samples_leaf param_max_leaf_nodes  \\\n",
       "50                       2                     10                    5   \n",
       "33                       2                     20                   10   \n",
       "23                      30                      1                   10   \n",
       "24                       2                     20                    5   \n",
       "68                      30                      5                   10   \n",
       "\n",
       "   param_max_depth param_criterion  mean_test_score  mean_train_score  \\\n",
       "50               5         entropy         0.786875          0.786875   \n",
       "33               5            gini         0.786875          0.786875   \n",
       "23               5         entropy         0.786875          0.786875   \n",
       "24              10            gini         0.786875          0.786875   \n",
       "68              25         entropy         0.786875          0.786875   \n",
       "\n",
       "    std_test_score  std_train_score  \n",
       "50        0.005427         0.001357  \n",
       "33        0.005427         0.001357  \n",
       "23        0.005427         0.001357  \n",
       "24        0.005427         0.001357  \n",
       "68        0.005427         0.001357  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results = pd.DataFrame(dt_classifier.cv_results_).dropna()\n",
    "cv_results = cv_results.iloc[:, [4,5,6,7,8,15,23,16,24]]\n",
    "cv_results.sort_values(by = 'mean_test_score', ascending = False).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Earlier, we had a training score of 0.98 and testing score of 0.68. The model has obviously overfit and after applying Regularization with Cross Validation, we see that the training score and the testing score are both now a modest 0.78. \n",
    "\n",
    "Let's take a look at how the best tree is built."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1MAAAIuCAYAAABac1I3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAADveElEQVR4nOzdd3xN9/8H8NcnexKRSMikiNhCEdTeEaGolQhFVWt9qVGj9PtVtFarqL1DW6ulVTVqVmyxEis2SaxSQRKS9++PJOfnypBE4ma8no/H+/Fwz/mcc97nXrn3vu/nnM9HiQiIiIiIiIgocwz0nQAREREREVFexGKKiIiIiIgoC1hMERERERERZQGLKSIiIiIioixgMUVERERERJQFLKaIiIiIiIiygMUUERERERFRFrCYIiIiIiIiygIWU0RERERERFnAYoqIiIiIiCgLWEwRERERERFlAYspIiIiIiKiLGAxRURERERElAUspoiIiIiIiLKAxRQREREREVEWsJgiIiIiIiLKAhZTREREREREWcBiioiIiIiIKAtYTBEREREREWUBiykiIiIiIqIsYDFFRERERESUBSymiIiIiIiIsoDFFBERERERURawmCIiIiIiIsoCFlNERERERERZwGKKiIiIiIgoC1hMERERERERZQGLKSIiIiIioixgMUVERERERJQFLKaIiIiIiIiygMUUERERERFRFrCYIiIiIiIiygIWU0RERERERFnAYoqIiIiIiCgLWEwRERERERFlAYspIiIiIiKiLGAxRURERERElAUspoiIiIiIiLKAxRQREREREVEWGOk7AaLcxMHB4bCJiYmzvvMgopwTFxd3Myoqqqa+8yAioryPxRTRS0xMTJxv3LhRXN95EFHOcXFx0XcKRESUT/AyPyIiIiIioixgMUVERERERJQFLKaI3pBSCuvWrXurx1y3bh2UUm/1mADg7u6OadOmvfXj5ibZ8Rzcu3cPSins3r07e5IiIiIivWAxRfSGIiIi4OvrCwC4evUqlFI4evSonrOijJgwYQIqVqyYqW2OHDmCTz75JIcyytv++ecfBAQEoHDhwihcuDACAgLw8OHD12534cIFvP/++7CxsYGFhQW8vLwQFhYGAHjw4AEGDhyIcuXKwdzcHC4uLujfvz/u37+vsw93d3copXRi1KhROXGaREREGg5AQfSGHB0d9Z0CvUX29vb6TiHbXL9+Ha6urtm2v27duuH69evYunUrAKBPnz4ICAjA5s2b09zmypUrqFu3Lnr06IG//voLNjY2OHfuHKysrAAAt2/fxq1bt/DNN9+gfPnyuHXrFj755BN07doV27Zt09nXF198gf79+2uPk/dBRESUY0SEwWAkhbOz8215RUJCgkybNk1Kly4tJiYm4uTkJKNGjdLWA5C1a9dq/345GjRoIHv27BEjIyOJiIjQ2e/o0aOlUqVKrx4uVcuXLxdXV1cxNzcXHx8fmT17tiT++f6/TZs2iZeXl5iamoq7u7uMHj1aYmNjRUTk888/Fy8vrxT79fb2loEDB2qPlyxZIp6enmJqaiplypSRGTNmSHx8vLbezc1Npk6dqj2+du2atGvXTqysrMTKykrat28vN27c0NaPHz9eKlSoIAsXLhQXFxcxMzMTPz8/uXv3rtYmMDBQfHx8ZMqUKeLg4CCFChWSkSNHSnx8vIwfP17s7e3FwcFBpkyZopP7w4cPpW/fvmJvby9WVlZSv359OXLkiLZ+6dKlYmlpKTt27JAKFSqIhYWFNGzYUC5fvqytf/X1Wrp06Wtfi1efAwAyf/586dixo1hYWEjJkiVl5cqVOtscPnxYe22qVq0qv/32mwCQXbt2aW3Onj0rrVu3FisrK7G3t5cuXbpo/2fOnTsnFhYWsmzZMq39H3/8IcbGxnLgwIHX5vyye/fuyezZs6VmzZri7u6eqW3TExoaKgBk//792rJ9+/YJADl37lya23Xt2lW6deuWqWP9/vvvopSSR48eactefV3Sk/R3rvf3GwaDwWDk/dB7AgxGborUiqlRo0ZJ4cKFZfHixXLx4kU5cOCAzJkzR1v/cjF1+PBhASBbt26ViIgIuX//voiIeHh4yNdff61tEx8fL87OzvLtt9++ergUDh48KEopmThxopw/f17mzZsntra2OsXU1q1bxdraWpYsWSKXLl2Sv/76S8qWLSvDhg0TkcQv6gAkLCxM2yY8PFwAyKFDh0REZMGCBeLo6Chr166Vy5cvy6ZNm8TBwUG+//57bZuXv7DGx8dL1apVxdvbW44cOSJHjhyRWrVqSfXq1SUhIUFEEospS0tLadCggRw/flz2798v5cuXF19fX22fgYGBYm1tLf369ZOwsDBZvXq1KKWkRYsWMmrUKDl//rz88MMPAkCOHj0qIokFbt26daV169Zy6NAhuXjxoowdO1asra3l9u3El3Dp0qViZGQkTZo0kUOHDsnJkyelatWq0rx5cxERefr0qQwbNkw8PDwkIiJCIiIi5OnTp699PVIrppycnGTlypVy8eJFGTVqlBgbG8u1a9dEROTx48dib28vHTt2lNOnT8vWrVulXLlyOsXU7du3pWjRojJixAgJDQ2VkydPSps2baRmzZpaMTt//nyxtraW8PBwuXPnjjg4OMiECRNem6+ISFxcnPzyyy/y/vvvi4mJiZQvX14mT54s169f19pcu3ZNLC0t041+/fqleYzFixeLlZWV9tonv06WlpayZMmSVLeJj48XKysrmThxorRo0ULs7OykRo0a8uOPP6Z7PmvWrBEzMzN5/vy5tszNzU0cHBzE1tZWqlSpIhMnTtR+THgViykGg8FgZFfoPQEGIzfFq8XU48ePxdTUVH744QdJy8vF1JUrVwSATg+JiMjUqVOlXLly2uMtW7aIiYmJ3Lt3L839Juvatas0bdpUZ1nv3r11iqn33ntP/vvf/+q02bhxo1haWmpfbqtVqyZjx47V1v/vf/+TsmXLao9dXFxkxYoVOvuYOXOmeHp6ao9fLiS2bdsmBgYGcuXKFW19eHi4KKVk+/btIpJYTBkYGGiFhcj/91ZcuHBBRBKLKWdnZ3nx4oXWpnr16lK5cmWdXF4+9s6dO8XS0jJF8VOlShWtaE3ueXq5V2TVqlViYmKiU+xVqFBBMiO1Yurlnsrnz5+Lubm51js1f/58KVy4sDx+/Fhrs3LlSp1iaty4cdK4cWOd4zx48ECn2BURadeundSqVUtat24tderU0XnOUnP8+HEZNGiQ2NnZiYODgwwZMkSOHTuWatvnz5/LxYsX042oqKg0j/XVV19JyZIlUywvWbKkTJo0KdVtIiIiBIBYWFjI9OnT5cSJEzJ9+nQxNDSU3377LdVt/vnnHyldurROj6qIyPTp0+Wvv/6SkydPysKFC6Vo0aLSu3fvVPfBYorBYDAY2RW8Z4ooHaGhoYiNjUWTJk3eaD+BgYEYM2YMDhw4gDp16mDJkiVo164dihYt+tptw8LCtAEuknl7e2Px4sXa42PHjuHw4cP4+uuvtWUJCQl49uwZIiMjUbx4cfj7+2POnDn43//+BwAICgpC9+7dAQB3797FjRs30K9fP517Tl68eAERSTOvEiVKwN3dXVtWqlQplChRAqGhoWjatCkAwMnJSee+nFq1asHAwABhYWEoU6YMAKB8+fIwNDTU2jg4OMDGxkbneA4ODrhz5452vk+fPk1x/1JMTAzCw8O1x6ampvDw8NAelyhRAnFxcfjnn39ga2ub6nllReXKlbV/GxkZwd7eXss1LCwMlStX1rl/x9vbW2f7Y8eOYe/evane4xMeHo6aNWsCABYtWgQPDw+cPXsWp06d0nnOUtO+fXvcuHEDo0ePxvjx42FklPZbvpGREUqXLv36k81GCQkJAAA/Pz8MHToUAFC1alUcPXoUs2fPho+Pj0776Oho+Pr6wsnJCd98843OuuTtgcTXo1ChQujcuTO+/vrrDP2dERERZQWLKaK3wN7eHm3btsWSJUvg4eGBTZs2pXtTfmYlJCRg/Pjx6NSpU6rHBoCuXbtixIgRCA4OhqmpKc6dOwd/f39tewCYN28e6tSp88b5ZHbYdmNj4xTbp7YsOc+EhAQ4ODhg3759KfZVqFAh7d+vFg/JeSXvJ7ukl2tGJCQkwMfHJ9Uh1x0cHLR/nzlzBo8ePQIA3Lp1CyVLlkx3v6tWrcKSJUvw3XffYc2aNfD394e/v3+qRdP169dRvnz5dPfn7++PefPmpbrO0dERd+/ehYhoz7OI4M6dO2kO0mJnZwcjI6MUx/X09MSPP/6osyw6OhqtW7cGAPz2228wMzNLN9datWoBAC5dusRiioiIcgyLKaJ0eHp6wtTUFDt37tR6UdJjYmICAIiPj0+xrm/fvujYsSNKlSoFR0dHrecmIzkcPHhQZ9mrj728vHDu3Ll0exaKFy+Oxo0bIygoCKampvD29kapUqUAJH5hL1GiBMLDw9GjR48M53X79m1cvXpV6526fPkybt++rfPl+NatW7hx4wZcXFwAAIcPH0ZCQgI8PT0zdJzUeHl5ISoqCgYGBto5ZIWJiUmqr1V28vT0xLJly/DkyRNYWloCSP31+/nnn+Hm5paiMEv28OFDBAQE4LPPPsOzZ88QEBCAkydP6hSPr6pXrx7q1auH2bNnY8OGDVixYgUmTpyId999FwEBAejcubNWaJQoUQIhISHpnkt6x/L29kZ0dDSCg4O1gjw4OBhPnjxJs0A3MTHBu+++i/Pnz+ssv3DhAtzc3LTHjx8/RqtWrSAi2Lp1a4ZG6Us+l+LFi7+2LRERUZbp+zpDBiM3RWoDUIwYMUJsbGy0wR0OHTokc+fO1dbjpXumku+X+fLLLyUyMlIePnyotUtISBB3d3cxMTGRL7744tXDpCk4OFiUUjJp0iS5cOGCLFiwQIoWLZpiAAojIyMZN26cnD59WsLCwmTt2rUyfPhwnX0tX75c7OzspESJEjqDaIiILFy4UMzMzGTGjBly7tw5OX36tCxfvlznfpeX7xdKSEiQqlWrSp06dbQBKGrXrp3qABSNGjWSEydOyIEDB6RixYri4+Oj7TN5NL+X+fj4SGBgoM6yWrVqaQNqJCQkSL169aRixYqyZcsWuXz5shw4cEC++OIL2bt3r4j8/2h+L9u1a5cA0EYTDAoKEnNzczl27JjcvXtXYmJiXvt6pHbPVPLrn1qbx48fi52dnXzwwQdy5swZ2bZtm3h6eurcM3Xr1i2xt7eX9u3by8GDByU8PFy2b98uffv2lX///VdERDp37ixeXl4SFxcnMTExUqlSJfH3939tvq+6efOmTJ48WTw9PaVUqVKZ3j49LVu2lIoVK8qBAwe017pNmzY6x/bw8JANGzZoyzZu3CjGxsYyf/58uXjxoixYsECMjIy0e6b+/fdfqV27tpQvX14uXLigDRYSERGhDTBx4MABmTFjhpw4cUIuX74sP/30k5QoUULatm2bap68Z4rBYDAY2RV6T4DByE2RWjEVHx8vkydPlpIlS4qxsbE4OzvL6NGjtfWvfplOHgbcwMBAGjRooLOvL7/8UpRSOoM2ZMSSJUu0ocVbtmwp33//fYqh0f/880+pV6+emJubi7W1tVSvXl1nJD6RxC/2FhYWYmxsnOrgF6tXr5Zq1aqJqamp2NjYSN26dWXNmjXa+tSGRvfz89OGRm/Xrl2qQ6PPnz9fnJ2dxczMTNq2bSt37tzR2mSlmBJJ/JI9aNAgcXJy0l6Xzp07y6VLl0QkY8VUTEyMdOjQQWxsbN5oaPT0iimRxBEZq1WrJiYmJlK5cmXZtGlTiqHRL1y4oOViZmYmZcuWlQEDBkhsbKysWLFCzM3NdUZjPH36tJiZmem8PpkVGhqa5W1T8+DBA+nevbtYW1uLtbW1dO/eXf755x9tffIALa8+z0uXLpUyZcqImZmZVKpUSVavXq2tS37NUovk5+/YsWNSq1YtKVy4sJiZmYmHh4eMHz9enjx5kmqeLKYYDAaDkV2hRFK/uZyoIHJxcbl948aNHLsuqH///rh06RK2b9+eU4fIVSZMmIB169bhzJkz+k6FSOPi4hJx48aNEvrOg4iI8j7eM0X0Fjx69AihoaFYsWIFfv75Z32nQ0RERETZwEDfCRAVBH5+fmjSpAk+/PDDFMM9t2rVClZWVqnGpEmT9JRxwRQUFJTma1GhQgV9p0dERES5DC/zI3pJTl/ml5pbt27h2bNnqa6ztbXN1vmQKH2PHz9GVFRUquuMjY11RpijvIuX+RERUXbhZX5Eeubk5KTvFCiJtbU1rK2t9Z0GERER5RG8zI+I9KZhw4YYMGBAprZxd3dPdXJbIiIioreNxRQR6c2GDRswefLkTG1z5MgRfPLJJzmUUaLr16/D19cXlpaWsLOzw6BBgxAXF5fuNn379sU777wDc3Nz2Nvbw8/PD2FhYdr63bt3QymVaqxduxYAkJCQgLZt28LV1RVmZmYoXrw4/P39cevWrRw9XyIiIsoaFlNEpDe2traZvqzO3t4eFhYWOZQREB8fDx8fHzx+/Bj79u3DmjVrsG7dOgwbNizd7WrUqIFly5YhLCwMf/75J0QETZs2xfPnzwEAderUQUREhE58/vnnsLKyQqtWrbT9NG7cGD///DPOnz+P9evX4/Lly2jfvn2OnS8RERFlHQegIHqJPgagyK+ePHmC/v37Y8OGDbC0tMSQIUPw999/w87ODsuWLQOQeJlfxYoVMXv2bACJl/D16dMHN27cwJo1a1CoUCEMHjwYw4cP1/br7u6OAQMG4LPPPsuRvP/44w/4+Pjg2rVrcHFxAQCsWrUKffr0wZ07d1CoUKEM7efUqVOoUqUKzp07Bw8Pj1TblC1bFg0bNsSCBQvS3M+mTZvg5+eHZ8+ewczMLPMnRClwAAoiIsou7JkiohwxbNgw7NmzBxs3bsRff/2FkydPYt++fa/dbubMmahUqRKOHz+OkSNHYsSIEQgODs7wcfft25fm8OYZGXI+ODgYnp6eWiEFAC1atEBsbCyOHTuWoRyePHmCpUuXwtXVFe7u7qm22b17Ny5evIiPPvoozf08ePAAQUFBqFWrFgspIiKiXIij+RFRtouOjsaSJUuwYsUKNGvWDACwePFiODs7v3bb5s2ba4NSDBw4ELNmzcLOnTvh7e2doWPXqFEDISEh6bZJb7j5yMhIODg46Cyzs7ODoaEhIiMj093v3LlzMWLECDx58gQeHh7YuXMnTE1NU227YMECVK1aFTVq1EixbuTIkZg9ezaePn2K2rVr47fffkv3uERERKQf7JkiomwXHh6O58+fo2bNmtoyS0tLVKxY8bXbVq5cWedxiRIlcOfOnQwf29zcHKVLl043cmruru7du+PEiRPYs2cPypYti06dOuHp06cp2t2/fx8bNmxA3759U93P8OHDceLECWzbtg2Ghobw9/cHL8kmIiLKfVhMEVGuYmxsrPNYKYWEhIQMb/+ml/k5OjqmmLj33r17iI+Ph6OjY7rHLly4MMqUKYP69etj3bp1uHDhAtavX5+i3YoVK2BoaIju3bunuh87OzuULVsWzZo1w48//og///wT+/fvz8DZExER0dvEy/yIKNu98847MDY2xpEjR1CqVCkAwNOnT3HmzBm88847OXrsN73Mz9vbGxMnTsTNmze1yxK3b98OU1NTVK9ePcN5iAhEBLGxsSnWLVq0CJ06dULhwoVfu5/kQjK1/RAREZF+sZgiomxnZWWFDz/8ECNHjoSdnR2KFy+OiRMnIiEhAUqpHD128mV+WdW8eXNUqFABPXr0wPTp03H//n0MHz4cffv21UbyO3z4MHr06IEVK1agZs2auHTpEtavX4+mTZvC3t4eN2/exJQpU2Bqaoo2bdro7H///v0IDQ1NdQS/4OBgHD9+HPXq1YONjQ3Cw8Mxbtw4uLu7o169elk+JyIiIsoZLKaIKEdMmzYNT548Qdu2bWFlZYX//Oc/iIqKyvWj0hkaGuL333/HJ598grp168Lc3Bzdu3fH1KlTtTZPnz7F+fPntfuhTE1NsXv3bkyfPh0PHz6Eg4MD6tevj+Dg4BSXBi5cuBCenp6oW7duimObm5tj3bp1+OKLL/DkyRMUL14cLVu2xE8//ZTrnzciIqKCiPNMEb2E80zlnNjYWLi5uWH48OGvnQCXKCdxnikiIsou7Jkiohxx4sQJhIWFoWbNmnj8+DG+/vprPH78GJ07d9Z3akRERETZgsUUEeWYGTNm4Pz58zAyMkLVqlWxd+/eDM01RURERJQXsJgiohxRrVo1HD16VN9pEBEREeUYzjNFRERERESUBSymiCjXu3r1KpRS7OkiIiKiXIXFFBFRNoiIiEC3bt1Qrlw5GBoaomfPninarF27FjVq1ICNjQ0sLS1RtWpVLF++PM19Tp48GUopDBgwQGf5hg0b0KJFC9jb20Mphd27d6fYNjY2FgMHDoSdnR0sLS3Rtm1b3Lx5801Pk4iIiF7CYoqIKBvExsbCzs4Oo0aNQq1atVJtU7RoUYwdOxYHDx7EqVOn0KtXL/Tu3RtbtmxJ0fbgwYNYsGABKleunGLdkydPUKdOHcyYMSPNfIYMGYL169djzZo12LdvH/7991+0adMG8fHxWT9JIiIi0sFiiog0e/fuRe3atWFlZYXChQujZs2aOHPmDADg/v376Nq1K5ydnWFubo4KFSpg6dKlOts3bNgQ/fv3x7Bhw2Brawt7e3t89913iI2NxaeffgobGxu4urpi5cqV2jbJl/CtXr0a9erVg5mZGcqVK4dt27alm2toaCh8fHxgbW2NYsWKoWvXroiMjNTWnz59Gk2aNEGhQoVgZWWFKlWqYNeuXdn4bOlyd3fHrFmz0LNnT9ja2qbapnHjxmjXrh3KlSuHd955B4MHD0blypWxb98+nXaPHj1C9+7dsWTJEhQpUiTFfgICAjB+/Hi0atUq1eM8evQIixcvxtSpU9GsWTN4eXlh5cqVOHXqFHbs2PHmJ0tEREQAWEwRUZIXL17Az88P9erVw8mTJ3Ho0CEMGTIEhoaGAICYmBh4eXnht99+w9mzZzF48GD069cPO3fu1NlPUFAQrK2tcejQIYwaNQpDhgxBu3btULZsWRw9ehSBgYHo06cPIiIidLYbMWIEBg0ahJCQEDRr1gx+fn64detWqrlGRESgfv36qFixIg4fPowdO3YgOjoafn5+SEhIAAB069YNxYsXx+HDhxESEoIJEybAzMwszfOfNGkSrKys0o1Xi543ISLYuXMnzp8/j/r16+us++ijj9CxY0c0atQoS/s+duwYnj9/jubNm2vLXFxc4OnpiQMHDrxR3kRERPQSEWEwGEnh7Ox8Wwqo+/fvCwDZvXt3hrfp3Lmz9O7dW3vcoEEDqV27tvY4ISFB7OzsxNfXV1sWFxcnxsbGsnbtWhERuXLligCQiRMnam3i4+OlTJkyMmbMGJ02R44cERGRcePGSePGjXVyefDggQCQQ4cOiYiItbW1LFu2LMPncv/+fbl48WK68fTp0wzty8fHRwIDA1Nd9/DhQ7G0tBQjIyMxNTWVRYsW6axfsGCBeHl5SVxcnIgkPqeffvppqvu6e/euAJBdu3bpLA8KChJDQ0NJSEjQWd6oUSP56KOPMnQO+VnS37ne328YDAaDkfeD80wREQDA1tYWPXv2RIsWLdCkSRM0adIEHTt2hKurKwAgPj4eU6ZMwU8//YRbt24hNjYWcXFxaNiwoc5+Xr7HRymFYsWKoVKlStoyY2NjFClSBHfu3NHZztvbW/u3gYEBatWqhdDQ0FRzPXbsGPbu3QsrK6sU68LDw1GzZk0MHToUffr0wfLly9GkSRN06NAB5cqVS/f807o8LztZW1sjJCQE0dHR2LlzJ4YOHQp3d3c0adIE58+fx+jRo7F//34YGxvneC5ERET0ZniZHxFpli5dikOHDqF+/frYtGkTPDw88OeffwIApk2bhunTp2P48OHYuXMnQkJC0K5dO8TFxens49UiQCmV6rLky/GyIiEhAT4+PggJCdGJixcvok2bNgCACRMmIDQ0FO3atcOBAwdQuXJlLFmyJM19vq3L/AwMDFC6dGlUrVoVw4YNQ6dOnTBp0iQAQHBwMO7du4cKFSrAyMgIRkZG2LNnD+bOnQsjIyPExsZm6BiOjo6Ij4/HvXv3dJZHRUXB0dHxjc+BiIiIErFnioh0VKlSBVWqVMHIkSPRqlUrLF++HC1atMD+/fvh6+uLgIAAAImXCF+4cAE2NjbZctyDBw+icePG2r4PHz6Mjh07ptrWy8sLP//8M9zc3NLtwSlTpgzKlCmDQYMGoX///li0aBE+/PDDVNt+/PHH+OCDD9LN0cnJKYNnk3EJCQlakdSuXTvUqFFDZ32vXr1QpkwZjB49GiYmJhnaZ/Xq1WFsbIzt27ejW7duAICbN28iLCwMderUyd4TICIiKsBYTBERAODKlSuYP38+2rZtCycnJ1y+fBmnTp1C//79AQBly5bFTz/9hP3798POzg7ff/89rly5gmrVqmXL8X/44QeULVsWlSpVwty5c3Ht2jXt2K/69NNPsXDhQnTu3BkjR46Evb09Ll++jJ9//hnTp0+HkZERPvvsM3Tq1Anu7u6IiorC/v370xyyHMiey/xCQkIAAP/++y8MDAwQEhICExMTlC9fHgDw1VdfoVatWihVqhRiY2OxZcsWrFy5Et9//z0AwMbGJkVxamlpCVtbW1SsWFFb9uDBA1y/fh0PHz4EAFy6dAk2NjZwdHSEo6MjChcujN69e2PEiBEoVqwYihYtiqFDh6Jy5cpo2rTpG50jERER/T8WU0QEALCwsMCFCxfQqVMn3Lt3Dw4ODujevTtGjhwJABg7diyuXLmCVq1awdzcHD179kT37t3TvK8ps6ZMmYIZM2bg+PHjcHNzw8aNG+Hs7Jxq2xIlSuDvv//G559/jpYtWyImJgaurq5o3rw5TE1NAQD//PMPevbsiYiICBQtWhRt2rTBtGnTsiXXtLxaWG7evBlubm64evUqACA6Ohr9+/fHzZs3YW5ujnLlymHFihXo2rVrpo6zadMm9OrVS3vct29fAMD48eMxYcIEAMC3334LIyMjdO7cGc+ePUOTJk2wYsUKbXRGIiIienNKRPSdA1Gu4eLicvvGjRvF9Z1HQXL16lWULFkSR44cSXGJG1FOcHFxibhx40YJfedBRER5HwegICIiIiIiygIWU0RERERERFnAe6aISK/c3d3By42JiIgoL2LPFBERERERURawmCIiIiIiIsoCFlNEpGnYsCEGDBig7zRea8KECVBKQSmFKVOm6DudAuPq1ava8/7yvFdEREQFFYspIsqTPDw8EBERgYEDB2rLNmzYgBYtWsDe3h5KKezevTvFdrGxsRg4cCDs7OxgaWmJtm3b4ubNmzptrl+/Dl9fX1haWsLOzg6DBg1CXFycTpvVq1ejatWqsLCwgKOjI/z9/REZGamtP3v2LDp27IhSpUpBKaXN/5RZc+bMQeXKlVGoUCEUKlQI3t7e+P3333XajBs3DuXKlYOlpSWKFCmCJk2a4MCBA9r6Bw8eYODAgShXrhzMzc3h4uKC/v374/79+zr7OX78OJo1awYbGxsULVoUH330EaKjo7X1Li4uiIiIwLBhw7J0LkRERPkNiykiypOMjIzg6OgIS0tLbdmTJ09Qp04dzJgxI83thgwZgvXr12PNmjXYt28f/v33X7Rp0wbx8fEAgPj4ePj4+ODx48fYt28f1qxZg3Xr1ukUEH///TcCAgIQGBiIs2fP4pdffkFoaCi6d++utXn69Cnc3d0xceJElCxZMsvn6ezsjK+//hrHjx/H0aNH0bhxY7Rr1w6nTp3S2nh4eGDOnDk4ffo09u/fj5IlS6Jly5aIiooCANy+fRu3bt3CN998g9OnT2PVqlXYu3evzmTBt2/fRtOmTVGqVCkcOnQIW7duxdmzZ9GzZ0+tjaGhIRwdHWFlZZXl8yEiIspXRITBYCSFs7PzbcmD5s+fL8WKFZMXL17oLO/atav4+vqKiMilS5ekbdu24uDgIBYWFlKtWjXZvHmzTvsGDRrIp59+qj12c3OTqVOnptsmNjZWRowYIU5OTmJubi41atSQrVu3Zvcp6hg/frxUqFAhzfV3794VALJr1y6d5Q8fPhRjY2NZtWqVtuz69euilNJy3rJliyil5Pr161qblStXiqmpqTx69EhERKZOnSqurq46+16yZIlYWlqmmk+FChVk/PjxmTnFdBUpUkTmzZuX5vpHjx4JgHRfh99//12UUto5zZ8/X4oWLarzf+jUqVMCQC5evKiz7eue/9wu6e9c7+83DAaDwcj7wZ4ponygU6dOePToEbZv364ti46Oxq+//gp/f3/tcatWrbB9+3acPHkSHTp0wPvvv49z58690bF79eqFPXv2YPXq1Thz5gwCAwPh6+uLkydPprnNpEmTYGVllW7s27fvjfJKzbFjx/D8+XM0b95cW+bi4gJPT0/tsrjg4GB4enrCxcVFa9OiRQvExsbi2LFjAIC6desiIiICmzdvhojg3r17+PHHH9G6detsz/ll8fHx+PHHHxEdHY06deqk2iYuLg4LFixAoUKFULVq1TT39e+//8LU1BQWFhYAEi9/NDY2hqGhodbG3NwcALB///7sOwkiIqJ8hPNMEeUDRYoUQevWrREUFISWLVsCAH755RcYGRmhbdu2AIAqVaqgSpUq2jZjxozB5s2bsW7dOowdOzZLxw0PD8eaNWtw9epVuLq6AgAGDBiAHTt2YP78+Zg7d26q23388cf44IMP0t23k5NTlnJKT2RkJAwNDWFnZ6ez3MHBQbvfKTIyEg4ODjrr7ezsYGhoqLXx9vbGjz/+iO7du+PZs2d48eIFmjVrhuXLl2d7zgBw+vRpeHt7IyYmBlZWVti4cSMqVaqk0+a3335Dly5d8PTpUxQvXhzbt29PcR7JHj58iHHjxqFv374wMkr8GGjcuDGGDh2KKVOmYOjQoXjy5AlGjRoFAIiIiMiR8yIiIsrr2DNFlE/4+/vjl19+wdOnTwEAQUFB6NChA8zMzAAk3k80YsQIlC9fHkWKFIGVlRWOHj2K69evZ/mYx48fh4igfPnyOr1Kv//+O8LDw9PcztbWFqVLl043kntFcqPQ0FAMHDgQ48aNw7Fjx7B161ZERkaiX79+OXI8Dw8PhISE4NChQ+jfvz8CAwNx5swZnTaNGjVCSEgIDhw4gJYtW+KDDz5ItQiKjo6Gr68vnJyc8M0332jLK1SogOXLl+Pbb7+Fubk5HB0dUbJkSTg4OMDAgB8VREREqWHPFFE+4ePjAyMjI/z6669o0qQJduzYgT///FNb/9lnn2Hr1q2YNm0aypQpAwsLC/To0SPFKHUvMzAwgIjoLHv+/Ln274SEBCilcOTIERgbG+u0S68YmjRpEiZNmpTu+fzxxx9477330m2TWY6OjoiPj8e9e/dgb2+vLY+KitKO5ejoiL///ltnu3v37iE+Ph6Ojo4AgMmTJ6NmzZoYPnw4AKBy5cqwtLTEe++9h0mTJsHZ2Tlb8zYxMUHp0qUBANWrV8eRI0cwc+ZMLF68WGtjaWmpFaK1a9dGmTJlsGjRIowbN05rEx0drV2K+Ntvv2mFdrJu3bqhW7duiIqKgqWlJZRSmDFjBkqVKpWt50NERJRfsJgiyidMTU3RqVMnBAUF4d69e3B0dETDhg219fv370ePHj3QoUMHAEBMTAzCw8NRtmzZNPdpb2+v07sRExODc+fOoVq1agCAatWqQUQQGRmJRo0aZThXfV3mV716dRgbG2P79u3o1q0bAODmzZsICwvT7kHy9vbGxIkTcfPmTa0o2r59O0xNTVG9enUAiSP1vXxvEQDtcUJCQrbn/aqEhATExsZmqs3jx4/RqlUriAi2bt2a7oh8yZcHLlmyBGZmZmjWrFn2JE5ERJTPsJgiykf8/f3RpEkTXLlyBV27dtW5PKts2bLYuHEj/Pz8YGxsjC+//BIxMTHp7q9x48ZYsmQJ2rZtC3t7e3z11Vd48eKFzj67d++Onj17Yvr06fDy8sKDBw+we/dulCpVCu+//36q+7W1tYWtrW32nPRLHjx4gOvXr+Phw4cAgEuXLsHGxgaOjo5wdHRE4cKF0bt3b4wYMQLFihVD0aJFMXToUFSuXBlNmzYFADRv3hwVKlRAjx49MH36dNy/fx/Dhw9H3759UahQIQCAr68v+vbtix9++AEtWrRAREQEhgwZAi8vL+3esbi4OISGhgJILEIjIyMREhICKysrrZcpI0aNGgUfHx+4uLjg8ePHWL16NXbv3q3NNfXvv//im2++ga+vL4oXL467d+9izpw5uHnzplawPn78GM2bN8e///6LX375BU+ePMGTJ08AJL4WJiYmAIDZs2fD29sb1tbW2L59O4YPH44pU6bAxsbmzV4YIiKi/ErfwwkyGLkp8urQ6MkSEhLEzc1NAMjJkyd11l29elWaNGkiFhYW4uTkJFOnThUfHx8JDAzU2rw67PmjR4+kS5cuUqhQISlRooTMmTMnRZu4uDgZP368lCxZUoyNjcXBwUF8fX3l6NGjOXaeaQ3NvXTpUgGQIl4eljwmJkYGDBggtra2Ym5uLm3atNEZBl1E5Nq1a+Lj4yPm5uZia2srAwcOlJiYGJ02s2bNkvLly4u5ubk4OjpKt27d5MaNG9r6K1eupJpLgwYNUuR75cqVNM81MDBQXF1dxcTEROzt7aVJkyY6Q54/efJE2rVrJ8WLFxcTExMpXry4tG3bVg4ePKi12bVrV6q54JXh4wMCAsTW1lZMTEykcuXKsmLFikw9/3kFh0ZnMBgMRnaFEpFX6yuiAsvFxeX2jRs3ius7D0rfhAkTsG7duhSDMOQ148ePx7p163Dy5EltVL28IK8//y4uLhE3btwooe88iIgo7+MQTUSUJ4WFhcHKygozZszQdypZtmXLFsyZMyfPFFLXr1+HlZXVawcPISIiKijYM0X0EvZM5Q0PHjzAgwcPACTOAcV7et6OFy9e4OrVqwASBzx5eWLjvIQ9U0RElF3yxs+hREQvyakBLCh9RkZGmRo8g4iIKL/jZX5ERERERERZwGKKiIiIiIgoC1hMERERERERZQHvmSJ6SVxc3M28elN9fiAiFiJibGBg8EjfueQEEbEEYKiU+lffuRRkcXFxN/WdAxER5Q8czY+IcgWlVC0AMwE0FpEYfeeTE5RSCsA6AD+JyM/6zoeIiIjeDIspItI7pZQdgB0A2ovIFX3nk5OUUoUB7ALQTUTO6TsfIiIiyjreM0VEeqWUMgSwEsCY/F5IAYCIPALQE8BKpZSVntMhIiKiN8Biioj0bSyA4yLyu74TeVtE5BSA2QDmJ136R0RERHkQiyki0hulVAsAdQF8oe9c3jYRWQ7gCYD++s6FiIiIsob3TBGRXiilXAFsAtBcRO7oOx99UEqZAfgLwBAROazvfIiIiChz2DNFRG+dUsoUwGoAnxTUQgoAkkYt7I7Ey/2K6jsfIiIiyhwWU0SkD9MBrBORA/pORN+SBt0Yi8QBKfieTERElIfwg5uI3iqlVDcAxQB8p+9ccoukwTdOILGoIiIiojyC90wR0VujlCqPxGHQG4nIv/rOJzdJGiL+DwDTRGSbvvMhIiKi12MxRURvhVLKGomT1fYUkTP6zic3UkoVA7ANgK+I3NB3PkRERJQ+XuZHRDkuaS6lhQC+ZSGVtqTBOD4FEKSUMtF3PkRERJQ+FlNE9DYMAPCPiKzSdyK5nYj8DWAjgGn6zoWIiIjSx2KKiHKUUqo2gC4Ahug5lbzkWwDFlVJd9J0IERERpY33TBFRjlFK2QPYDqCdiFzVczp5ilKqEBLvMQsQkVB950NEREQpsWeKiHJE0uh0qwCMZiGVeUmjHfYEsEIpZaXndIiIiCgVLKaIKKeMB3BIRLboO5G8SkROI3E+roVJg3gQERFRLsJiioiynVKqFYBaAL7Udy55nYisBPAQiaP8ERERUS7Ce6aIKFsppdwA/AqgmYjc1Xc++YFSyhTAXwCGichBfedDREREidgzRUTZJulL/2oA/VlIZR8RiQXQHcA8pZSdvvMhIiKiRCymiCg7zQTwk4gE6zuR/CZpEI/RAFYmDe5BREREesZiioiyhVKqOwBbAN/rO5f8Kmkwj6MAxuk7FyIiIuI9U0SUDZRSFQAsB9BIRB7rO5/8LKlXaguAmSKyVd/5EBERFWQspojojSilrAHsBtBDRM7qOZ0C4aXJkNuKyHV950NERFRQ8TI/IsqypLmPFgOYzkLq7Uka3KM/gNVJg34QERGRHrCYIqI3MQjAXRFZre9ECpqkQT7WApih71yIiIgKKhZTRJQlSqk6ADoCGKrvXAqwWQDslFLd9J0IERFRQcR7pogo05RSxQBsA+/Z0buke9Z2AQjkpZZERERvF3umiChTkkaTWwVgJAsp/UsaPTEQwPKkwoqIiIjeEhZTRJRZXwI4ICJ/6jsRSpTUIzUTwKKkQUGIiIjoLWAxRUQZppTyAVAdwP/0nQvpEpEgAPcADNR3LkRERAUF75kiogxRSrkD+AVAUxG5p99sKDVJw6TvBDA8abQ/IiIiykHsmSKi11JKmQFYDaAfC6ncS0RiAXQH8EPSxL5ERESUg1hMEVFGfAtgtYgc0ncilD4RuQZgFIBVSYOFEBERUQ5hMUVE6VJKBQAoBGCOvnOhjBGRrQAOAhiv71yIiIjyM94zRURpUkpVArAUQEMRidZ3PpRxSb1SvwGYJSJ/6DsfIiKi/IjFFBGlSilVCImTwfqLSJi+86HMU0rZAdgBoJ2IXNVzOkRERPkOL/MjohSS5ipaDOAbFlJ5V9JgIR8DCEoa6Y+IiIiyEYspIkrNEAARIvKTvhOhNyMiBwH8iMRBRIiIiCgbsZgiIh1KqXoA2gP4TN+5ULaZDcBGKeWv70SIiIjyE94zRUQapZQDgG0A2ojIDX3nQ9lHKWWFxHvgeonIGX3nQ0RElB+wZ4qIAABKKSMAqwB8xkIq/0kajbEHgOVJg4sQERHRG2IxRUTJ/gtgn4hs13cilDOSBhOZCmBR0iAjRERE9AZYTBERlFK+AKoCmKjnVCiHiciPAKIADNZ3LkRERHkd75kiKuCUUqUAbADQRETu6zsfynlKKRMAOwGMEpG/9Z0PERFRXsViiqgAU0qZIXFQgsEicljf+dDbo5RyBbAJQHMRuaPvfIiIiPIiXuZHVLDNArCChVTBIyLXAYwAsEopZajvfIiIiPIiFlNEBZRSKhCABYB5+s6F9ENEtgH4G8CX+s6FiIgoL+JlfkQFkFKqMoBFABqJyBN950P6o5QyAPAbgDki8ru+8yEiIspLWEwRFTBKqcIA/gLQTUTO6zsf0j+llB2AHQDai8gVfedDRESUV/AyP6ICJGluoSUAprCQomQicg/ARwCCkgYlISIiogxgMUVUsAwDcENE1uo7EcpdkgYhWQXgO33nQkRElFewmCIqIJRS9QH4InEEN6LU/ADASinVQ9+JEBER5QW8Z4qoAFBKOQL4E4CPiNzUdz6UeymlLJE491hvETmt73yIiIhyM/ZMEeVzSikjAEEAhrGQotdJGt2xB4BlSYOVQClVQynlq9/MiIiIch8WU0T530QAu0Rkh74TobxBRM4B+BrA4qRBSwRAe/1mRURElPuwmCLKx5RSbQFUBDBJ37lQ3iIiPwO4BeA/AM4AqKTfjIiIiHIfI30nQEQ5Qyn1DoAvATQRkQR950N5h1KqKoCLAIYjcf6pwwBEKWUmIjH6zI2IiCg3Yc8UUT6klDJH4jDXfUXkgb7zoTynCoADAIYicf6p2UgsrirqMykiIqLchsUUUf70PYBlInJU34lQ3iMiywHUBvAYwE8ATgGoBaCGPvMiIiLKbVhMEeUzSqleAEwALNB3LpR3icgzEZkD4F0AewAUAeCv36yIiIhyF84zRZSPJN3rMh9AIxF5qud0KB9JGmLfSUSu6TsXIiKi3II9U0R5mFLqPaWUYdK/bQAsARDAQoqym4i8YCFFRESkiz1TRHmYUuowgPcAxAHYACBIRNbpN6u8ycHB4bCJiYmzvvMgyivi4uJuRkVF1dR3HkRE+sSh0YnyKKWUKQCISKxSagSAKyykss7ExMT5xo0bxfWdB1Fe4eLiou8UiIj0jsUUUd5VEcAZpVQDAK0BNNNzPkREREQFCu+ZIsq7vABcAvAdgO4Amiil9iilOus3LSIiIqKCgT1TRHlXdSQOW70OwFoAZwH0FpFLes2KiIiIqIBgzxRR3tUWgBMABwBdRKQvC6ncYcKECahYsaK+0ygQGjZsiAEDBug7DSIiKqBYTBHlXasAeInIQBG5ru9kKPfbvXs3lFK4d++evlPJNhs2bMDkyZP1nUautH79epQvXx6mpqYoX748Nm7cmG77q1evQimVIrZu3fqWMiYiyntYTBHlUSIyQkRu6zuPgiQuLk7fKVCS5NfC1tYW1tbWes7mzcXExODu3bvZtr/g4GB07twZ3bt3R0hICLp3745OnTrh0KFDr91269atiIiI0KJx48bZlhcRUX7DYoqIKA0NGzZE//798dlnn8He3h5169YFAISGhsLHxwfW1tYoVqwYunbtisjIyHT3tXTpUpQvXx5mZmYoW7YsZs6ciYSEBABAt27d0KFDB532CQkJcHFxwYwZMwAAIoJvvvkG77zzDszNzVGpUiWsWrVKa5/cq7B+/Xo0a9YMFhYWKF++PLZv366tb9SoEQDA3t4eSin07Nnztc/B647bu3dvVKhQAc+ePQMAxMfH47333kObNm108lq9ejXq1asHMzMzlCtXDtu2bdM5zuue0549e6JNmzb4+uuv4ezsDGdnZ+01evkyv7i4OIwcORLOzs6wsLDAu+++iz///FNbn9w7t3PnTtSqVQsWFhaoUaMGjh8/rpPPwYMH0bhxY1haWqJw4cJo3Lgxbt++naHnJDP+/vtvfPTRR3B0dMSWLVuytI/UfPvtt2jUqBHGjBkDT09PjBkzBg0bNsS333772m2LFi0KR0dHLUxMTLItLyKi/IYDUFCWcIJTym65dQLQVatW4aOPPsK+ffsgIoiIiED9+vXRu3dvTJs2Dc+fP8eYMWPg5+eH4OBgGBik/I1q4cKF+OKLL/D999+jevXqOHPmDPr27QtjY2MMGDAA/v7+eP/99/Ho0SMULlwYALBnzx5ERESga9euAICxY8di3bp1mDNnDjw8PBAcHIy+ffuiSJEi8PHx0Y41ZswYTJ06FXPnzsXEiRPRpUsXXLt2DS4uLli/fj06dOiAs2fPwtbWFubm5q89/9cdd9asWahWrRo+++wzzJkzB1999RUuXryIU6dO6exnxIgRmDFjBipXrow5c+bAz88Ply5dgpOTU4af0z179qBw4cLYunUr0ppwvlevXggPD8fq1avh7OyMLVu2wNfXF0eOHEGVKlW0dp9//jm+/vprFC9eHIMHD0b37t0RGhoKpRROnjyJRo0aISAgADNmzICpqSn27t2LFy9eZOq1SMuVK1ewcuVKrFixArdv34afnx/WrFmD5s2ba20+/vjj1xZooaGhcHV1TXVdcHAwBg4cqLOsRYsWmD179mvze//99xETE4MyZcrgP//5Dzp27PjabYiICiwRYTAyHc7OzreFKBsl/Z/KVf+nGzRoIJUqVdJZNm7cOGncuLHOsgcPHggAOXTokIiIjB8/XipUqKCtd3FxkRUrVuhsM3PmTPH09BQRkefPn0uxYsVk0aJF2vrevXtLs2bNREQkOjpazMzMZO/evTr7GDx4sLRq1UpERK5cuSIAZN68edr6mzdvCgDZt2+fiIjs2rVLAMjdu3dfPdVUZeS4IiKHDx8WY2NjGTdunBgZGcmWLVu0dcl5TZw4UVsWHx8vZcqUkTFjxohIxp7TwMBAsbOzk5iYGJ12DRo0kE8//VRERC5duiRKKbl27ZpOGz8/P+nfv7/Oc7B161Zt/f79+wWA3LhxQ0REunXrJrVr136j5+RVjx8/lsWLF0v9+vXF0NBQGjVqJEuWLJF///031fZRUVFy8eLFdOP58+dpHs/Y2FiWL1+us2z58uViYmKS5jZ3796VadOmSXBwsBw5ckTGjRsnBgYGsnLlylTb6/tvlsFgMHJDsGeKiCgd1atX13l87Ngx7N27F1ZWVinahoeHo2ZN3c61u3fv4saNG+jXrx/69++vLX/x4gVEEntXjIyM0LlzZwQFBaF3796IjY3F+vXr8d133wFI7IGIiYlBy5YtoZTS9vH8+XO4u7vrHK9y5crav0uUKAEAuHPnThbOPOPHfffddzFmzBhMmDABn3zyCVq1apViX97e3tq/DQwMUKtWLYSGhgLI+HNasWJFmJqappnv8ePHISIoX768zvLY2NgU9/2k9Tw5OzvjxIkTaN++farHyMxr8bJ169ahd+/e8PT0xLFjx3R6yVJTrFgxFCtWLN022c3Ozg7Dhg3THteoUQP37t3DN998A39//7eaCxFRXsFiiogoHZaWljqPExIS4OPjg2nTpqVo6+DgkGJZ8n1R8+bNQ506ddI8jr+/P7y9vXHr1i0cOnQIcXFxeP/993X2sXnz5hSXdRkbG6f5OPnLfvL2mZXR44oI9u/fD0NDQ4SHh0NEdAqNjBwnI8/pq69FavtRSuHIkSMpnpdXL2nM6vOUmdfiZX5+fvj222+xfPly1KxZEz4+PvD394ePj0+qBeKbXubn6OiIqKgonWVRUVFwdHRMd5+vqlWrFpYuXZqpbYiIChIWU1SgNGzYEBUrVszQfQPJ3N3dMWDAAHz22Wc5mBnlFV5eXvj555/h5uaW7pfnZA4ODihRogTCw8PRo0ePNNvVrFkTpUuXxpo1axAcHAw/Pz+tpyZ5eOtr16690chqyQMJxMfHZ6h9Ro87Y8YMHD9+HHv37kXr1q3x/fffY9CgQTptkgd0ABKLr8OHD2v34mT2OU1LtWrVICKIjIzUBtvI6n7++uuvVNdl9bUoUqQIBg8ejMGDB+PMmTNYvnw5BgwYgD59+qBTp04ICAhA3bp1tcLuv//972vfc5J71FLj7e2N7du3Y/jw4dqy7du3p1vQpyYkJATFixfP1DZERAWKvq8zZOTNyKv3TN2/fz/NexTScufOHXny5EkOZZTo2rVr0qZNG7GwsJCiRYvKwIEDJTY2NkPbJiQkSMuWLQWArF27Vmfd+fPnxc/PT4oWLSpWVlZSq1Yt+eOPP3LiFN6Yvu+/SOueqeT7cZLdunVL7O3tpX379nLw4EEJDw+X7du3S9++fbX/W6/eM7Vw4UIxMzOTGTNmyLlz5+T06dOyfPlymTRpks6+v/zyS/Hw8BBTU1P5/fffddaNGTNGbG1tZfHixXLx4kU5ceKE/PDDDzJ//nwR+f97k44cOaKz3cv/L27evClKKVm8eLHcuXNHHj9+nObrkdHjhoSEiKmpqaxfv15ERIKCgsTMzEzOnDmjk5ezs7OsXbtWzp07J4MGDRJTU1PtHqWMPKeBgYHi4+OTIr9XX6Pu3buLq6urrF27VsLDw+XIkSMydepULb/U7ht79bk7ceKEmJqaSt++fSUkJETOnTsnCxcu1O7Fet1zklEvXryQP/74Q7p06SLm5uYp7qt7E3///bcYGhrK5MmTJSwsTCZNmiRGRkZy8OBBrc2oUaN07lVbtmyZBAUFSWhoqJw7d06mTp0qxsbGMmPGjFSPoe+/WQaDwcgNofcEGHkz8moxlRu9ePFCKlasKA0aNJBjx47Jtm3bpHjx4jJgwIAMbT916lRp3bp1qsVUmTJlpEWLFhISEiIXL16UYcOGiYmJiVy6dCknTuWN6PuLWUaLKRGRCxcuSIcOHcTGxkbMzMykbNmyMmDAAK0AfrWYEhFZvXq1VKtWTUxNTcXGxkbq1q0ra9as0WkTHh4uAKRYsWIpBhdISEiQWbNmiaenp5iYmIidnZ00bdpUtm3bJiIZK6ZERP773/+Ko6OjKKUkMDAwxbm9Kr3jPnv2TCpUqCC9evXS2cbf318qV64sMTExWl6rVq0Sb29vMTU1lbJly+oMUpGR5zSjxVRcXJyMHz9eSpYsKcbGxuLg4CC+vr5y9OhREclYMSUism/fPnnvvffEzMxMChcuLE2aNJHbt29n6LXIikePHsmtW7eyvH1q1q5dKx4eHmJsbCzlypXTCspkgYGB4ubmpj1etmyZeHp6ioWFhVhbW0v16tXTHHxCRP9/swwGg5EbQu8JMPJm5MZiKjo6WgICAsTS0lKKFSsmkyZNEh8fH50vjK9+8XJzc5P//e9/8tFHH4m1tbU4OTnJN998o7NfNzc3mTp1ao7lvWXLFlFKyfXr17VlK1euFFNTU3n06FG62x4+fFicnZ0lKioqxZfmu3fvCgD566+/tGXPnz8XAwODFEVXbqDvL2a58f90fpBWkUd5n77/ZhkMBiM3BCftpXxj2LBh2LNnDzZu3Ii//voLJ0+exL59+1673cyZM1GpUiUcP34cI0eOxIgRIxAcHJzh4+7btw9WVlbpxqRJk9LcPjg4GJ6ennBxcdGWtWjRArGxsTh27Fia2z1+/BjdunXDggULUh31q2jRovD09MTKlSsRHR2N+Ph4LFiwANbW1trks0RERESUdRyAgvKF6OhoLFmyBCtWrECzZs0AAIsXL4az8+vnFW7evDkGDBgAABg4cCBmzZqFnTt36gzlnJ4aNWogJCQk3Ta2trZprouMjEwxCpydnR0MDQ0RGRmZ5nYff/wxWrZsmeow1EDiCGXbt29H+/btUahQIRgYGMDW1hZ//PEHbygnAMD169dTDCP+svRGiyMiIiIWU5RPhIeH4/nz5zpz/FhaWqJixYqv3fbl+WaAxBGyMjMvj7m5OUqXLp3xZLPBypUrcfLkSRw9ejTNNiKCTz75BEWLFsW+fftgbm6ORYsWoUOHDjhy5AicnJzeYsaUG5UoUSLdHwLSGy0uo9zd3SEib7wfIiKi3IjFFBV4rw7FrJTK1Lw8+/btS7N3KNno0aMxevToVNc5Ojri77//1ll27949xMfHpzknzM6dOxEaGppiktPOnTvD29sb+/fvx19//YXNmzfjwYMHsLGxAQDMnTsX27dvx9KlSzF27NgMniHlV0ZGRm/9hwAiIqL8hMUU5QvvvPMOjI2NceTIEZQqVQoA8PTpU5w5cwbvvPNOjh77TS/z8/b2xsSJE3Hz5k3tssTt27fD1NQU1atXT3Wbr776KsUcNJUqVcK0adPg5+cHIPH8AcDAQPfWSAMDgyxP4koFD+dmIyIiShsHoKB8wcrKCh9++CFGjhyp9dr06dMHCQkJ2iSYOSX5Mr/0Ir1iqnnz5qhQoQJ69OiBEydOYMeOHRg+fDj69u2LQoUKAQAOHz6McuXK4fDhwwAAJycnVKxYUScAwMXFRSsmvb29YWtri169euHkyZO4cOEChg8fjsuXL6NNmzY5+pxQ/rFhwwZMnjw5U9scOXIEn3zySQ5llOj69evw9fWFpaUl7OzsMGjQIMTFxaXZ/sGDBxg4cCDKlSsHc3NzuLi4oH///rh//75OuwsXLqBdu3aws7ODtbU1ateuja1bt6bY36pVq1C1alWYmZnBzs4u3QmZiYgo/2LPFOUb06ZNw5MnT9C2bVtYWVnhP//5D6KiomBmZqbv1NJlaGiI33//HZ988gnq1q0Lc3NzdO/eHVOnTtXaPH36FOfPn9d6mzLCzs4OW7duxZgxY9C4cWM8f/4cnp6e+OWXX+Dl5ZUTp0L5UHo/BKTF3t4+BzL5f/Hx8fDx8dHuB7x//z4CAwMhIvj+++9T3eb27du4desWvvnmG5QvXx63bt3CJ598gq5du2Lbtm1auzZt2qBUqVLYuXMnLC0tMW/ePPj5+SE0NFTr5Z41axYmT56MqVOnonbt2nj27BkuXLiQo+dMRES5lL7HZmfkzcgLc/LExMSIg4ODTJs2Td+pUAboe86avPB/OrsVxLnZXvb777+LUkrbJiNzs/3zzz9iYWHxRhP05hf6/ptlMBiM3BC8zI/yjRMnTmD16tW4dOkSTpw4gcDAQDx+/BidO3fWd2pEuVJBm5vtVf/++y9MTU1hYWEBIGNzs23btg3x8fGIiopC+fLl4eTkhPbt2+Py5csZPi4REeUfvMyP8pUZM2bg/PnzMDIyQtWqVbF3794MzTVFVNAUxLnZXvbw4UOMGzcOffv2hZFR4kdhRuZmu3z5MhISEjBx4kR8++23sLW1xX//+180atQIYWFhWmFGREQFA4spyjeqVauW7rxLRPT/CtrcbC+Ljo6Gr68vnJyc8M0332jLRV4/N1tCQgKeP3+OWbNmoXnz5gCAoKAgODo6YvPmzewJJyIqYHiZHxERZUp2zM32Jpf5OTo6IioqSmfZ6+ZmSxYdHY3WrVsDAH777TedAWqS52Zbs2YN6tatCy8vL8ydOxeWlpZYunQpAGg9VOXLl9e2K1y4MEqUKIHr169n+DkgIqL8gT1TRKm4evUqSpYsiSNHjqBGjRr6Toco2xW0udkA4PHjx2jVqhVEBFu3bk0x6XVG5mZLvnfq/Pnz2rGjo6MREREBNze3dM+JiIjyH/ZMEeVRGzZsQPPmzWFvbw9ra2vUqlULmzZt0mmzdu1a1KhRAzY2NrC0tETVqlWxfPlynTbu7u5QSqUIHx8frc3evXvRtm1bODk5QSmFZcuWvY1TpBxU0OZme/z4MZo3b45//vkHy5Ytw5MnTxAZGYnIyEhtfqqMzM1WtmxZ+Pn5YfDgwfj7778RGhqKXr16oVixYpy/jYioAGIxRZRH7dmzB40bN8bvv/+OEydOoHXr1mjfvr3OaGxFixbF2LFjcfDgQZw6dQq9evVC7969sWXLFq3NkSNHEBERocXx48ehlMIHH3ygtYmOjkbFihXx3Xffwdzc/K2eJ+WcadOm4b333kPbtm3RqFEjVK5cGTVq1Mgzc7NZWFigbt266Ny5Mzp06IBp06ZpbV6dm+3YsWM4ePAgQkNDUbZsWRQvXlyLAwcOAPj/udmio6PRuHFj1KhRA3v37k0xN9vKlSvh7e0NX19f1K1bFzExMdi5cycHnyAiKoj0PTY7I29Gds3Js2fPHqlVq5ZYWlpKoUKF5N1335XTp0+LiMi9e/ekS5cu4uTkJGZmZlK+fHlZsmSJzvYNGjSQjz/+WIYOHSpFihQROzs7+fbbbyUmJkY++eQTKVy4sLi4uMiKFSu0ba5cuSIAJCgoSOrWrSumpqbi4eEhf/75Z4o2R44c0ZadPXtWWrduLVZWVmJvby9dunSRiIgIbf2pU6ekcePGYm1tLZaWllK5cmWd+WrehnfffVeGDh2abptq1arJqFGj0lw/ceJEKVy4sDx9+jTV9ZaWlrJ06dI3STNV+p6zpiDOM/Uqzs1GmaHvv1kGg8HIDcGeKdKbFy9ewM/PD/Xq1cPJkydx6NAhDBkyBIaGhgCAmJgYeHl54bfffsPZs2cxePBg9OvXDzt37tTZT1BQEKytrXHo0CGMGjUKQ4YMQbt27VC2bFkcPXoUgYGB6NOnDyIiInS2GzFiBAYNGoSQkBA0a9YMfn5+uHXrVqq5RkREoH79+qhYsSIOHz6MHTt2IDo6Gn5+ftq9FN26dUPx4sVx+PBhhISEYMKECen+wj9p0qTX3oSfkTl/Xvb48WMUKVIk1XUigp07d+L8+fOoX79+mm0WL14Mf39/9kAVAJybjYiI6A3pu5pj5M3Ijl/x79+/LwBk9+7dGd6mc+fO0rt3b+1xgwYNpHbt2trjhIQEsbOzE19fX21ZXFycGBsby9q1a0Xk/3udJk6cqLWJj4+XMmXKyJgxY3TaJPdMjRs3Tho3bqyTy4MHDwSAHDp0SERErK2tZdmyZRk+l/v378vFixfTjbR6h1Ize/ZssbKykqtXr+osf/jwoVhaWoqRkZGYmprKokWL0tzHn3/+KQAkJCQkzTbsmco/jh8/LtWrVxcrKyuxsbGRhg0bytGjR/WdFuUR+v6bZTAYjNwQHM2P9MbW1hY9e/ZEixYt0KRJEzRp0gQdO3aEq6srACA+Ph5TpkzBTz/9hFu3biE2NhZxcXFo2LChzn5envNGKYVixYqhUqVK2jJjY2MUKVIkxTw4L08wamBggFq1aiE0NDTVXI8dO4a9e/emGP0LSJyvp2bNmhg6dCj69OmD5cuXo0mTJujQoQPKlSuX7vmnd5N9Zqxfvx7Dhw/HTz/9lGJEMWtra4SEhCA6Oho7d+7E0KFD4e7ujiZNmqTYz8KFC/Huu++iSpUq2ZIX5W6cm42IiOjN8DI/0qulS5fi0KFDqF+/PjZt2gQPDw/8+eefABJvjp8+fTqGDx+OnTt3IiQkBO3atdNG3kqW2pw3bzoPzqsSEhLg4+ODkJAQnbh48aI2gteECRMQGhqKdu3a4cCBA6hcuTKWLFmS5j6z6zK/devWISAgACtWrICvr2+K9QYGBihdujSqVq2KYcOGoVOnTqnO4XPnzh38+uuv6Nu3byaeGSIiIqKCiz1TpHdVqlRBlSpVMHLkSLRq1QrLly9HixYtsH//fvj6+iIgIABA4iWpFy5cgI2NTbYc9+DBg2jcuLG278OHD6Njx46ptvXy8sLPP/8MNze3FIXay8qUKYMyZcpg0KBB6N+/PxYtWoQPP/ww1bYff/yxzoh5qXFyckp3/c8//4zAwEAsX748zdxflZCQgNjY2BTLly1bBlNTU3Tt2jVD+yF6GedmIyKigog9U6Q3V65cwahRo3DgwAFcu3YNu3btwqlTp1C+fHkAifO57Ny5E/v378e5c+cwYMAAXLlyJduO/8MPP2DdunU4f/48hgwZgmvXrqF///6ptv3000/x6NEjdO7cGYcOHcLly5exY8cOfPTRR3j8+DGePXuGTz/9FLt378bVq1dx6NAh7N+/XzuX1Nja2r52rp30BoH48ccf0b17d0yZMgX169fX5sx58OCB1uarr77Cjh07cPnyZYSFhWH69OlYuXIl/P39dfYlIli0aBG6dOmS6qWM0dHRWm9cQkICrl+/jpCQEFy/fv11TzNRrhEREYFu3bqhXLlyMDQ0RM+ePVO0OXv2LDp27IhSpUpBKYUJEyakuq+5c+eiZMmSMDMzQ/Xq1VP0Ivft2xfvvPMOzM3NYW9vDz8/P4SFhem0OX78OJo1awYbGxsULVoUH330EaKjo7PrdImI6C1gMUV6Y2FhgQsXLqBTp04oW7YsAgMD0b17d4wcORIAMHbsWNSsWROtWrVC/fr1YWlpie7du2fb8adMmYIZM2agSpUq2Lp1KzZu3AhnZ+dU25YoUQJ///03DAwM0LJlS1SoUAGffvopTE1NYWpqCkNDQ/zzzz/o2bMnPDw80L59e3h7e2PGjBnZlu+r5s2bhxcvXmDIkCE6c+a8//77Wpvo6Gj0798fFSpUQN26dbF+/XqsWLECH3/8sc6+du/ejYsXL6Z5id/Ro0dRrVo1VKtWDc+ePcP48eNRrVo1fPHFFzl2fkTZLTY2FnZ2dhg1ahRq1aqVapunT5/C3d0dEydORMmSJVNt89NPP2Hw4MEYPXo0Tpw4gTp16qBVq1Y6Py7UqFEDy5YtQ1hYGP7880+ICJo2bYrnz58DAG7fvo2mTZuiVKlSOHToELZu3YqzZ8+mWuAREVEupu8RMBh5M/LyyGepzSFF+qfvkcHyyv9pzs2WPXx8fCQwMDDdNhUqVJDx48enWF6zZk3p06ePzrLSpUunO3/byZMnBYCcO3dORETmz58vRYsWlRcvXmhtTp06JQDk4sWLGT8RPdL33yyDwWDkhmDPFBFRHsG52bJ/brbMiouLw7Fjx9C8eXOd5c2bN8eBAwdS3ebJkydYunQpXF1d4e7uDiCxl8zY2Fh77QBol/Xu378/Z5InIqLsp+9qjpE3I6/8ip8a9kzlTvr+lTsv/J/m3GzZNzdbVnumbt26JQBkz549Osu//PJLKVu2rM6yOXPmiKWlpQAQDw8PnR6nM2fOiJGRkUyePFliY2PlwYMH0qFDBwEgkyZNytA56Ju+/2YZDAYjNwR7pqjAcXd3h4hwxDHKc16em83HxwczZszQuU8nPj4eX331FSpXroyiRYvCysoKGzZsSDFQyNuemy05XFxcACTOzQZAm5utcePG+Oqrr3Du3LnXnv+bDNrytnXv3h0nTpzAnj17ULZsWXTq1AlPnz4FAFSoUAHLly/Ht99+C3Nzczg6OqJkyZJwcHCAgQE/momI8gq+YxMR5SGcm02/l/nZ2dnB0NAQUVFROsujoqLg6Oios6xw4cIoU6YM6tevj3Xr1uHChQtYv369tr5bt26IjIzE7du3cf/+fUyYMAF3795FqVKlcvQciIgo+7CYolylYcOGGDBggL7TeK0JEyZAKQWlFKZMmaLvdPKEhg0bas/Z0aNH9Z1OnpY8L9vu3bvRsGFDLF++HAB05marWrUq3nnnHVy4cCHbjnvw4EHt3yKJc7N5enqm2tbLywtnz56Fm5tbit4ja2trrV3yvGy///47evfujUWLFqV5/I8//jhFcfZq5HSPs4mJCapXr47t27frLN++fTvq1KmT5nbJl4OkNsebg4MDrKys8NNPP8HMzAzNmjXL9ryJiChnsJgiyiIPDw9ERERg4MCB2rKoqCj07NkTJUqUgIWFBVq2bImLFy/qbPdyUZEcXbp00Wnzzz//ICAgAIULF0bhwoUREBCAhw8fZiq/mJgY9OzZE5UrV4axsTEaNmyYbvv9+/fDyMgIFStW1Fm+cOFCvPfeeyhSpAhsbGzQqFGjFDfI7927F23btoWTkxOUUli2bFmK/W/YsAGHDx/O1DmQLs7N9uaX+SUXXf/++y8ePHiAkJAQnUsV4+LitDYxMTGIjIxESEgILl26pLUZOnQoli1bhkWLFiEsLAyDBw/G7du3tSkHLl26hK+//hrHjh3D9evXceDAAXTq1AmmpqZarxwAzJ49G8eOHcOFCxcwZ84cDBgwAJMnT862icmJiCjnGek7AaK8ysjISOeyHhFBu3btYGBggF9++QWFCxfGjBkz0LRpU4SGhsLS0lJr26tXL0yaNEl7/OoXwG7duuH69evYunUrAKBPnz4ICAjA5s2bM5xffHw8zMzMMGDAAGzZsiXdYuyff/5Bjx490KRJkxSjs+3evRudO3fGrFmzYGFhgZkzZ6JFixYICQlBmTJlACTOZ1WxYkX06NEDPXr0SPUYtra2+PfffzOcP6X08txs9+7dg4ODQ4q52a5cuYJWrVrB3NwcPXv2RPfu3dO8rymzkudmO378ONzc3DI0N9vnn3+Oli1bIiYmBq6urmjevDlMTU0BQJubLSIiAkWLFkWbNm0wbdq0bMk1LdWqVdN5vHnzZri5ueHq1asAEud/erlNeHg45s+fjwYNGmD37t0AgM6dO+P+/fuYOHEiIiIiULFiRWzZsgVubm4AAFNTU+zevRvTp0/Hw4cP4eDggPr16yM4OFjnPePw4cMYP348oqOjUa5cOcyfPx8BAQE5ev5ERJTN9D0CBiNvxqsjn82fP1+KFSumM2eKiEjXrl21UcIuXbokbdu2FQcHB7GwsJBq1arJ5s2bddo3aNBAPv30U+2xm5ubTJ06Nd02sbGxMmLECHFychJzc3OpUaOGbN26VXLS+PHjpUKFCjrLzp8/LwAkJCREWxYfHy/29vaycOHCNPN/VWhoqACQ/fv3a8v27dunM0dNZn366afSoEGDNNe3b99eJkyYkOp5vSohIUEcHBxk1qxZqa63tLSUpUuXprouvZEU9T0yWF4YzU9fOAImpUbff7MMBoORG4KX+VG26NSpEx49eqRzH0F0dDR+/fVX+Pv7a49btWqF7du34+TJk+jQoQPef//9147g9Tq9evXCnj17sHr1apw5cwaBgYHw9fXFyZMn09wmJ25kT74X4uV5cgwMDGBqaprisrgff/wRdnZ2qFChAj777DM8fvxYWxccHAwrKyud+y/q1q0LS0vLNOexeRNz585FVFQUxo4dm6H2cXFxiImJQZEiRbI9FyIiIqK8hJf5UbYoUqQIWrdujaCgILRs2RIA8Msvv8DIyAht27YFkHjTfJUqVbRtxowZg82bN2PdunUZ/iL/qvDwcKxZswZXr16Fq6srAGDAgAHYsWMH5s+fj7lz56a63ccff4wPPvgg3X07OTllKpdy5crB1dUVo0ePxsKFC2FlZYWZM2fi5s2bOpOfduvWDW5ubihRogTOnj2Lzz//HKdOncK2bdsAAJGRkbC3t4dSStsmefjqyMjITOX0OqdPn8aXX36JgwcP6kwemp6xY8fCyspKe12JiIiICioWU5Rt/P39ERgYiKdPn8LCwgJBQUHo0KGD1lPz5MkTfPnll/jtt98QERGB58+fIyYmRmfOm8w6fvw4RCTFTeuxsbFo3LhxmtvZ2trC1tY2y8dNjbGxMTZs2IDevXujaNGiMDQ0RNOmTdGqVSuIiNbuo48+0v5dqVIllCpVCrVq1cLx48fh5eWVrTmlJzY2Fp07d8a0adNQsmTJDG3z3XffYf78+dixYwcKFSqUwxlSbpE8NxsRERHpYjFF2cbHxwdGRkb49ddf0aRJE+zYsUOb/wYAPvvsM2zduhXTpk1DmTJlYGFhgR49eqSYA+dlBgYGKb7EPX/+XPt3QkIClFI4cuRIinly0hvVa9KkSToDQKTmjz/+wHvvvZdum1dVr14dISEhePToEeLi4mBvb49atWqlO1xzjRo1YGhoiIsXL8LLywuOjo64e/cuRETrnRIR3LlzJ8U8Nm8iIiICYWFh6NWrF3r16gUg8fkUERgZGWHLli1o3ry51v7bb7/FuHHj8Mcff6BmzZrZlgcRERFRXsViirKNqakpOnXqhKCgINy7dw+Ojo46w3Hv378fPXr0QIcOHQAkDt0dHh6OsmXLprlPe3t7nUvkYmJicO7cOW20rWrVqkFEEBkZiUaNGmU415y4zO9lhQsXBgBcvHgRR48exf/+9780254+fRrx8fEoXrw4AMDb2xvR0dEIDg7W7psKDg7GkydP0p3HJrOcnJxw+vRpnWVz587F9u3bsXHjRri7u2vLZ8yYgfHjx+P3339HvXr1si0HIiIioryMxRRlK39/fzRp0gRXrlxB165dYWDw/2OclC1bFhs3boSfnx+MjY3x5ZdfIiYmJt39NW7cGEuWLEHbtm1hb2+Pr776Ci9evNDZZ/fu3dGzZ09Mnz4dXl5eePDgAXbv3o1SpUrh/fffT3W/OXGZHwCsXbsWdnZ2cHNzw+nTpzF48GC0a9dO6+EJDw9HUFAQWrduDTs7O4SGhmLYsGGoVq0a6tatCwDw9PREy5Yt0a9fPyxYsAAA0K9fP7Rp0wYeHh6Zyic0NBRxcXG4d+8eoqOjERISAgCoWrUqjI2NU8wpVaxYMZiamuosnzp1KsaMGYNVq1ahbNmy2n1b5ubmWtEYHR2tzcOTkJCA69evIyQkBLa2ttq9bJQzGjZsiIoVK2L27Nn6TiVdEyZMwJdffgkAmDx5MkaNGqXnjPIHd3d3XLt2DQBw9+5d2NnZ6TkjIqKChaP5UbZ677334OTkhNDQUG0Uv2QzZsxAsWLF8N5776FVq1aoXbv2ay+j+/zzz9G4cWP4+fmhefPmqFevXop5YpYuXYpevXphxIgRKFeuHNq0aYO9e/dqc768TREREejRowfKlSuHQYMGISAgAGvWrNHWm5iYYOfOnWjRogU8PDwwaNAgNG/eHDt27NAZAGL16tWoUqUKWrRogRYtWqBKlSpYuXKlzrGUUpgwYUK6+bRu3RrVqlXDTz/9hGPHjqFatWopnr/XmTNnDp4/f47OnTujePHiWgwePFhrc/ToUW3fz549w/jx41GtWjV88cUXmToW5W9Zneh6wYIFaNSoEWxsbKCU0uaESpaQkIC2bdvC1dUVZmZmKF68OPz9/VPMmTZ48GDUqFEDZmZmOj2vmZHRybDnzJkDT09PmJubw8PDAytWrEjRZv369ShfvjxMTU1Rvnx5bNy4UWd9dHQ0Bg4cCGdnZ20/M2fO1Glz5MgRrF+/PkvnQkRE2UDfY7Mz8mYU9Dl5MjIfU066fPmyKKV05qLKCzjPVPZ73bxluUVqfzMJCQlSu3ZtqVOnjhw6dEjOnTsnH330kbi6ukp0dLTWbubMmTJp0iSZOXOmAJArV67o7Cc+Pl5mzpwpwcHBcvXqVfn777/F29tb3n33XZ12AwYMkFmzZknfvn3Fzc0tS+cRHR0t/fr1k/nz54ufn1+q87fNnTtXLC0tZfXq1RIeHi5r1qwRKysr2bRpk9bmwIEDYmhoKBMnTpTQ0FCZOHGiGBoaysGDB7U2ffv2lZIlS8pff/0lV65ckeXLl4uJiYmsWLFC53i7du0SAHL37t0snVNW6ftvlsFgMHJD6D0BRt6MvPrFM7uMHz9eDAwMxNLSUqZPn/7Wjz979mwJDAx868d9Ey1bthQLCwsWU0k40XXGJ7pOduTIkVSLqdT8+uuvAkCePXuWYt3UqVOzXEy9LK3JsL29vWXIkCE6y4YOHSp169bVHn/wwQfStGlTnTZNmjSRLl26aI8rVKggX3zxhU6b+vXrpyieWUwxGAyG/oKX+RFlwaBBg3D+/HmEhITgww8/fOvH//TTT7Fs2bK3ftw3sWjRIpw8eRIXL15EpUqV9J2O3nGi68xNdJ0ZDx48QFBQEGrVqqWz77clNjY2xXHNzc1x+PBhbTTS4OBgndEyAaBFixY6E3PXq1cPmzdvxo0bNwAABw4cQEhIiDaXHxER6R8HoCDKgpwawCI/e5PREfMjTnSd8YmuM2rkyJGYPXs2nj59itq1a+O3337L9D6yQ4sWLbB48WK8//77qFGjBo4dO4ZFixbh+fPnuHfvHooXL47IyEg4ODjobOfg4KAzMfesWbPQr18/uLq6wsgo8eP6+++/R5s2bd7q+RARUdpYTBER6Qknus7YRNcZNXz4cPTu3RvXrl3Dl19+CX9/f/zxxx/afG1vy7hx4xAZGYk6depARODg4IDAwEB88803OiOcvs7333+PAwcOYNOmTXBzc8PevXvx2Wefwd3dnb1TRES5BIspIiI94UTXWZvoOi12dnaws7ND2bJl4enpCRcXF+zfvz/TOb0pc3NzLFmyBPPnz0dUVBSKFy+OBQsWwNraGvb29gAAR0dHREVF6WwXFRWlTcz97NkzfP7551i7di18fX0BAJUrV0ZISAimTZvGYoqIKJdgMUVEpCec6Pr/ZWai64xISEgA8P/3ZemDsbExnJ2dAQA//vgj2rRpo/VMeXt7Y/v27Rg+fLjWfvv27drE3M+fP8fz5891pkwAAENDQ+3ciIhI/1hMUZbExcXddHFx0XcalI/ExcXd1HcO+sCJrtOf6BoAIiMjERkZiQsXLgBInIz64cOHcHV1ha2tLYKDg3H8+HHUq1cPNjY2CA8Px7hx4+Du7o569epp+7l06RKio6Nx+/ZtxMXFaZNYly9fHiYmJhnOOb3JsAHgwoULOHToEGrXro1//vkHM2bMwJkzZ7B8+XJtH4MHD0b9+vUxZcoUtGvXDhs3bsSuXbu0gTcKFSqEBg0aYNSoUbCysoKbmxv27NmDFStW4JtvvsnKU01ERDlB38MJMhgMRm4IfQ33n5CQIG5ubgJATp48qbPu6tWr0qRJE7GwsBAnJyeZOnWq+Pj46AyL/+qw548ePZIuXbpIoUKFpESJEjJnzpwUbeLi4mT8+PFSsmRJMTY2FgcHB/H19ZWjR4/m2HmmNTfbd999J87OzmJsbCyurq4yduxYiY2NTbEtgBSxdOlSERE5ceKENGzYUGxtbcXU1FTc3d3l448/lhs3bujsp0GDBqnu5+Wh1gHI+PHj0z2X5Nfr1UgWGhoqVatWFXNzcylUqJD4+fnJuXPnUuxn7dq14uHhIcbGxlKuXDlZv369zvqIiAjp2bOnlChRQszMzMTDw0OmTp0qCQkJOu04NDqDwWDoL5RI5m/yJSLKb1xcXG7fuHGjuL7zyK8mTJiAdevW4cyZM/pOJU1XrlzBO++8g3379qFu3br6TifDdu/ejUaNGuHu3buws7N7a8d1cXGJuHHjRom3dkAiolyI80wREdFbERYWBisrK8yYMUPfqaRqy5Yt6NGjR54qpCpUqIBWrVrpOw0iogKLPVNERGDPVE578OABHjx4ACBx1D0bGxv9JpRPXLt2TRutsVSpUpkaev1NsWeKiIgDUBAR0VvAia5zhpubm75TICIq0HiZHxERERERURawmCIiIiIiIsoCXuZHRIQ8PXeaSkhIKATA0MDA4BGAeH0nRK8nIlYiYpb0mj3Xdz5ZUVDnhiMiehkHoCAiyqOUUi0BTAbwHYDlwjf0PEUp5QFgHoDDACaIyDM9p0RERJnEy/yIiPIYpVRRpdRyAH0BtBKRZSyk8h4ROQ+gCYBrAPYrpRroOSUiIsok9kwREeURSikFoCOAsQD+KyLr9ZwSZROllBuAH5BYWI0UkX/1nBIREWUAe6aIiPIApVRxAOsB+ABoxEIqfxGRa0h8bQ8A2KuUaq3nlIiIKAPYM0VElIsl9Ub1AjAQwCgR+VPPKVEOU0o5APgeQByAISJyT88pERFRGtgzRUSUSymlSgLYCqAqgPdYSBUMIhIlIh8gsSdyh1Kqc1JRTUREuQx7poiIchmllCGAAQACAAwWkb/1nBLpiVKqCICpAOwAfCIit/WcEhERvYQ9U0REuYhSqjyAXQCKAajHQqpgE5F/RKQPgNkAfldK9WEvFRFR7sGeKSKiXEApZQJgJIDWAPqLSIh+M6LcRillCWAigIoAPhaRcD2nRERU4LFniohIz5RSNQDsA/AcQH0WUpQaEXkiIv8B8AWAH5VSQ5MuCSUiIj1hzxQRkZ4opcwBfAngXST2NJzXc0qURyilTAGMAdAIiT2ZZ/ScEhFRgcRiiohID5RSDQDMALAYwDwRSdBzSpQHKaUqI3Gy320AJotInJ5TIiIqUFhMERG9RUqpQgC+BuCGxB6Fa3pOifI4pZQRgP8A6AhggIgc0XNKREQFBospIqK3RCnVGsAkANMBrBK+AVM2UkqVATAPwAkAX4jIUz2nRESU73EACiKiHKaUslNKrQLQE0ALEVnJQoqym4hcBNAMwCUA+5RSjfScEhFRvseeKSKiHJI0H1BnAJ8DmCAiG/WcEhUQSikXAHMB3AYwQkQe6TklIqJ8iT1TREQ5QCnlBOAXAE0BNGQhRW+TiNwA0BbAHgB7lFK+ek6JiChfYs8UEVE2SuqN6gPgEwDDRWSHnlOiAk4pZQ9gFgABMFhE7uo5JSKifIM9U0RE2UQp9Q4Sh6guD6AeCynKDUTkroh0BfAjgO1KqW5JRT8REb0h9kwREb0hpZQhgMEAugIYJCLBek6JKFVKqcIApgJwBPCJiNzUc0pERHkae6aIiN6AUqoiEu9LsUFibxQLKcq1ROSRiHwE4FsAm5RS/ZRS/C5ARJRF7JkiIsoCpZQJEkfpa47EyXdP6TklokxRSlkA+C+AagA+ThpanYiIMoG/RhERZZJSqiaA/QCeAmjAQoryIhF5KiKfIfFHgSCl1GdKKSN950VElJewZ4qIKIP4Sz7lV+xpJSLKGhZTREQZoJRqBGA6gAUAFohIgp5TIsp2SfcA/gBgF4CvRCRWzykREeVqvMyPiCgdSqnCSqkFAIYB8BOReSykKL8SkTMAGgJ4CGC/Uqq2XhMiIsrl2DNFRJQGpZQvgP8B+AbAGuEbJhUgSfOmzQNwBsBYEXmi55SIiHId9kwREb1CKWWvlFoDoBuAZiKymoUUFTQiEo7Ee6jOAtinlGqi55SIiHId9kwRESVRSikkTrw7AsAXIrJJzykR5QpKKScAcwHcBfCZiDzUb0ZERLkDe6aIiAAopVwAbELi/SINWEgR/T8RuQWgHYDtAHYrpfz0mxERUe7AnikiKtCUUgYAPkqKz0TkLz2nRJSrKaXsAHwLwBjAIBGJ0m9GRET6w54pIiqwlFJlkPhLe2kA9VhIEb2eiNwTEX8AKwD8qZQKSLpEloiowGHPFBEVOEopIwD/AdAJwAAROaznlIjyJKVUIQBTALghcbLf63pOiYjorWLPFBEVKEqpygD2ALBAYm8UCymiLBKRf0XkEyROH7BRKfVJ0qWzREQFAnumiKhAUEqZAhgDoDGAj5MmJyWibKKUMgcwAcC7SPwbu6DfjIiIch5/PSKifE8p5Q1gP4CHSBypj4UUUTYTkWciMhKJUwusVEqNTLqklogo32LPFBHlW0opSwBfAagIoF/SJKRElMOUUiZILKpaA/hEREL0mxERUc5gMUVE+ZJSqimAqUicaHSR8M2O6K1TSpUHMA/APgD/E5EYPadERJSteJkfEeUrSikbpdRiAAMBtBGRhSykiPRDREIBNAJwB8B+pVQdPadERJSt2DNFRPmGUqodEm+AnwzgZxZRRLmHUqokEnupzgMYLSLRek6JiOiNsWeKiPI8pZSDUupnAB0BNBWRn1hIEeUuInIFQEsAIQD2KaWa6zcjIqI3x54pIsqzlFIKgD+AYQDGiMjvek6JiDJAKVUCwGwAjwAME5EHek6JiChL2DNFRHmSUsoVwO8A6gCoz0KKKO8QkdsAOiDxb/gvpVQHPadERJQl7JkiojxFKWUAoD+A3gD+IyJ79JwSEb0BpVRRADMAWAIYICKRek6JiCjD2DNFRHmGUsoDwE4ArgDqspAiyvtE5L6IBAJYBOAPpVTPpEt4iYhyPfZMEVGup5QyAvAZgPYAPhWRo3pOiYhygFLKGomjcb4DoL+IXNVvRkRE6WPPFBHlakqpqgD2AjAC8B4LKaL8S0Qei8gAAJMArFNKDUy6tJeIKFdizxQR5UpKKTMA4wDUB9AvafJPIiogkt4DxiNxkJmPRSRMzykREaXAX3uIKNdRStUFsB/AHQANWUgRFTwiEiMinwP4D4AlSqnRSiljfedFRPQy9kwRUa6hlLJC4uU9Hkj8JfqKnlMiolwgqYj6DEBbJN43eVzPKRERAWAxRUS5hFKqOYCvAXwPYKnwzYmIXqGUKgdgHoBgAP8VkWd6TomICjhe5kdEeqWUslVKLQXwMQAfEVnCQoqIUiMi5wA0BnALwD6lVD09p0REBRx7pohIb5RSHQB8AWAigHUsoogoo5RS7gB+ABAO4HMReazfjIioIGLPFBG9dUqp4kqp9Ui8/6GxiKxlIUVEmZE0B1VrAEcA7FVKtdJvRkRUELFniojeGqWUAhAIYAiAUSKyVb8ZEVF+oJRyROL9ls8A/EdE7us5JSIqINgzRURvRdIlOX8AqIHEyXdZSBFRthCRSBHpBOAXADuVUp2SfrwhIspR7JkiomyT9OWllogcfGmZAYBPAfQEMERE9ukpPSIqAJRSRQBMB2CDxGHUI15a5wbguYjc1lN6RJTPsGeKiLJTewC9kh8opTwB7AJQAkBdFlJElNNE5B8R+RCJQ6hvUUp9+FIvVTEAi9lrRUTZhT1TRJQtlFLmAA4AaAHgHwAjAPgC+IQTbBKRPiRNBP4VgHIA+ovIZaXUMgAbRGSTXpMjonyBxRQRZQul1HgkFlH7AcwF8CuAaSLyXK+JEVGBp5SqA+A7AEEAfgbwOwBvEYnRa2JElOexmCKiN5Z0H8J6AH8BqAWgX9LkmkREuYJSygzAWAD1ARwE8EhEvtJvVkSU17GYIqI3ppTaDaAkEnujjgNwBuAC4B8RGaXH1IiogFNKNUHivZw3ksIIQB8AxZE4YM5lPaZHRHkciykieiNJI2eFATgK4Dz+/wvLTQCXON8LEemTUsoEQBkk/sDjgsQfe1wB1AWwSUQ+02N6RJTHsZgiIiIiIiLKAiN9J0C5g4ODw2ETExNnfedBlFFxcXE3o6Kiauo7D6K8hu/3RPkfPyPfHhZTBAAwMTFxvnHjRnF950GUUS4uLvpOgShP4vs9Uf7Hz8i3h5P2EhERERERZQGLKSIiIiIioixgMUVERERERJQFLKYoW927dw9KKezevVvfqWDZsmWwsrJ6q8dUSmHdunVv9ZgTJkxAxYoV3+oxiahg0cd727p166CUeqvHBAB3d3dMmzbtrR83N8mO5yA3fR8gykkspihfSO2DvnPnzrh8mXMx6sM///yDgIAAFC5cGIULF0ZAQAAePnyY7jY9e/aEUkonateu/XYSJqJ0RUREwNfXFwBw9epVKKVw9OhRPWdFGZGVH9yOHDmCTz75JIcyytuy8vn26mdbcnz66ac67S5cuID3338fNjY2sLCwgJeXF8LCwrT1kZGRCAgIgKOjIywsLFClShUEBQXlxGlSJnA0P8q3zM3NYW5uru808oTr16/D1dU12/bXrVs3XL9+HVu3bgUA9OnTBwEBAdi8eXO62zVt2hQrV67UHpuYmGRbTkSUdY6OjvpOgd4ie3t7faeQbXLD51tERITO46NHj8LX1xcffPCBtuzKlSuoW7cuevTogb/++gs2NjY4d+6czhU2PXr0wIMHD/Drr7/C3t4eGzduREBAAFxcXFC/fv1sO0fKJBFhMODs7HxbXpGQkCBff/21lCpVSszMzKRixYqycuVKnTaHDx8WLy8vMTU1lapVq8pvv/0mAGTXrl0iIrJr1y4BIHfv3tW2uXLligCQI0eOaMvCwsLE19dXChUqJJaWllK7dm05deqUdoxmzZpJ0aJFxdraWurWrSsHDhzQtnVzcxMAWri5uYmIyNKlS8XS0lIn33nz5sk777wjxsbG8s4778iCBQt01gOQ+fPnS8eOHcXCwkJKliyZ4pzTA0DWrl2rPb5586Z07txZbGxsxMbGRlq3bi0XLlwQEZHz588LAO08k82fP1+KFi0qcXFxIiJy9uxZad26tVhZWYm9vb106dJFIiIitPbjx4+XChUqZDjHZPfu3ZPZs2dLzZo1xd3dPdPbpyU0NFQAyP79+7Vl+/btEwBy7ty5NLcLDAwUHx+fDB8n6f+s3v92GIy8Fmm930+bNk1Kly4tJiYm4uTkJKNGjdLWv/ze9vL7LQBp0KCB7NmzR4yMjHTem0RERo8eLZUqVXr1cKlavny5uLq6irm5ufj4+Mjs2bMl8WvK/9u0aZP2mePu7i6jR4+W2NhYERH5/PPPxcvLK8V+vb29ZeDAgdrjJUuWiKenp5iamkqZMmVkxowZEh8fr613c3OTqVOnao+vXbsm7dq1EysrK7GyspL27dvLjRs3tPXJ78ELFy4UFxcXMTMzEz8/P53PveT3tylTpoiDg4MUKlRIRo4cKfHx8TJ+/Hixt7cXBwcHmTJlik7uDx8+lL59+4q9vb1YWVlJ/fr1dT47kz/nduzYIRUqVBALCwtp2LChXL58WVv/6uu1dOnS174Wrz4HGflsfN33AZH0P8/OnTsnFhYWsmzZMq39H3/8IcbGxjqf+RmR2z7fXtWnTx8pW7aszrKuXbtKt27d0t3O0tJSlixZorPM1dVV57VKxs/Itxd6T4CROyK1D9fRo0dL2bJl5Y8//pDLly9LUFCQWFhYyG+//SYiIo8fPxZ7e3vp2LGjnD59WrZu3SrlypXLdDF169YtKVq0qLRt21YOHTok58+fl5UrV8qJEydERGTnzp2yYsUKCQ0NlbCwMPn000/FxsZG7t27JyIid+7cEQCycOFCiYiIkDt37ohIymJqw4YNYmRkJN9//72cP39eZs2aJUZGRrJp0yatDQBxcnKSlStXysWLF2XUqFFibGws165de/XpSdXLXziePHkiZcqUkcDAQDl58qSEhYVJ7969xdXVVZ48eSIiIu+++66MHDlSZx/169eX/v37i4jI7du3pWjRojJixAgJDQ2VkydPSps2baRmzZrah39miqm4uDj55Zdf5P333xcTExMpX768TJ48Wa5fv661uXbtmlhaWqYb/fr1S/MYixcvFisrK0lISNCWJSQkpPoh8LLAwEApXLiw2NvbS5kyZaRPnz4SFRWVZnt+UDAYWYvU3u9HjRolhQsXlsWLF8vFixflwIEDMmfOHG39y+9thw8fFgCydetWiYiIkPv374uIiIeHh3z99dfaNvHx8eLs7Czffvvtq4dL4eDBg6KUkokTJ8r58+dl3rx5Ymtrq1NMbd26VaytrWXJkiVy6dIl+euvv6Rs2bIybNgwEUn8og5AwsLCtG3Cw8MFgBw6dEhERBYsWCCOjo6ydu1auXz5smzatEkcHBzk+++/17Z5uZCIj4+XqlWrire3txw5ckSOHDkitWrVkurVq2vvcePHjxdLS0tp0KCBHD9+XPbv3y/ly5cXX19fbZ+BgYFibW0t/fr1k7CwMFm9erUopaRFixYyatQoOX/+vPzwww8CQI4ePSoiie+bdevWldatW8uhQ4fk4sWLMnbsWLG2tpbbtxNfwqVLl4qRkZE0adJEDh06JCdPnpSqVatK8+bNRUTk6dOnMmzYMPHw8JCIiAiJiIiQp0+fvvb1SK2YSu+zMSPfBzLyeTZ//nyxtraW8PBwuXPnjjg4OMiECRNem69I7v58e9njx4/FyspK5/mNj48XKysrmThxorRo0ULs7OykRo0a8uOPP+ps26JFC2nZsqXcu3dP4uPj5ZdffhFzc3Pt+9LL+Bn59kLvCTByR7z64RodHS1mZmayd+/elxfL4MGDpVWrViKS+KZXuHBhefz4sbZ+5cqVmS6mRo8eLa6urtqvi6+TkJAgjo6OOr+KvdojJJKymKpTp4706tVLp01gYKDUrVtXZz8v/xr7/PlzMTc3z3Dv1Mt5LF68WEqXLq3zpvvixQuxtbWVn376SUREvvvuO3F1ddXaXLt2TZRS8vfff4uIyLhx46Rx48Y6x3jw4IHOl4OMFFPHjx+XQYMGiZ2dnTg4OMiQIUPk2LFjqbZ9/vy5XLx4Md1Ir8j56quvpGTJkimWlyxZUiZNmpTmdmvWrJFff/1VTp06JZs2bZLKlStLhQoVJCYmJtX2/KBgMLIWr77fP378WExNTeWHH36QtLz83pba1QUiIlOnTpVy5cppj7ds2SImJibaD1/p6dq1qzRt2lRnWe/evXWKqffee0/++9//6rTZuHGjWFpaau+h1apVk7Fjx2rr//e//+n0ALi4uMiKFSt09jFz5kzx9PTUHr9cSGzbtk0MDAzkypUr2vrw8HBRSsn27dtFJPE92MDAQOdHt+TeiuQrEQIDA8XZ2VlevHihtalevbpUrlxZJ5eXj71z506xtLRMUfxUqVJFK1qTe55e7hVZtWqVmJiY6BR7mb16IbViKr3Pxox8H8jI55mISLt27aRWrVrSunVrqVOnjs5zlpq88Pn2svnz54uJiYn2w6+ISEREhAAQCwsLmT59upw4cUKmT58uhoaG2g/YIiKPHj2SVq1aCQAxMjISS0tL+eWXX1I9Dj8j317wnilKVWhoKGJiYtCyZUud0ZSeP38Od3d3AEBYWBgqV66scz2vt7d3po914sQJ1KtXL837Y+7cuYNx48Zh165diIqKQnx8PJ49e4br169n6jhhYWH48MMPdZbVq1cPmzZt0llWuXJl7d9GRkawt7fHnTt3MnUsADh27BiuXLkCa2trneVPnz5FeHg4AKBLly4YNmwY9u3bh/r162PNmjUoWbIk6tSpo+1j7969qY5KGB4ejpo1a2Yol/bt2+PGjRsYPXo0xo8fDyOjtP/0jYyMULp06YyeZrbp0qWL9u9KlSqhevXqcHNzw++//47333//redDVFCEhoYiNjYWTZo0eaP9BAYGYsyYMThw4ADq1KmDJUuWoF27dihatOhrtw0LC9MGuEjm7e2NxYsXa4+PHTuGw4cP4+uvv9aWJSQk4NmzZ4iMjETx4sXh7++POXPm4H//+x8AICgoCN27dwcA3L17Fzdu3EC/fv3Qv39/bR8vXryAiKSZV4kSJbTPPQAoVaoUSpQogdDQUDRt2hQA4OTkpHNfTq1atWBgYICwsDCUKVMGAFC+fHkYGhpqbRwcHGBjY6NzPAcHB+3z5tixY3j69GmK+5diYmK0zxAAMDU1hYeHh/a4RIkSiIuLwz///ANbW9tUzysr0vtszMj3gYx+ni1atAgeHh44e/YsTp06pfOcpSYvfL69bOHChfDz89N5XRMSEgAAfn5+GDp0KACgatWqOHr0KGbPng0fHx8AwNixY3Hv3j3s2LEDdnZ2+OWXX9CjRw/s3bsXVapUefsnQwA4AAWlIfkPe/PmzSlu3DQ2Ns7wfgwMEgeMfPmD6vnz55nKJTAwEFFRUZg5cybc3d1hamqKJk2aIC4uLlP7ScurQ+++en5KKe35yIyEhARUrVoVP/74Y4p1yR9wxYoVQ7NmzRAUFIT69evrfPAn78PHxyfVIWodHBwynMuqVauwZMkSfPfdd1izZg38/f3h7++f6ofK9evXUb58+XT35+/vj3nz5qW6ztHREXfv3oWIaM+tiODOnTuZuom9RIkScHZ2xsWLFzO8DRHpj729Pdq2bYslS5bAw8MDmzZteu2gM5mRkJCA8ePHo1OnTqkeGwC6du2KESNGIDg4GKampjh37hz8/f217QFg3rx52g9WbyKzw7an9tmS3udNQkICHBwcsG/fvhT7KlSokPbvV4uH5Lyy8rmVnjf9bMzo59mZM2fw6NEjAMCtW7dQsmTJdPeblz7fQkJCcPToUUyaNElnuZ2dHYyMjFLk5unpqX2HCA8Px/fff4+QkBCtcKpSpQr27duH77//HosWLXrt8SlnsJiiVJUvXx6mpqa4du0aGjdunGobT09PLFu2DE+ePIGlpSUA4ODBgzptkj/gIiIitH+HhITotKlWrRpWrVqFuLi4VHun9u/fj1mzZmm/zERFRaUYGcfY2Bjx8fHpnpOnpyf+/vtv9O7dW2ffr3tjzSovLy+sWbMGdnZ2KX59fJm/vz8GDBiAjz76CKdPn9YZ4t3Lyws///wz3NzcMlXEvqpevXqoV68eZs+ejQ0bNmDFihWYOHEi3n33XQQEBKBz587ar8clSpRI8Rq96uUP8ld5e3sjOjoawcHB2heW4OBgPHnyJFNfYO7du4dbt26hePHiGd6GiDLP09MTpqam2Llzp9aLkp7k9+nU3nP79u2Ljh07olSpUnB0dNR6bjKSw6ufH68+9vLywrlz59LtWShevDgaN26MoKAgmJqawtvbG6VKlQKQ+IW9RIkSCA8PR48ePTKc1+3bt3H16lWtd+ry5cu4ffu2zmfHrVu3cOPGDbi4uAAADh8+jISEBHh6emboOKnx8vJCVFQUDAwMtHPIChMTk9d+Pr6pjHwfyMjn2cOHDxEQEIDPPvsMz549Q0BAAE6ePJnuZ05e+nxbsGABSpYsmeLvwsTEBO+++y7Onz+vs/zChQtwc3MDkHhVC4AUPXWGhobZXjhTJun7OkNG7ojUbkgeM2aM2NraajcknzhxQn744QeZP3++iCReZ29nZycffPCBnDlzRrZt2yaenp4610jHxcWJi4uLtG/fXs6fPy9//vmnVK5cWed6+5s3b4qtra34+fnJ4cOH5eLFi7J69WrthkovLy9p3LixnD17Vg4fPiwNGzYUS0tLGT9+vJZrmTJlpG/fvhIRESEPHjwQkZT3TG3cuFGMjIxk9uzZcuHChTQHoHj13qtXrx1Pz8vbP3nyRMqWLSv169eX3bt3y+XLl2XPnj0ydOhQ7Tr65HZWVlZSpUoVeffdd3X2d+vWLbG3t5f27dvLwYMHJTw8XLZv3y59+/aVf//9V0SyPprfzZs3ZfLkyeLp6SmlSpXK9PbpadmypVSsWFEOHDggBw4ckIoVK0qbNm10ju3h4SEbNmwQkcT/S8OGDZMDBw7IlStXZNeuXVK7dm1xcnLSzvNVvB6cwchapPZ+P2LECLGxsdEGdzh06JDMnTtXW//ye1vy/TJffvmlREZGysOHD7V2CQkJ4u7uLiYmJvLFF1+8epg0BQcHi1JKJk2aJBcuXJAFCxZI0aJFUwxAYWRkJOPGjZPTp09LWFiYrF27VoYPH66zr+XLl4udnZ2UKFFCZxANEZGFCxeKmZmZzJgxQ86dOyenT5+W5cuX69zv8vJ7fkJCglStWlXq1KmjDUBRu3btVAegaNSokZw4cUJ7z3t5dNLURiv18fGRwMBAnWW1atXSBtRISEiQevXqScWKFWXLli1y+fJlOXDggHzxxRfa/cypjVr76r3KQUFBYm5uLseOHZO7d++meR/qy1K7Zyq9z8aMfB/IyOdZ586dxcvLS+Li4iQmJkYqVaok/v7+r833Vbnp8y3ZkydPpFChQjJx4sRU97tx40YxNjaW+fPny8WLF2XBggViZGSk3TMVFxcnpUuXlvfee08OHTokly5dkmnTpolSSud7TDJ+Rr690HsCjNwRaQ2VO2vWLPH09BQTExOxs7OTpk2byrZt27Q2Bw8elGrVqomJiYlU/r/27jssiuvrA/j30naXooggICKCgoqxxkY0FlRQEbHEggholMSuQVGTWBM1RiHVEstPo6ixx6gxGmKLGgsWSBSV2BtglFiACAjn/WNhXlcWhAUcFs7neeZJmLkzc2blcPfO3Lm3USPauXNnnqFQjx07Ro0bNyalUkmtW7eWhkt98eXl8+fPU7du3cjMzIzMzc3Jw8OD/vrrLyIiiomJoZYtW5JSqSQXFxdau3YtNWjQQKMxtXPnTqpTpw4ZGRkVODT60qVLqXbt2mRkZJTv0Ogl1ZgiIkpMTKQhQ4aQjY0NmZiYUK1atWjo0KEaA3IQEQUGBhIA+vrrr/McMz4+nvr27UuWlpakVCrJzc2NxowZIw3YoWtj6kVxcXHF2v9lycnJFBAQQBYWFmRhYUEBAQH077//SttzX2DPHaI3LS2NvLy8yMbGhoyNjalmzZoUHBysMQrTy7ii4IUX3RZtf++zsrLos88+I2dnZzI2NqYaNWrQRx99JG1/+W9b7jDgBgYG1L59e41jzZ49m4QQGoM2FMaqVaukocW7du1K3377bZ6h0fft20dt27YllUpFFhYW9Oabb2qMxEek/mJvampKxsbGWge/2LBhAzVt2pQUCgVZWlpSmzZt6IcffpC2axsa3c/PTxoavVevXlqHRl+2bBnVqFGDlEol9ezZU2OAAV0aU0RET548oXHjxpGDg4P07zJgwAC6cuUKERWuMfXs2TOpDnnx725BitqYIirc94GC6rO1a9eSSqXSGI3xr7/+IqVSqfHvU1Ry12+5Vq1aRYaGhnT37t18j7169WpydXUlpVJJDRs2pA0bNmhsj4+Ppz59+lC1atXI1NSUGjVqpDGU/Iu4jnx9iyDS/tIlq1gcHR3v3b59m/tTMb3h6OiYcPv27epyx8GYvintv/cjR47ElStXEBUVVVqnKFNmzZqFrVu34vz583KHwpiE68jXh9+ZYowxxlixPX78GHFxcVi7di02b94sdziMMfZaGMgdAGP6Yt68eTA3N9e6dOvWTe7wGGNMVn5+fujUqRPeffddacCgXN26dcv37+fLI5ux0rV+/fp8/y0aNGggd3iM6R3u5scAcDe/wkhOTkZycrLWbSqVCg4ODq85ooqNuzAwphs5/t7fvXsX//33n9ZtVlZWJTofEivY06dPkZSUpHWbsbGxNHoc029cR74+3M2PsULiCp8xxnTDN5vKDgsLizyTyTPGdMfd/Fi51aFDB4wZM6ZI+9SqVUvrhIKMMcaYLrguYqx848YUK7e2b9+Ozz77rEj7REdHY9SoUaUUkdqtW7fg6+sLMzMzWFtbY9y4ccjIyChwn5CQENSuXRsqlQo2Njbw8/PDxYsXNcqcPXsWXbp0gaWlJapWrYr33nsPKSkp0vbY2Fj4+/vD0dERKpUKdevWxYIFC3iyP8YYK0XlpS5KTk7G2LFjUa9ePahUKjg6OmLkyJF4+PChVCY7Oxs9e/ZEzZo1oVQqYW9vj8GDB+Pu3btSme+//x5CCK1LdHR0qV4zY6WBG1Os3LKysipyVwYbGxuYmpqWUkRAVlYWfHx88PTpUxw5cgQ//PADtm7diokTJxa4X/PmzfH999/j4sWL2LdvH4gInTt3RmZmJgDg3r176Ny5M1xcXHDy5Ens3bsXFy5cwJAhQ6RjnDlzBjY2NoiMjMSFCxcwe/ZsfPrpp5g/f36pXS9jjFV05aUuunfvHu7evYsFCxbgr7/+wrp16/D777/D399fo5ynpyc2b96My5cvY9u2bbh27Rp69+4tbR8wYAASEhI0lsGDB8PFxQXNmzcvtWtmrNTIPdEVL2Vj0TaJY1mWkpJCgYGBZGZmRtWqVaN58+blmQCxffv2NHr0aOlnJycn+vTTT+m9994jCwsLcnBwoAULFmgctygT9Opiz549JITQmIw2MjKSFAoFPX78uNDHiY2NJQB06dIlIiJatmwZVa1alZ4/fy6V+fPPPwkA/f333/keJywsjJo1a6bDlciPJyTkhRfdFn37e1+WVfS66OeffyYhRIH7/PTTTwSA/vvvP63bU1NTqXLlyjR37tzCXwB7Ja4jX9/CT6aYXpo4cSIOHz6MH3/8EQcOHEBsbCyOHDnyyv2+/PJLNGzYEGfPnsWUKVMwefJkHD9+vNDnPXLkSL5DyhZmmN/jx4+jfv36cHR0lNZ5e3sjPT0dZ86cKVQMqampWL16NWrWrIlatWoBANLT02FsbAxDQ0OpnEqlAgAcPXo032M9efIEVapUKdR5GWOMaarIdRGgrkMUCkW+T9GSk5Oxfv16tGrVCkqlUmuZzZs3IzU1Fe+++26hz8tYWcKj+TG9k5KSglWrVmHt2rXo0qULAOB///sfatSo8cp9vby8pBeBx44di2+++Qb79++Hh4dHoc7dvHlzxMTEFFimoBH/EhMTYWtrq7HO2toahoaGSExMLPC4S5YsweTJk5Gamoq6deti//79UCgUANTdKkJDQzF//nyEhoYiNTUVU6dOBQAkJCRoPd7Zs2fx/fffY/369QWelzHGWF4VtS7K9ejRI0yfPh0hISEwMtL8OjllyhQsWrQIaWlpaN26NXbv3p3vcZYvX44ePXrAzs6uUOdlrKzhxhTTO1evXkVmZiZatmwprTMzM8Mbb7zxyn0bNWqk8XP16tVx//79Qp9bpVKhTp06hQ+2BAUEBKBLly5ISEhAeHg4+vXrh2PHjsHU1BQNGjTAmjVrEBoaio8//hhGRkYYN24cbG1tYWCQ9wH05cuX4ePjgwkTJqBv374yXA1jjOm3iloXAeqGpK+vLxwcHLBgwYI828PCwjBs2DDcvHkTs2fPxuDBg/HLL79ACKFR7sKFCzh+/Dh+/vnn1xU6YyWOu/mxCsXY2FjjZyFEkUazK27XCjs7uzyTJT548ABZWVmvvCtXuXJluLq6ol27dti6dSvi4+Oxbds2afugQYOQmJiIe/fu4eHDh5g1axb++ecfuLi4aBzn0qVL6NChAwYOHMiDTzDGmAz0uS5KSUlB9+7dAQC7d+/W2n3P2toabm5u6NKlCzZu3Ih9+/Zp7XK+fPlyODo6omvXroW5bMbKJH4yxfRO7dq1YWxsjOjoaKmhkJaWhvPnz6N27dqleu7idq3w8PDAnDlzcOfOHakrSFRUFBQKBd58881Cx5H70mN6enqebbldN1atWgWlUil1PwGAuLg4eHp6on///vjyyy8LfT7GGGOaKmJd9PTpU3Tr1g1EhL1798Lc3PyVseY2El+ur549e4bIyEiMGzdOaw8KxvQFN6aY3jE3N8e7776LKVOmwNraGvb29pgzZw6ys7PzdCEoacXtWuHl5YUGDRogKCgIERERePjwIcLCwhASEoJKlSoBAE6dOoWgoCCsXbsWLVu2xJUrV7Bt2zZ07twZNjY2uHPnDubPnw+FQoEePXpIx160aBE8PDxgYWGBqKgohIWFYf78+bC0tASg7k7h6emJjh074qOPPtLoF8991RljrGgqWl309OlTeHl54cmTJ9ixYwdSU1ORmpoKQN1wMzExwfHjx3H27Fm0bdsWlpaWuHr1KqZPn45atWqhbdu2GjFs3boVjx8/5oEnmN7jxhTTS+Hh4UhNTUXPnj1hbm6ODz74AElJSfmOFlRWGBoa4ueff8aoUaPQpk0bqFQqBAQEYOHChVKZtLQ0XL58GWlpaQAAhUKBQ4cOISIiAo8ePYKtrS3atWuH48ePazSCTp06hZkzZyIlJQX16tXDsmXLEBgYKG3fsmUL7t+/j02bNmHTpk0acRFRKV85Y4yVPxWpLjpz5gxOnDgBAHBzc9M43sGDB9GhQweoVCps3boVM2bMQGpqKuzt7dG1a1ds2rQpz2eyYsUKeHt7o2bNmqV8tYyVLsFfohgAODo63rt9+7a93HHoKj09HU5OTggLC3vlBLisfHB0dEy4fft2dbnjYEzf6Pvf+7KM6yJWVnAd+frwkymml86dO4eLFy9KXQ8+//xzPH36FAMGDJA7NMYYYxUE10WMMW5MMb31xRdf4PLlyzAyMkKTJk3w+++/F2p+D8YYY6ykcF3EWMXGjSmml5o2bYrTp0/LHQZjjLEKjOsixhiPRckYY4wxxhhjOuDGFGNa3LhxA0IIvuPIGGNMNlwXMVb2cWOKMT21fft2eHl5wcbGBhYWFmjVqhV27typUSYzMxOffPIJateuDaVSicaNG2Pv3r0aZWrVqgUhRJ7Fx8dHKpOVlYXp06fD2dkZSqUSzs7OmDZtGp4/f/5arpUxxljZtWHDBjRp0gSmpqaws7PD4MGDNeYyBIBt27bB3d0dCoUC7u7u+PHHH/McJz4+Hn369IGlpSVMTU3RrFkzXLx4Udq+fPlydOzYEZaWlhBC4MaNG6V9aYy9EjemGNNThw8fhqenJ37++WecO3cO3bt3R+/evXHkyBGpzLRp07B06VJ88803iIuLw4gRI9C7d2+cO3dOKhMdHY2EhARpOXv2LIQQ6N+/v1Tm888/x+LFi/HNN9/g0qVL+Prrr7F48WJ89tlnr/WaGWOMlS3Hjh1DYGAggoODceHCBezYsQNxcXEICAiQyhw/fhwDBgxAQEAAYmJiEBAQgH79+uHkyZNSmevXr6NNmzZwdnbGgQMHcP78ecyZMwfm5uZSmbS0NHh5eWHWrFmv8xIZKxgR8cILatSocY9kcPjwYWrVqhWZmZlRpUqVqEWLFvTXX38REdGDBw9o4MCB5ODgQEqlktzd3WnVqlUa+7dv355GjBhBoaGhVKVKFbK2tqavvvqKnj17RqNGjaLKlSuTo6MjrV27Vtrn+vXrBIDWr19Pbdq0IYVCQXXr1qV9+/blKRMdHS2tu3DhAnXv3p3Mzc3JxsaGBg4cSAkJCdL2P//8kzw9PcnCwoLMzMyoUaNGdODAgdL66LRq0aIFhYaGSj/b29vTV199pVGmT58+FBAQkO8x5syZQ5UrV6a0tDRpnY+PDwUFBWmUCwoKIh8fnxKKvOhyfmdlzx1eeNG3Ra6/92UZ10W6W7hwIdWsWVNj3apVq8jMzEz6uX///tS5c2eNMp06daKBAwdKP/v7+9OgQYMKdc7o6GgCQNevX9c98HKO68jXt/CTKSab58+fw8/PD23btkVsbCxOnjyJCRMmwNDQEADw7NkzNGvWDLt378aFCxcwfvx4vP/++9i/f7/GcdavXw8LCwucPHkSU6dOxYQJE9CrVy+4ubnh9OnTCA4OxvDhw5GQkKCx3+TJkzFu3DjExMSgS5cu8PPzw927d7XGmpCQgHbt2uGNN97AqVOn8NtvvyElJQV+fn7Izs4GAAwaNAj29vY4deoUYmJiMGvWrDwzvr9o3rx5MDc3L3B58SlTYTx9+hRVqlSRfk5PT88Tg0qlwtGjR7XuT0T43//+h8GDB0OlUknr27Zti4MHD+LSpUsAgLi4OBw4cADdu3cvUnyMMVbWcF1UvLqoTZs2SEhIwK5du0BEePDgATZu3KhRPxw/fhxeXl4a+3l7e+OPP/4AAGRnZ2PXrl1wd3dH165dYWNjgxYtWmDTpk35npexMkPu1hwvZWOR407lw4cPCQAdOnSo0PsMGDCAhg0bJv3cvn17at26tfRzdnY2WVtbk6+vr7QuIyODjI2NacuWLUT0/3f65syZI5XJysoiV1dX+vjjjzXK5N4NnD59Onl6emrEkpycTADo5MmTRERkYWFB33//faGv5eHDh/T3338XuLz4dOhVFi1aRObm5nTjxg1pnb+/P9WrV48uXbpEWVlZ9Ouvv5JKpSITExOtx9i3bx8BoJiYGI312dnZ9NFHH5EQgoyMjAiA9FnJhe+68cKLbgs/mdLEdVHx66Jt27aRhYWFVD906dJFYx9jY2Nas2aNxj5r1qyR6qKEhAQCQKamphQREUHnzp2jiIgIMjQ0pN27d+c5Hz+ZejWuI1/fwvNMMdlYWVlhyJAh8Pb2RqdOndCpUye88847qFmzJgD1oAfz58/Hpk2bcPfuXaSnpyMjIwMdOnTQOE6jRo2k/xdCoFq1amjYsKG0ztjYGFWqVMH9+/c19vPw8JD+38DAAK1atUJcXJzWWM+cOYPff/9do+92rqtXr6Jly5YIDQ3F8OHDsWbNGnTq1Al9+/ZFvXr1Crx+Kyur/D+gIti2bRvCwsKwadMmODk5Seu//vprhISEwN3dHUII1K5dG0OHDsWqVau0HmfFihVo0aIFGjdurLF+06ZNWLt2LTZs2IAGDRogJiYG48ePh7OzM4YNG1Yi18AYY3Lguqh4dVFcXBzGjh2L6dOnw9vbGwkJCQgLC8P777+PtWvXFuoYuU/V/Pz8EBoaCgBo0qQJTp8+jUWLFmkMiMRYWcPd/JisVq9ejZMnT6Jdu3bYuXMn6tati3379gEAwsPDERERgbCwMOzfvx8xMTHo1asXMjIyNI5hbGys8bMQQuu63D/WusjOzoaPjw9iYmI0lr///hs9evQAAMyaNQtxcXHo1asX/vjjDzRq1CjfRgtQct38tm7disDAQKxduxa+vr4a22xsbLBjxw6kpqbi5s2buHTpEszNzeHi4pLnOPfv38dPP/2EkJCQPNvCwsIwadIkDBw4EA0bNkRgYCBCQ0N5AArGWLnAdZHuddFnn32Gli1bIiwsDI0aNYK3tzeWLFmCyMhI3LlzBwBgZ2eHpKQkjf2SkpJgZ2cHALC2toaRkRHc3d01ytSvXx+3bt3S6bNi7HXhJ1NMdo0bN0bjxo0xZcoUdOvWDWvWrIG3tzeOHj0KX19fBAYGAlB3SY2Pj4elpWWJnPfEiRPw9PSUjn3q1Cm88847Wss2a9YMmzdvhpOTU57K8UWurq5wdXXFuHHjMHLkSKxcuRLvvvuu1rIjRozQGDFPGwcHhwK3b968GcHBwVizZk2+sQOAUqmEg4MDMjMzsW3bNq3n/f7776FQKODv759nW1pamvT+QC5DQ8NifSlgjLGyhOui/BVUF+VXPwD//8TJw8MDUVFRCAsLk8pERUXhrbfeAgCYmJigRYsWuHz5ssZx4uPjNXpbMFYWcWOKyeb69etYtmwZevbsCQcHB1y7dg1//vknRo4cCQBwc3PDpk2bcPToUVhbW+Pbb7/F9evX0bRp0xI5/9KlS+Hm5oaGDRtiyZIluHnzpnTul40ePRorVqzAgAEDMGXKFNjY2ODatWvYvHkzIiIiYGRkhEmTJqFfv36oVasWkpKScPToUbRq1Srf8xe3a8XGjRsRGBiI8PBwtGvXTprTw8TERDruyZMncffuXTRp0gR3797FrFmzkJ2djcmTJ2sci4iwcuVKDBw4UGv3EV9fX8yfPx/Ozs5o0KABzp07hy+++AJBQUE6x88YY2UB10XFq4t8fX0REhKCpUuXSt38JkyYgGbNmkldJcePH4927dph/vz56NWrF3788UccPHhQYzCkyZMno3///nj77bfh6emJgwcPYuPGjdixY4dUJjExEYmJiYiPjweg7mL46NEj1KxZs8S6zTNWZHK/tMVL2VjkeCE5MTGRevfuTdWrVycTExNydHSksLAwysjIICL1S7W9e/eWhn8NCwujkSNHUvv27aVjtG/fnkaPHq1x3AYNGtDMmTM11tna2tK3335LRP//Qu+6devIw8ODFAoFubm50Z49e6Ty2oajjY+Pp759+5KlpSUplUpyc3OjMWPGUHp6OqWnp5O/vz85OTmRiYkJ2dvbU0hICD1+/LiEP7X/1759ewKQZ3nx8zl06BDVr1+fFAoFVa1alQIDA+nu3bt5jnXgwAGNF5hf9uTJExo/fjzVrFmTlEolOTs704cffkj//fdfaV3eK/HLtbzwotvCA1Bo4rqo+L755htyd3cnlUpFdnZ2NGjQILp9+7ZGmS1btlDdunXJ2NiY6tWrR9u2bctznNWrV5OrqysplUpq2LAhbdiwQWP7zJkztdZ7q1evLs3L00tcR76+RRCRDE04VtY4Ojreu337tr3ccbwON27cgLOzM6Kjo9G8eXO5w2E6cnR0TLh9+3Z1ueNgTN9UpL/3ZRnXRaw0cR35+vAAFIwxxhhjjDGmA25MMcYYY4wxxpgOeAAKVuHUqlUL3L2VMcaYnLguYqx84CdTjDHGGGOMMaYDbkyxMqVDhw4YM2aM3GG80qxZsyCEgBAC8+fPlzscvTBkyBDpM9u6davc4TDGGACudyq6WrVqSZ/rgwcP5A6H6SFuTDGmo7p16yIhIQFjx44FAGRmZmLKlClo1KgRzMzMYG9vj0GDBuWZvX358uXo2LEjLC0tIYTAjRs38hw7Pj4evXr1grW1NSwsLNC6dWvs3btXo8ytW7fg6+sLMzMzWFtbY9y4ccjIyCjSNbzYwHlxMTMz0yi3ePFi1K9fHyqVCnXr1sXatWvzHOvrr79GvXr1oFKpUKNGDYwePRopKSka2xMSEooUH2OMsf/3cr2TKz4+Hn369IGlpSVMTU3RrFkzXLx4Udp+9epV9O7dGzY2NqhUqRL69++PpKQkjWO82KjIXaZOnapRZvz48WjevDmUSiVq1aql0zU8e/YMQ4YMQaNGjWBsbIwOHTrkKZOQkIBBgwahXr16MDQ0xJAhQ/KUuXDhAt555x24uLhACIFZs2ZpPd+SJUvg7OwMpVKJN998E0eOHNHYHh0djW3btul0LYwB3JhiTGdGRkaws7OTGh5paWk4e/YsPv74Y5w9exY//fQTbt++ja5du+L58+fSfmlpafDy8sr3Dz8A9OjRA8+ePcP+/ftx7tw5tG3bFn5+frh69SoAICsrCz4+Pnj69CmOHDmCH374AVu3bsXEiROLdA25DZwXFxcXF/Tv318qs3TpUkyZMgUzZszAhQsXMHv2bIwePRq7du2SymzYsAGTJ0/Gxx9/jIsXL2Lt2rXYs2cPxo8fL5WpXLky7OzsihQfY4yx//dyvQOoJx1u06YNnJ2dceDAAZw/fx5z5syRJmBPTU2Fl5cXiAgHDhzAsWPHkJGRAV9fX2RnZ2scf8aMGRr1wbRp0zS2Z2dnIzg4uFgTtmdlZUGpVGLMmDHw8fHRWiY9PR3W1taYOnVqvhMOp6WloVatWpgzZw6cnZ21ltm0aRPGjx+Pjz76COfOncNbb72Fbt26adzktLGx4Ql/WfHIPdEVL2VjKe4kjsuWLaNq1arR8+fPNdb7+/uTr68vERFduXKFevbsSba2tmRqakpNmzalXbt2aZR/eeJDJycnWrhwYYFl0tPTafLkyeTg4EAqlYqaN29Oe/fuLc7lvNLMmTOpQYMGryx34cIFAkB//vlnnm3R0dEEgK5fv66x/p9//iEAdODAAWldZmYmGRgY0JYtW4iIaM+ePSSEoFu3bkllIiMjSaFQFGtyxqNHjxIAOnbsmLTOw8ODJkyYoFEuNDSU2rRpI/08evRoateunUaZGTNmaP2MAEjXURw8ISEvvOi2lJdJe7neUfP396dBgwblu9++fftICEHJycnSukePHpEQgqKioqR12q47PwsXLiQnJ6fCB5+P0aNHa0x+rI2Pjw8FBwcXWEbbBMlERC1btqThw4drrKtTpw5NnTpVY93BgwcJAP3zzz+FCVsvcB35+hZ+MsVKRL9+/fD48WNERUVJ61JSUvDTTz9h8ODB0s/dunVDVFQUYmNj0bdvX/Tp0weXLl0q1rmHDh2Kw4cPY8OGDTh//jyCg4Ph6+uL2NjYfPeZN28ezM3NC1xe7gqgiydPngAAqlSpUuh9qlativr16yMyMhIpKSnIysrC8uXLYWFhgTZt2gAAjh8/jvr168PR0VHaz9vbG+np6Thz5ozO8a5YsQINGjTAW2+9Ja1LT0+HUqnUKKdSqXDq1ClkZmYCANq2bYuYmBicOHECgLoL4s6dO9G9e3edY2GMsYJwvaN+UrRr1y64u7uja9eusLGxQYsWLbBp0yapTHp6OoQQGn/HlUolDAwMcPToUY3jhYeHo2rVqmjSpAnmzp1b5K7jZUlGRgbOnDkDLy8vjfVeXl74448/ZIqKlUc8NDorEVWqVEH37t2xfv16dO3aFQCwY8cOGBkZoWfPngCAxo0bo3HjxtI+H3/8MXbt2oWtW7fm6UpQWFevXsUPP/yAGzduoGbNmgCAMWPG4LfffsOyZcuwZMkSrfuNGDFCoyubNg4ODjrFlCsjIwMTJ06Er68vatSoUej9hBCIiopC7969UalSJRgYGMDKygq//PIL7O3tAQCJiYmwtbXV2M/a2hqGhoZITEzUKd7Hjx9j8+bN+OyzzzTWe3t743//+x/69OmD5s2b48yZM1i5ciUyMzPx4MED2NvbY+DAgXj48CHatWsHIsLz588RGBiIzz//XKdYGGPsVbjeAe7fv4+UlBTMmzcPn376KebPn48DBw4gICAA5ubm8PHxQevWrWFubo6wsDDpb/LUqVORlZWl8R7ruHHj0LRpU1StWhWnTp3C1KlTcf36daxcubJIMZUVDx48QFZWVp660tbWFr/99ptMUbHyiBtTrMQMHjwYwcHBSEtLg6mpKdavX4++fftKd8NSU1Mxe/Zs7N69GwkJCcjMzMSzZ8/QqFEjnc959uxZEBHc3d011qenp8PT0zPf/aysrEq1j/Tz588xePBgPHr0CDt37izSvkSEUaNGoWrVqjhy5AhUKhVWrlyJvn37Ijo6utiNvPysW7cO2dnZCAwM1Fg/ffp0JCYm4q233gIRwdbWFsHBwViwYAEMDNQPtw8fPoxPP/0US5YsQatWrXDlyhWMHz8eM2fOxCeffFIq8TLGWEWvd3LfefLz80NoaCgAoEmTJjh9+jQWLVoEHx8f2NjYYMuWLRg5ciSWLFkCAwMD+Pv7o1mzZtLfcADS/gDQqFEjVKpUCQMGDMDnn3+OqlWrlmjcjJUn3JhiJcbHxwdGRkb46aef0KlTJ/z222/Yt2+ftH3SpEnYu3cvwsPD4erqClNTUwQFBRXYjcDAwABEmpMa5nYtA9QViRAC0dHRMDY21iinUqnyPe68efMwb968Aq/nl19+wdtvv11gGW2eP38Of39//PXXXzh06FCRK6EDBw5g165dSE5OhqWlJQD1aERRUVFYvXo1pk2bBjs7Oxw7dkxjv9y7cLoO8rBixQr07ds3T2WvUqmwatUqLFu2DElJSbC3t5e6HdrY2AAApk2bBn9/fwwfPhwA0LBhQ6SmpmL48OGYMWMGjIz4Tw1jrORV9HrH2toaRkZGeRp29evXx8aNG6Wfvby8cPXqVTx48ABGRkawtLSEnZ0dXFxc8j127sAPV65c0cvGVG5vjZdHLUxKSuLBkFiJ4m84rMQoFAr069cP69evx4MHD2BnZ6cx5OnRo0cRFBSEvn37AlAPj3r16lW4ubnle0wbGxuNbgjPnj3DpUuX0LRpUwBA06ZNQURITExEx44dCx1raXXzy8zMxMCBA3H+/HkcOnRIpz/YaWlpAKBxxzD359y7kB4eHpgzZw7u3LkjdSGMioqCQqHAm2++WeRznjp1CrGxsfjqq6/yLWNsbCyda+PGjejRo4cUY1paGgwNDTXKGxoa5vlCwhhjJami1zsmJiZo0aIFLl++rLE+Pj4eTk5OecpbW1sDUN+0u3//vtQdUpuYmBgAkLqX6xsTExO8+eabiIqKQr9+/aT1UVFR0u8DYyWBG1OsRA0ePBidOnXC9evX4e/vr9EgcHNzw48//gg/Pz8YGxtj9uzZePbsWYHH8/T0xKpVq9CzZ0/Y2Nhg7ty5GsOMu7m5ISAgAEOGDEFERASaNWuG5ORkHDp0CC4uLujTp4/W45ZGd4vnz5+jX79+iI6Oxq5duyCEkN5fqly5snTHMjExEYmJiYiPjwcAxMXF4dGjR6hZsyasrKzg4eEBKysrDB06FDNmzIBKpcKKFStw7do19OjRA4D6LmODBg0QFBSEiIgIPHz4EGFhYQgJCUGlSpWKHPvy5cvh6uqqdb6P+Ph4nDx5Eq1bt8a///6LL774AufPn8eaNWukMr6+vvjiiy/QvHlzqZvf9OnT0aNHD34qxRgrVRW53gGAyZMno3///nj77bfh6emJgwcPYuPGjdixY4dUZvXq1ahXrx6qVauG48ePY/z48fjggw9Qt25dAOpBjU6cOIGOHTuicuXKiI6OxgcffICePXtK74UB6qdUKSkpuHfvHjIyMqQGl7u7O0xMTAodc1xcHDIyMvDgwQOkpKRIx2nSpIlUJnfdkydPYGBggJiYGJiYmEhP4TIyMhAXFwdA3eBNTExETEwMzM3NUadOHQDqrouBgYFo2bIl2rRpg++++w737t3DiBEjivIRM1YwuYcT5KVsLCU1VG52djY5OTkRAIqNjdXYduPGDerUqROZmpqSg4MDLVy4MM+Qpy8PP/v48WMaOHAgVapUiapXr06LFy/OUyYjI4NmzpxJzs7OZGxsTLa2tuTr60unT58uiUvSStsQtdevXycAWpfVq1dr7PuqMtHR0eTl5UVWVlZkYWFBLVu2pN27d2uc7+bNm+Tj40MqlYqsrKxo7Nix9OzZszzxvHhcbZ48eUJmZmb0+eefa90eFxdHTZo0IZVKRZUqVSI/Pz+6dOmSRpnMzEyaNWsW1alTh5RKJdWoUYNGjhypMRRvLvDQ6LzwIutSXoZGz1WR651cq1evJldXV1IqldSwYUPasGGDxvYpU6aQra0tGRsbk6urK0VERFB2dra0/cyZM9SqVSuqXLkyKZVKqlu3Ls2cOZNSU1M1jtO+fXut9deLU3wA0DpM+Yty/71eXl6kbfuLw7HnV+e+PNT64sWLycnJiUxMTKhZs2Z0+PDhPPHw0Oi8FGcRRNwNhwGOjo73bt++rZ/P8mUwa9YsbN26FefPn5c7lHwdPHgQ3bt3x4ULFwrsF/+6CSGwZcsWvPPOO8U6jqOjY8Lt27erl1BYjFUY/PdeP+lDvXP9+nXUrl0bR44ckaby0AeHDh1Cx44d8c8//0hdIfUd15GvD88zxZiOLl68CHNzc3zxxRdyh6LVnj17MGXKlDLTkBoxYgTMzc3lDoMxxvSWPtQ7QUFBetWQatCgAbp16yZ3GEyP8ZMpBoDvVBZVcnIykpOTAahf6M0ddY/l7/79+9Ikxvb29jAzMyvW8fiuG2O64b/3+onrndJx8+ZNabRGFxeXPIM/6SuuI18ffjOcMR2U9jxV5VG1atVQrVo1ucNgjDG9xPVO6dA26iFjRVE+mt+MMcYYY4wx9ppxY4oxxhhjjDHGdMCNKcYYY4wxxhjTAb8zxQAAGRkZdxwdHeUOo9QRkRkRGRkYGDyWO5aSlp2dXUUI8Z8QouAZKcuJjIyMO3LHwJg+qih/78swkZ2dXdXAwOBfAFlyB1PCjLOzsysZGBg8lDuQio7ryNeHR/NjFYYQwgNAOABPIkqXO56SJoSwBHAAwEAiipc5HMYYYy8RQhgA+BHAGiLaLnc8pUEIMRpAQyIaIXcsjL0O3M2PVQhCCBsASwEMKo8NKQAgokcAhgKIFEIUb9xxxhhjpWEygL/La0MqxxIAlYQQgXIHwtjrwE+mWLknhDAEsAfAV0T0i9zxlDYhxFAAHQEEEyc4Y4yVCUKIjgBmAPAioky54ylNQghzAAcBvEtEf8kdD2OliZ9MsYpgJoCTFaEhBQBEtBpABoD35I6FMcYYIISoDuArAAHlvSEFAESUAiAIwPdCiEpyx8NYaeLGFCvXhBDdALQEMFvuWF6zsQCGCCGayx0IY4xVZEIIYwDrAYwnontyx/O6ENFFAAsArBJCCLnjYay0cDc/Vm4JIZwA/ASgMxE9kDue100I4QJgO4BORMQjKzHGmAyEEOEA/iGiz+WORQ5CiG8A3CCiL+SOhbHSwE+mWLkkhFAA2ABgREVsSAEAEV2Dun/+mpwRpBhjjL1GQog+AOpA/YSmopoEoJcQoq3cgTBWGvjJFCuXhBBLAcQR0bdyxyI3IcRnAFKJaI7csTDGWEUhhHADsBHq6TgeyRyOrIQQjgB2AfAmoiS542GsJPHdalbuCCEGA7AEsEjmUMqK6QA6CCE6yx0IY4xVBEIIUwCRAIZV9IYUABDRbQBhANYJIYzkjoexksRPpli5IoR4A8D3ADrkjCbEAAghbAHsA9CDiHhWdMYYKyU5gy18D+B3IvqfzOGUKUKI6QBURPSR3LEwVlL4yRQrN3KGX10DIIgbUppyulWMBbBeCGEidzyMMVaOhQDI4oaUVnMBNBZC+ModCGMlhZ9MsXIh507gJgA/EtEPcsdTVgkhJgJwJKIJcsfCGGPljRDiTQCLAXQkov/kjqcsEkJUBfAbgL45AyUxptf4yRQrL8YDuM8NqVf6AoCjEKK/3IEwxlh5IoSwArASwGBuSOUvZ6qO96B+f0opdzyMFRc/mWJ6TwjRBsB8qOeTSpc7nrJOCFEZwAEAAUR0Se54GGNM3+VMP/ETgP8R0Q6Zw9ELQoiRAJoS0Xtyx8JYcfCTKabXhBDVoO5SEcANqcIhoscAhgKIFEKYyx0PY4yVA1MBXOSGVJF8B8BUCDFE7kAYKw5+MsX0lhDCEMAvACKIaJ/c8egbIUQwAC+ou6TwHwLGGNOBEKITgI8BeBHRc7nj0SdCCDMABwGEEFGs3PEwpgt+MsX02ScA/uCGlG6IaA2AVAAj5Y6FMcb0kRDCAep3UQO4IVV0RJQKIBDA6pwu6IzpHW5MMb0khPAB0AzAp3LHoufGARgshGgpdyCMMaZPhBDGANYDGEtECXLHo6+I6DKAz6BuUAm542GsqLibH9M7QghnAD9CPeDEA7nj0Xc5n+d2qD/Ph3LHwxhj+kAI8SWAe0S0UO5YyoOcz/MuEYXLHQtjRcFPppheyRlGdT2A97ghVTKI6DqAaVAPSMF/Exhj7BWEEP0AOAHgL/4lZzIAXyFEO7kDYawo+MkU0ytCiGUA/iSixXLHUt4IIeYCSCeiT+SOhTHGyiohRF0APwDwJKJHModTruS8g/YzgK5ElCh3PIwVBt+FZnpDCBEEwALAErljKadmAGgrhPCSOxDGGCuLckafiwQwlBtSJY+I7gKYCGC9EMJI7ngYKwx+MsX0ghCiIYBVADoSUYrc8ZRXOfN2/QrAl4huyx0PY4yVFTmDI6wFcICIVssdT3kmhPgYgAURTZU7FsZehZ9MsTIvZ7jU7wEEcUOqdBHRfQCjob4raCJ3PIwxVoa8D3VXaG5Ilb7PADQQQvSUOxDGXoWfTLEyLedO4BYAW4hok9zxVBRCiA8AOBPROLljYYwxuQkhmgP4Fur3pP6TO56KQAhhBWA/gHeI6Krc8TCWH34yxcq6D6AeepYbUq/XVwDshRAD5Q6EMcbkJISoCmAl1BPzckPqNSGiZAAhANYJIVRyx8NYfvjJFCuzhBBtAcyDev6jDLnjqWiEEJUAHAQwmIguyh0PY4y9bjnTRewEsJyIdsodT0UkhHgfQAsiGi53LIxpw0+mWJkkhLAFsAjqO4HckJIBET0BMATAWiGEuczhMMaYHD4C8Bc3pGS1HICJEGKo3IEwpg03pliZkzMc6noAYTyinLyI6C8A3wBYnvP+GmOMVQhCiC4AOgCYLnMoFRqpu1CNADBCCNFE5nAYy4O7+bEyRwjxGYA0IvpU7liYmhBiKYALRLRI7lgYY6y0CSEcAewG4EVESXLHwwAhhBuAjeDJklkZw0+mWJmSMwxqIwBz5Y6FaZgAwF8I0VruQBhjrDTlTAuxHsAYbkiVHUQUD2AOgNXcU4KVJfxkipUZQggXANsBdCKih3LHwzQJIWoB2AH1gCAP5I2GMcZKhxDiawC3iChC7lhYXkKICABJRLRA7lgYA/jJFCsjhBBKqO8EhnBDqmwiohtQv4wdKYQwlDkcxhgrcUKIAQBqAPhC7lhYvqYC6CaEaC93IIwB/GSKlRFCiBUAzhLRUrljYQUTQnwKIIuIZskdC2OMlRQhRD0AGwB0JKLHcsfD8ieEqA5gD4BuRJQgdzysYuMnU0x2QoghAFQAvpM5FFY4swB4CCG6yh0IY4yVhJzpHyIBDOGGVNlHRPcAfABgfc4IwIzJhp9MMVkJIRoDWAH1ncBUueNhhSOEsAEQBaAnEd2SOx7GGNNVzmAG6wD8SkRr5I6HFZ4Q4kMAVYhostyxsIqLn0wx2QghKgNYDSCQG1L6hYj+ATAS6ruCCrnjYYyxYhgJIJUbUnrpcwD1hBC95A6EVVz8ZIrJIudO4DYAPxDRFrnjYboRQowH4EpEY+SOhTHGikoI0RLAV1DPXfRM5nCYDoQQVQAcANCPiK7IHQ+rePjJFJPLRAA3uSGl974BYCOEGCR3IIwxVhRCiKoAlgEI4IaU/iKifwEMA7BOCKGSOx5W8fCTKfbaCSHaAfgU6vmKMuWOhxWPEMICwEEAQUQUJ3c8jDH2KkIIAwC7ASwmop/ljocVnxAiBIAHEb0rdyysYuEnU+y1EkLYQf00YxA3pMoHInoKIBjAmpyGFWOMlXXTAJzjhlS5shKAgRBimNyBsIqFG1PstckZvnQ9gIlEdFfueFjJIaILUL93sCLnfTjGGCuThBBeANoCmCF3LKzkkLqr1SgAIUKIpnLHwyoO7ubHXhshxHwAT4lortyxsNIhhFgM4BIRfSt3LIwx9jIhhCOAXQC8iOi+3PGwkieEqANgM9SDijySORxWAfCTKfZaCCH8ADQA8JncsbBSFQqgvxDCQ+5AGGPsRUIIEwAbAIzmhlT5lTOi36dQdz3n77ms1PGTKVbqhBC1AWwF0ImIkuWOh5UuIYQTgJ8AdMmZj4oxxmQnhPgWwDUi+lLuWFjpE0IsBPCQiObLHQsr37jFzkpVzjCl6wGEcEOqYiCimwCmQj1MraHc8TDGmBBiIAA7qN/tZBXDhwC8hBAd5Q6ElW/8ZIqVKiHE/wCcIqJlcsfCXi8hxGyo/8bwS96MMdkIIdwBRALoSERP5I6HvT5CCHsAvwDoTkT35I6HlU/8ZIqVGiHEuwCMACyXOxYmi08AtBBCdJM7EMZYxSSEMAewBkAwN6QqHiJKADABwHohhLHM4bByip9MsVIhhGgC9czyHYkoTeZwmEyEENYAfgPgl9P9jzHGXoucaRo2APiZiNbJHQ+TjxBiCgAbIpokdyys/OEnU6zECSEsAawCEMgNqYqNiB4AGAH1XUGF3PEwxiqU0QAecUOKAVgAoI4Qoo/cgbDyh59MsRKVcydwO4BIItoudzysbBBCjAVQn4hGyR0LY6z8E0K0BhAB9VxD6XLHw+SXc6P3AIABRPS3zOGwcoSfTLGSFgb10LPckGIvWgSgihAiQO5AGGPlW0734u8ABHBDiuXKmcB3GNQjzZrKHA4rR/jJFCsxQoj2AGZBPbN8pszhsDIm50XwgwCGENEFueNhjJU/OdMx7AbwLRHtkTseVvYIIYYBeBvAUOIvwawE8JMpViJyhh/9Guo7gdyQYnkQUQqAYKhnpbeQOx7GWLk0HcBpbkix/BDR/wBkAxgudyysfODGFCs2IYQR1BPzfsDzOLCCEFEc1O8xrMx5v44xxkqEEKIrAA+oe0gwVpDRAN4VQjSTOxCm/7ibHys2IcQCAP8S0Wdyx8L0gxDiWwBXiOhruWNhjOk/IURNADsBdCGif+SOh5V9QojaALYA6ERE/8odD9Nf/GSKFYsQoheAegA+lzkUpl8mAugrhHhL7kAYY/otZ9qF9QBGckOKFRYRXQUwG+qu5/x9mOmMn0wxnQkh6gDYDL6rw3Twwp1kLyK6L3c8jDH9JIRYBCCeiL6ROxamf4QQn0M9Hxn3rmE64ZY400nOsKLrAAzjhhTTBRHdAjAZQGTOCFyMMVYkQohBAKwBfCt3LExvfQygsxDCU+5AmH7iJ1OsyHIGDlgF4BgRrZQ7HqbfhBAzARgT0TS5Y2GM6Q8hRAMAawB0JKKncsfD9JcQwg7AXgA+RHRX7niYfuEnU0wXw3L++z9Zo2DlxacAmgkhfOQOhDGmH3KmV1gDIJgbUqy4iCgRwDgA64UQxnLHw/QLP5liRZIzjOgSAJ5ElCZ3PKx8EEJYA/gNQC8iuiFzOIyxMiynd8RGAD8R0Qa542HlhxAiDIA9EYXKHQvTH/xkihWaEKIKgJUABnNDipUkInoA4D2o7woq5Y6HMVamjQXwgBtSrBSEA6glhHhH7kCY/uDGFMuXEKKBEKJ9zv8bAPgewKdEdEXWwFi5RESnAGwA8FXuOiGEnxDCQbagGGOyE0KMfOH/PQD0B8BPDliJI3V3rXcBfCSEcAMAIYSxECJE3shYWcaNKVYQPwB2Of8/GeqhZ3+UMR5W/i0BUEkIEZjzc20AnWWMhzEmIyGEEYCQnP+3AbAUQAARpcsaGCu3iOgRgKFQjzRrBiALwGhZg2JlGjemWEGaAjgrhOgIwAvAhzLHw8q5nLuC7wEYJ4RoCOAsgGbyRsUYk1FdABdzpk9YB2AqEd2UOSZWzhFRLIDvoG68E4BHQggreaNiZRU3plhB6gBIA/AlgAAAWTkTrTJWKoQQNYkoBUAggNUArkLdqGeMVUzNAJwDMBPACSLaK4SwyhnNj7ESJ4SoIoSwIKLVANKhvsF3DlwXsXxwY4pplTPYRArUdwLHA2gO4AiACTKGxcqxnBG65gkhfoW6e+kCAF8AqJzzzh5jrOJpBsAIQAsAy4QQC6Ee+dNF1qhYedYYwFEhxAwA0wEEA/gX3EuC5YOHRmda5XTtiwBwAYAr1N2tFvCw1ay0CSHcoe5S6gjgEQBnAP2J6LKccTHGXj8hxAkAlaG+mdcE6hssW4goS864WPkmhFAACAIwCsBxqN/d/YuI+soaGCuT+G4vy89AAA0APADQl4hGcUOKvQ5EFEdEgQCGA3gI9e9hb3mjYoy9bjlPq98E8BzALwBaE9FGbkix0kZE6US0AuonoscBGALoIm9UrKziJ1NMKyFEIwBPuAHF5CaEeAPAf0R0Ve5YGGOvlxDCB8Ae4i8rTEY5A6B4EdEvcsfCyh5uTDHGGGOMMcaYDribH2OMMcYYY4zpwEjuAArL1tb2lImJSQ2542D6ISMj405SUlJLueOQC+cLKw2cV5xXrGg4ZzhnWMkqizmlN40pExOTGrdv37aXOw6mHxwdHeUOQVacL6w0cF5xXrGi4ZzhnGElqyzmFHfzY4wxxhhjjDEdcGOKMcYYY4wxxnTAjali6NChA8aMGVOkfWrVqoXw8PBSiogx/cI5xFjxcA4xVjycQ6y4uDFVDNu3b8dnn31WpH2io6MxatSoUopI7datW/D19YWZmRmsra0xbtw4ZGRk5Fs+OTkZY8eORb169aBSqeDo6IiRI0fi4cOHUpns7Gz07NkTNWvWhFKphL29PQYPHoy7d+/mOd66devQpEkTKJVKWFtbIygoqFSuk+m/8pJDLyIidOvWDUIIbN26VWuZZ8+eoXHjxhBC4PTp09L6f/75B97e3qhevToUCgUcHR0xevRoPH78uESui5U/5SmHQkJCULt2bahUKtjY2MDPzw8XL17UWja/HAIAIUSe5bvvviuxa2PlS3nJocJ8lwOAuXPnok2bNjAzM4N6Xuy89u/fj7feegsWFhaws7PDlClT8Pz58xK9vvKEG1PFYGVlBQsLiyLtY2NjA1NT01KKCMjKyoKPjw+ePn2KI0eO4IcffsDWrVsxceLEfPe5d+8e7t69iwULFuCvv/7CunXr8Pvvv8Pf31+jnKenJzZv3ozLly9j27ZtuHbtGnr37q1R5ptvvkFYWBgmTZqE8+fP4+DBg/Dz8yuVa2X6r7zk0IsiIiJgYFDwn9ZJkyahRo28A1wZGBigd+/e2LVrF+Lj4/H9999j//79CAkJ0elaWPlXnnKoefPm+P7773Hx4kXs27cPRITOnTsjMzMzT9n8cijXihUrkJCQIC3BwcHFvi5WPpWXHCrsd7n09HT06dMHEyZM0Hqc2NhYdO/eHV26dMG5c+ewadMm7Ny5E1OnTi3JSyxfiEgvlho1atyj1yglJYUCAwPJzMyMqlWrRvPmzSMfHx8KDg6WyrRv355Gjx4t/ezk5ESffvopvffee2RhYUEODg60YMECjeM6OTnRwoULSy3uPXv2kBCCbt26Ja2LjIwkhUJBjx8/LvRxfv75ZxJCFLjPTz/9RADov//+IyKif//9l0xNTenXX3/V/QJKSM7vi+y/t3ItrztftKmIOXTq1CmqUaMGJSUlEQDasmVLnjI7duwgd3d3iouLIwAUHR1d4DG//vprsrOz0+1iShjnFddDhVFS9VBsbCwBoEuXLmmsf1UO5Zd7cuCckbcuqug5VNB3uS1btpC6GaDpww8/pCZNmmis27lzJymVSnry5EkRrqJ0lMWc4idT+Zg4cSIOHz6MH3/8EQcOHEBsbCyOHDnyyv2+/PJLNGzYEGfPnsWUKVMwefJkHD9+vNDnPXLkCMzNzQtc5s2bl+/+x48fR/369TWGjvT29kZ6ejrOnDlT6DiePHkChUKR752X5ORkrF+/Hq1atYJSqQQA/Prrr8jKykJSUhLc3d3h4OCA3r1749q1a4U+Lys/KloOPX36FIMGDcLy5ctRrVo1rWXu3LmDkSNHYsOGDVCpVK+8lnv37mH79u1o3779K8uy8qei5dCLUlNTsXr1atSsWRO1atWS1hc2h8aPHw9ra2u0aNEC3333HbKzswt1Xla+VOQcAl79XU6b9PR06XtdLpVKhWfPnhXp3BWJ3swz9TqlpKRg1apVWLt2Lbp06QIA+N///ldgl4JcXl5e0ouMY8eOxTfffIP9+/fDw8OjUOdu3rw5YmJiCixjZWWV77bExETY2tpqrLO2toahoSESExMLFcOjR48wffp0hISEwMhI81dkypQpWLRoEdLS0tC6dWvs3r1b2nbt2jVkZ2djzpw5+Oqrr2BlZYVPPvkEHTt2xMWLF0v1kTgrWypiDo0YMQJdu3ZFt27dtG7PyspCQEAAJk6ciMaNG+PGjRv5Hsvf3x8//fQT/vvvP/To0QOrV68u8HpY+VMRcwgAlixZgsmTJyM1NRV169bF/v37oVAoABQ+h3LrHXNzc+zfvx8TJ07EgwcPMG3atALPzcqXippDuQr6LlcQb29vfPnll1i3bh0GDhyIpKQkfPLJJwCAhISEQh+nIuHGlBZXr15FZmYmWrb8/wmWzczM8MYbb7xy30aNGmn8XL16ddy/f7/Q51apVKhTp07hgy1hKSkp8PX1hYODAxYsWJBne1hYGIYNG4abN29i9uzZGDx4MH755RcIIZCdnY3MzEx888038PLyAgCsX78ednZ22LVrFwYMGPC6L4fJpKLlUGRkJGJjY/O8CP+iefPmwcTEBKGhoa883pdffomZM2ciPj4eH374ISZMmIBly5aVZMisjKtoOZQrICAAXbp0QUJCAsLDw9GvXz8cO3YMpqamhc6h6dOnS//fpEkTZGVlYe7cudyYqmAqag4Br/4uVxAvLy+Eh4dj9OjRGDJkCBQKBaZPn44jR4688n3gioo/lRJmbGys8XNuI6Owivto2M7ODklJSRrrHjx4gKysLNjZ2RV47pSUFHTv3h0AsHv37jyPeQH1nRE3Nzd06dIFGzduxL59+3D06FEAgL29epJzd3d3qXzlypVRvXp13Lp1q3AfAKvw9DGH9u/fj7i4OJibm8PIyEi6CzhgwAC0bdtWKnPgwAEYGxvDyMhIqmhbt26NgICAPDHUq1cPPXv2xLJly7B8+XLcvn270J8Bq9j0MYdyVa5cGa6urmjXrh22bt2K+Ph4bNu2DUDRcuhFrVq1wpMnT/LExFh+9DmHCvNd7lVCQ0Px6NEj3Lp1Cw8ePJAGEnNxcSnysSoCfjKlRe3atWFsbIzo6GjpFyctLQ3nz59H7dq1S/XcxX007OHhgTlz5uDOnTvSo+yoqCgoFAq8+eab+e739OlTdOvWDUSEvXv3wtzc/JWx5v5hSU9PBwC0adMGAHD58mXp3CkpKUhISICTk9Mrj8fKj4qWQ3PnzsWkSZM01jVs2BDh4eFSJbR69WqkpqZK2+/duwdvb2+sX79eyh1tXs4zVjFUtBzSJvfl7tzffV1zKCYmBkqlEpaWloU+N9N/FTGHdPkulx8hBKpXrw4A+OGHH+Do6IhmzZrpfLzyjBtTWpibm+Pdd9/FlClTYG1tDXt7e8yZMwfZ2dn5jslfUor7aNjLywsNGjRAUFAQIiIi8PDhQ4SFhSEkJASVKlUCAJw6dQpBQUFYu3YtWrZsiadPn8LLywtPnjzBjh07kJqaKlVYVlZWMDExwfHjx3H27Fm0bdsWlpaWuHr1KqZPn45atWpJd97d3Nzg5+eH8ePHY9myZahSpQpmzpyJatWqoUePHsX/cJjeqGg55ODgAAcHhzzHcnR0lCpxZ2dnjW25lVzt2rWlynL37t14+PAh3nzzTZibm+PChQsICwtD69atZe0ywl6/ipZDV65cwbZt29C5c2fY2Njgzp07mD9/PhQKhVR/FCaHdu3ahcTERHh4eEClUuHgwYOYMWMG3nvvPendK1YxVLQcKsx3OUA9f1VycrL0zmFuo69OnTpSTi1cuBBdu3aFgYEBtm/fjvnz52Pz5s0wNDTU+ZrKM25M5SM8PBypqano2bMnzM3N8cEHHyApKUmnx6Wvk6GhIX7++WeMGjUKbdq0gUqlQkBAABYuXCiVSUtLw+XLl5GWlgYAOHPmDE6cOAFA3SB60cGDB9GhQweoVCps3boVM2bMQGpqKuzt7dG1a1ds2rRJ4zOJjIxEaGgofH19QURo27Yt9u/fz4NPVEAVKYdKilKpxHfffYeLFy8iPT0djo6O6N27N8/vUUFVpBxSKBQ4dOgQIiIi8OjRI9ja2qJdu3Y4fvz4K7s1vcjY2BhLlixBaGgosrOz4eLigk8++QSjR48u8etkZV9FyqHCfJcDgBkzZmDNmjXStqZNm+Yp88svv2Du3LlIT09H48aN8dNPP+U7uBIDBBHJHUOhODo63rt9+7a9XOdPT0+Hk5MTwsLCCj15J5OPo6Njwu3bt6vLHYdc5M4XbTiH9B/nFddDrGg4Z8pWXcQ5pP/KYk7xk6l8nDt3DhcvXpQenX7++ed4+vQpj0jHWCFxDjFWPJxDjBUP5xB7HbgxVYAvvvgCly9fhpGREZo0aYLff/+9UPMTMMbUOIcYKx7OIcaKh3OIlTZuTOWjadOmBc4ZwxgrGOcQY8XDOcRY8XAOsdeB55lijDHGGGOMMR1wY6qMuHHjBoQQfAeFsULinGGseDiHGCseziEGcGOKFdHixYtRv359qFQq1K1bF2vXrtXY3qFDBwgh8iwNGjSQymzZsgXNmzeHpaUlzMzM0KRJE41hOhkrL7Zv3w4vLy/Y2NjAwsICrVq1ws6dO/OU27ZtG9zd3aFQKODu7o4ff/xRYzsRYdasWahevTpUKhU6dOiACxcuaJSpVatWnrzjIdWZvktISMCgQYNQr149GBoaYsiQIVrLvSqHACA+Ph59+vSBpaUlTE1N0axZM1y8eDFPOSJCt27dIITA1q1bS/qSGHvtMjIyMGPGDDg7O0OhUKBmzZr45ptvpO2ZmZn45JNPULt2bSiVSjRu3Bh79+7VOMbTp08xYcIEODk5QaVS4a233kJ0dLRGGW3f/4QQ5X5qAm5MsUJbunQppkyZghkzZuDChQuYPXs2Ro8ejV27dklltm/fjoSEBGm5ceMGLCws0L9/f6lM1apVMW3aNJw4cQJ//vknhg4dimHDhmHPnj1yXBZjpebw4cPw9PTEzz//jHPnzqF79+7o3bs3jhw5IpU5fvw4BgwYgICAAMTExCAgIAD9+vXDyZMnpTILFixAREQEvv32W0RHR6NatWro0qULnj59qnG+GTNmaOTftGnTXtu1MlYa0tPTYW1tjalTp6JVq1ZayxQmh65fv442bdrA2dkZBw4cwPnz5zFnzhxpktIXRUREwMCAvx6x8mPgwIHYu3cvli9fjsuXL2PLli1o1KiRtH3atGlYunQpvvnmG8TFxWHEiBHo3bs3zp07J5UZPnw49u3bhzVr1uCvv/6Cl5cXOnfujLt370plXqx/EhISpO+HL34HLJeISC+WGjVq3KMScPjwYWrVqhWZmZlRpUqVqEWLFvTXX38REdGDBw9o4MCB5ODgQEqlktzd3WnVqlUa+7dv355GjBhBoaGhVKVKFbK2tqavvvqKnj17RqNGjaLKlSuTo6MjrV27Vtrn+vXrBIDWr19Pbdq0IYVCQXXr1qV9+/blKRMdHS2tu3DhAnXv3p3Mzc3JxsaGBg4cSAkJCdL2P//8kzw9PcnCwoLMzMyoUaNGdODAgZL4mLTy8PCgCRMmaKwLDQ2lNm3a5LvPunXryNDQkG7dulXgsZs2bUpTp04tkTiJiHJ+X2T/vZVrKal8IeKcKWktWrSg0NBQ6ef+/ftT586dNcp06tSJBg4cSERE2dnZZGdnR3PmzJG2p6Wlkbm5OX333XfSOicnJ1q4cGGpxs55pVtecQ6VDB8fHwoODs6z/lU5RETk7+9PgwYNeuU5Tp06RTVq1KCkpCQCQFu2bClWzJwz/N1N7hzat28fVapUif755598y9jb29NXX32lsa5Pnz4UEBBAROo6x9DQkHbs2KFRplmzZvTxxx/ne9zhw4eTm5tbMaLPqyzmVIW69fL8+XP4+fmhbdu2iI2NxcmTJzFhwgQYGhoCAJ49e4ZmzZph9+7duHDhAsaPH4/3338f+/fv1zjO+vXrYWFhgZMnT2Lq1KmYMGECevXqBTc3N5w+fRrBwcEYPnw4EhISNPabPHkyxo0bh5iYGHTp0gV+fn4aLfoXJSQkoF27dnjjjTdw6tQp/Pbbb0hJSYGfnx+ys7MBAIMGDYK9vT1OnTqFmJgYzJo1q8BZvefNmwdzc/MClxfvmL8sPT09z/FVKhVOnTqFzMxMrfusWLECXbt2haOjo9btRIT9+/fj8uXLaNeuXb7nZvLgnClezmjz9OlTVKlSRfr5+PHj8PLy0ijj7e2NP/74A4D6jnpiYqJGGZVKhXbt2kllcoWHh6Nq1apo0qQJ5s6di4yMjCLFxkoe51DJ59DLXpVD2dnZ2LVrF9zd3dG1a1fY2NigRYsW2LRpk8Y+T58+xaBBg7B8+XJUq1atWDGxksM5VLwc2rFjB1q0aIEvvvgCNWrUgKurK8aNG4eUlBSpTH7f744ePSr9G2RlZRVY5mUpKSnYuHEjQkJC8o2t3JC7NVfYpSTubjx8+JAA0KFDhwq9z4ABA2jYsGHSz+3bt6fWrVtLP2dnZ5O1tTX5+vpK6zIyMsjY2Fi6o5V75+LFO8tZWVnk6uoqtehfvrsxffp08vT01IglOTmZANDJkyeJiMjCwoK+//77Ql/Lw4cP6e+//y5wSUtLy3f/Dz/8kKpVq0anTp2i7Oxsio6OJltbWwJA9+7l/ee5fPkyAchzJ4OI6NGjR2RmZkZGRkakUCho5cqVhb6OwiiLdy5e51JSdwM5Z4qXMy9btGgRmZub040bN6R1xsbGtGbNGo1ya9asIRMTEyIiOnbsGAGgmzdvapQZOnQoeXl5ST9HRETQgQMHKDY2llasWEFVq1bV+HcoCZxXRc8rzqGSy6H8nky9KocSEhIIAJmamlJERASdO3eOIiIiyNDQkHbv3i3tM2jQIBozZoz0M/jJlCw58zLOoeLlkLe3NykUCurevTudOHGC9u7dS66urtS3b1+pjL+/P9WrV48uXbpEWVlZ9Ouvv5JKpZJyiEjdO6lt27Z0584dev78OUVGRpKBgUG+T56WLVtGJiYmdP/+/UJfa2GUxZyqUPNMWVlZYciQIfD29kanTp3QqVMnvPPOO6hZsyYAICsrC/Pnz8emTZtw9+5dpKenIyMjAx06dNA4zov9TIUQqFatGho2bCitMzY2RpUqVXD//n2N/Tw8PKT/NzAwQKtWrRAXF6c11jNnzuD333/X2p/76tWraNmyJUJDQzF8+HCsWbMGnTp1Qt++fVGvXr0Cr9/Kyir/D+gVpk+fjsTERLz11lsgItja2iI4OBgLFizQ2r98xYoVsLe3h4+PT55tFhYWiImJQUpKCvbv34/Q0FDUqlULnTp10jk+VvI4Z4qXMy/atm0bwsLCsGnTJjg5OZXIMV8UGhoq/X+jRo1QqVIlDBgwAJ9//jmqVq1a4udjhcM5VHI5pKvcJwJ+fn5SnjRp0gSnT5/GokWL4OPjg8jISMTGxvKobGUQ51Dxcig7OxtCCGzYsAGVK1cGACxatAje3t5ISkqCra0tvv76a4SEhMDd3R1CCNSuXRtDhw7FqlWrpONERkbi3XffRY0aNWBoaIhmzZrB398fZ86c0XreFStWwM/PDzY2NjrHri8qVDc/AFi9ejVOnjyJdu3aYefOnahbty727dsHQN1FJiIiAmFhYdi/fz9iYmLQq1evPF1ljI2NNX4WQmhdl/sHXBfZ2dnw8fFBTEyMxvL333+jR48eAIBZs2YhLi4OvXr1wh9//IFGjRpp/OK/rLiPilUqFVatWoW0tDTcuHEDt27dQq1atWBhYZEnWTIyMrBmzRoMHToURkZ52+wGBgaoU6cOmjRpgokTJ6Jfv36YN2+ejp8WK02cM8XvorR161YEBgZi7dq18PX11dhmZ2eHpKQkjXVJSUmws7OTtueuy6+MNrkv61+5cuWV8bHSxTlUut38XpVD1tbWMDIygru7u0aZ+vXr49atWwCA/fv3Iy4uDubm5jAyMpLqrQEDBqBt27bFio8VH+eQ7jlkb28PBwcHqSEFqH/3AUi//zY2NtixYwdSU1Nx8+ZNXLp0Cebm5nBxcZH2qV27Ng4fPoyUlBTcvn1besXjxTK5YmJicPr06YrRxQ+oWE+mcjVu3BiNGzfGlClT0K1bN6xZswbe3t44evQofH19ERgYCEDdBTI+Ph6WlpYlct4TJ07A09NTOvapU6fwzjvvaC3brFkzbN68GU5OTnmS/UWurq5S/9eRI0di5cqVePfdd7WWHTFixCtHVHFwcHjldRgbG6NGjRoAgI0bN6JHjx55nkzt2LEDDx48wLBhw155PED9Byg9Pb1QZdnrxzmTv1flzObNmxEcHIw1a9Zojd3DwwNRUVEICwuT1kVFReGtt94CADg7O8POzg5RUVFo0aIFAPU7AkeOHMHChQvzPW9MTAwAdUXK5Mc5lL/C1DsFeVUOmZiYoEWLFrh8+bLGfvHx8dJT4rlz52LSpEka2xs2bIjw8HD4+fkVKz5WMjiH8ldQDrVp0wZbtmxBSkqK9MQsPj4eAPL0klAqlXBwcEBmZia2bdum9bxmZmYwMzPDv//+i3379mHBggV5yixfvhzOzs7o3LlzgXGXFxWqMXX9+nUsW7YMPXv2hIODA65du4Y///wTI0eOBAC4ublh06ZNOHr0KKytrfHtt9/i+vXraNq0aYmcf+nSpXBzc0PDhg2xZMkS3Lx5Uzr3y0aPHo0VK1ZgwIABmDJlCmxsbHDt2jVs3rwZERERMDIywqRJk9CvXz/UqlULSUlJOHr0aL5DxwLFf1QcHx+PkydPonXr1vj333/xxRdf4Pz581rniFq+fDk6deqk9Y7F3Llz0apVK7i4uCA9PR179uxBZGQkvv32W51jY6WDc6Z4ObNx40YEBgYiPDwc7dq1Q2JiIgD1l7vc444fPx7t2rXD/Pnz0atXL/z44484ePCg9FKvEAITJkzAvHnzUK9ePbi5uUlDOg8aNAiA+gX8EydOoGPHjqhcuTKio6PxwQcfoGfPnlJXGCYPzqHid/PLvTHw5MkTGBgYICYmBiYmJtKTplflEKAeRKB///54++234enpiYMHD2Ljxo3YsWMHAPWXUW1fSB0dHbXWY+z14RwqXg4NGjQIn376KYYOHYpZs2bh0aNHGD9+PN555x1poJWTJ0/i7t27aNKkCe7evYtZs2YhOzsbkydPlo6zb98+ZGdno169erhy5QrCwsJQr149DB06VON8aWlpWL9+PSZPngwhhM5x6xW5X9oq7FISLzEmJiZS7969qXr16mRiYkKOjo4UFhZGGRkZRKR+SbB3797ScJZhYWE0cuRIat++vXSM9u3b0+jRozWO26BBA5o5c6bGOltbW/r222+J6P9fUFy3bh15eHiQQqEgNzc32rNnj1Re2/Ca8fHx1LdvX7K0tCSlUklubm40ZswYSk9Pp/T0dPL39ycnJycyMTEhe3t7CgkJocePHxf3Y8pXXFwcNWnShFQqFVWqVIn8/Pzo0qVLecpdvXqVhBC0adMmrceZOnUq1alTh5RKJVWpUoU8PDxow4YNJRprWXxB8XUuJTUABedM8bRv354A5Fle/HyIiLZs2UJ169YlY2NjqlevHm3btk1je3Z2Ns2cOZPs7OxIoVBQu3btpGGBiYjOnDlDrVq1osqVK5NSqaS6devSzJkzKTU1tUSvh/Oq6HnFOVR82nLIyclJo8yrcoiIaPXq1eTq6kpKpZIaNmz4ynoHPABFmaiLOIeK79KlS9SlSxdSqVRUvXp1GjVqFD158kTafujQIapfvz4pFAqqWrUqBQYG0t27dzWOsWnTJnJxcSETExOys7Oj0aNH06NHj/Kca9WqVWRoaJhn/5JSFnNKENFrbr7pxtHR8d7t27f1sr/KjRs34OzsjOjoaDRv3lzucCoER0fHhNu3b1eXOw656HO+AJwzZRXnlf7kFedQ2cA5oz858zLOobKpLOZUhRuAgjHGGGOMMcZKAjemGGOMMcYYY0wHFWoACrnUqlUL+tKdkrGygHOGseLhHGKseDiHWGHxkynGGGOMMcYY00GFb0x16NABY8aMkTuMV5o1axaEEBBCYP78+XKHI7shQ4ZIn8fWrVvlDqfC4Hwpvzp06CB9ZqdPn5Y7nAqF80o/vfh5hIeHyx1OhcI5U37pY15V+MaUPqlbty4SEhIwduxYaV1KSgrGjh2LGjVqQKVSoW7duvjyyy+17k9E6Natm9YGyL///ovAwEBUrlwZlStXRmBgIB49elSk+J49e4YhQ4agUaNGMDY2RocOHfKUOXz4MN566y1UrVoVKpUK9erVy5MsW7ZsQfPmzWFpaQkzMzM0adIkz1xWX3/9NRISEooUH6tYtOXL9u3b4e3tDRsbGwghcOjQoTz7hYSEoHbt2lCpVLCxsYGfnx8uXryo9RzPnj1D48aNtTZAbt26BV9fX5iZmcHa2hrjxo1DRkZGka8jIyMDM2bMgLOzMxQKBWrWrIlvvvlGo8yTJ08wbtw4VK9eHQqFAnXq1MHmzZul7S9WTrmLnZ2dxjG2b9+OU6dOFTk+VrHoUg8lJydj7NixqFevHlQqFRwdHTFy5Eg8fPhQ49jx8fHo1asXrK2tYWFhgdatW2Pv3r1Fii82Nhb+/v5wdHSUYlmwYAGys7OlMnFxcejYsSNsbW2hVCrh4uKCjz76KE9+vir3Jk2ahISEBGkSe8a00ZYzgPr3vU+fPrC0tISpqSmaNWumUdcsX74cHTt2hKWlJYQQuHHjhsb+N27cwLBhw+Di4gKVSgUXFxd8+OGH+O+//6Qy//zzD7y9vaW6wdHREaNHj8bjx4+LfB379u2Dh4cHLCwsYG1tDT8/P2nyX0DzJveLi5mZmdbjHT16FEZGRnjjjTc01utjXvE7U3rEyMgozxeg0NBQ/Pbbb4iMjISzszN+//13hISEwNraWpoNPFdERAQMDLS3nwcNGoRbt25JFdfw4cMRGBiIXbt2FTq+rKwsKJVKjBkzBnv27NHaGDM3N8e4cePQsGFDmJqa4tixY3j//fdhamqKUaNGAQCqVq2KadOmoV69ejA2Nsbu3bsxbNgw2NjYoHv37gAgNfoYy4+2fElNTcVbb72FwYMHIygoSOt+zZs3R1BQEBwdHZGcnIxZs2ahc+fOuHHjRp4Z7SdNmoQaNWrgzz//1FiflZUFHx8fVK1aFUeOHMHDhw8RHBwMIiry5NQDBw7EnTt3sHz5cri6uiIpKUmjsszMzESXLl1gZWWFzZs3o0aNGrhz5w4UCoXGcerWravReDQ0NNTYbmVlhSdPnhQpNlbx6FIP3bt3D3fv3sWCBQvg7u6Ou3fvYtSoUfD398evv/4qHadHjx5wcXHB/v37YWZmhu+++w5+fn6Ii4tD7dq1CxXfmTNnYGNjg8jISNSsWROnTp1CSEgInj9/jo8++giAetLs4OBgNG3aFJaWloiNjZXKLFiwQDrWq3LP3Nwc5ubmeXKJsRdpy5nr16+jTZs2CAoKwoEDB2BpaYlLly7B3NxcKpOWlgYvLy/4+fnhgw8+yHPcS5cuISsrC0uXLoWrqysuXryI9957Dw8fPsTy5csBAAYGBujduzfmzZsHa2trXLlyBaNHj0ZISIjGDbdXuX79Ovz8/DBu3DhERkYiJSUFkydPRvfu3XHlyhUA6pvcLz99a9OmDdq1a5fneP/++y+CgoLQqVMn3L17V2ObXuaV3BNdFXZ5eeK3ZcuWUbVq1ej58+cak3n5+/uTr68vERFduXKFevbsSba2tmRqakpNmzalXbt2aZR/eSI3JycnWrhwYYFl0tPTafLkyeTg4EAqlYqaN29Oe/fupdI0c+ZMatCgQZ71DRo0oBkzZmisa9euXZ7J6U6dOkU1atSgpKSkPBMRxsXFEQA6evSotO7IkSMEQOukvIUxevToPBOT5qd37940cODAAss0bdqUpk6dmmf9y9eSqyxO6vY6F84X7fmS659//iEAdPDgwVceKzY2Vmsu7Nixg9zd3aX8eXHSxj179pAQgm7duiWti4yMJIVCUaTJGfft20eVKlWif/75J98yy5YtI2dnZ0pPT8+3zKs+j1zaJqB8EecV51Vx6qEX/fzzzySEkPIhNycPHDgglcnMzCQDA4NiT5wbFhZGzZo1K7DMBx98QK1bt5Z+Lkzu5dL275WLc4ZzRlvO+Pv706BBgwp1jOjoaAJA169ff2XZxYsXk5WVVYFlvv76a7KzsyvUuXNt2bKFDAwMNP7dDhw4QADyzZGjR48SADp27Fiebb1796ZZs2YVWDfll1dlMaf0tptfv3798PjxY0RFRUnrUlJS8NNPP2Hw4MHSz926dUNUVBRiY2PRt29f9OnTB5cuXSrWuYcOHYrDhw9jw4YNOH/+PIKDg+Hr64vY2Nh895k3b57U2s5vOXLkSJFjadu2LXbt2oXbt28DAP744w/ExMSga9euUpmnT59i0KBBWL58OapVq5bnGMePH4e5uTneeustaV2bNm1gZmaGP/74o8gxFcW5c+fwxx9/oH379lq3ExH279+Py5cva727wQqH80U3qampWL16NWrWrIlatWpJ6+/cuYORI0diw4YNUKlUefY7fvw46tevD0dHR2mdt7c30tPTcebMmUKff8eOHWjRogW++OIL1KhRA66urhg3bhxSUlI0yrRp0wZjx46FnZ0d3N3dMWvWLGRmZmoc69q1a6hevTqcnZ0xcOBAXLt2rQifBNOG80qtMPXQy548eQKFQgFTU1MA6h4J9evXl+56Z2VlYfny5bCwsECbNm2KHNPL56pSpUq+269cuYK9e/dq1EOFyT1WdJwzQHZ2Nnbt2gV3d3d07doVNjY2aNGiBTZt2lSs6wNe/bt+7949bN++Pd/vXPlp0aIFjI2NsXLlSmRlZeHp06dYs2YNWrRoAWtra637rFixAg0aNND4bgkAS5YsQVJSEqZNm1akGMo0uVtzhV1evruR27IdPHiw9HNkZCRVqlSJ/vvvvzwt2VytWrWiTz/9VPq5qHc3rly5QkIIunnzpkYZPz8/GjlyZL7nffjwIf39998FLmlpafnun1/rPT09nYYMGUIAyMjIiIyMjGjp0qUaZQYNGkRjxoyRfsZLT3Pmzp1Lzs7OeY7t7OxM8+bNyzemgrzqyZSDgwOZmJiQgYEBzZ49O8/2R48ekZmZGRkZGZFCoaCVK1dqPc7L15KrLN65eJ0L50vxnkwtXryYzMzMCADVrVuX/v77b2nb8+fPqV27dhQeHk5E2p/mhISEUMeOHTWOmZ2dTYaGhrRhw4Z843qZt7c3KRQK6t69O504cYL27t1Lrq6u1LdvX6lM3bp1SaFQ0NChQ+n06dO0detWsrW1pYkTJ0pl9uzZQ5s2baLY2FiKioqi9u3bk62tLT148EDjfPxkivOqtOqhF/37779Up04dGjt2rMb6O3fuUIsWLUgIQYaGhmRjY0N//PFHvscpjDNnzpBCoaCtW7fm2ebh4UEKhYIAUEhICGVlZUnbCpN7ufjJFOdMfrTlTEJCAgEgU1NTioiIoHPnzlFERAQZGhrS7t278xyjsE+mbty4QVWrVqWIiIg82wYOHEgqlYoAUI8ePQqMOT+///472drakqGhIQkhqFmzZpSUlKS17KNHj0ilUtFXX32lsf7PP/+katWq0bVr14io4Lpan55M6fU7U4MHD0ZwcDDS0tJgamqK9evXo2/fvlAqlQDUd5Vnz56N3bt3IyEhAZmZmXj27BkaNWqk8znPnj0LIoK7u7vG+vT0dHh6eua7n5WVFaysrHQ+b36+/fZb/PHHH9i5cyecnJzw+++/Y9KkSahVqxa6du2KyMhIxMbGlrnRuY4cOYKUlBScOHECU6ZMgbOzs8Y7XhYWFoiJiUFKSgr279+P0NBQ1KpVC506dZIxav3G+VJ4AQEB6NKlCxISEhAeHo5+/frh2LFjMDU1xbx582BiYoLQ0NBSjyM7OxtCCGzYsEF6R3DRokXw9vZGUlISbG1tkZ2djWrVqmHFihUwNDTEm2++iYcPH+KDDz7AwoULIYRAt27dNI7bunVruLi4YM2aNa/lOsozzqtX10MvSklJga+vLxwcHDTeTyIijBo1SnrPUKVSYeXKlejbty+io6Ph4OBQ5LguX74MHx8fTJgwAX379s2zfdOmTXj69CliY2MRFhaGzz//HB9++CGAwuUe001Fz5ncwVD8/Pykv79NmjTB6dOnsWjRIvj4+BT5mElJSejatSu6dOmi9f2qL7/8EjNnzkR8fDw+/PBDTJgwAcuWLSv08RMTEzFs2DAEBQXB398fT58+xYwZM9C/f38cOHAgz/v469atQ3Z2tsb3uvT0dAwYMADh4eFwdnYu8jWWZXrdmPLx8YGRkRF++ukndOrUCb/99hv27dsnbZ80aRL27t2L8PBwuLq6wtTUFEFBQQWOqGVgYAAizUnaXuwuk/sHNjo6Os/L6Nq6++SaN28e5s2bV+D1/PLLL3j77bcLLPOi//77Dx9++CG2bNkCX19fAECjRo0QExOD8PBwdO3aFfv370dcXJzGS40AMGDAAHh4eODo0aOws7PDP//8AyKCEAKAumK7f/9+npcmS0puIjVs2BBJSUmYNWuWRtIZGBigTp06ANR/ZC5evIh58+ZxY6oYKnq+FEXuACeurq5o3bo1qlSpgm3btiEwMBD79+/HkSNH8lxP69atMWDAAKxfvx52dnY4duyYxvYHDx4gKyurSDllb28PBwcHjcFW6tevD0A9WqCtrS3s7e1hbGys8bJu/fr1kZaWhgcPHsDGxibPcc3NzdGgQQP8/fffhY6FaVfR86ow9VCulJQUaRCh3bt3S1+eAeDAgQPYtWsXkpOTYWlpCUDdHSgqKgqrV68ucpegS5cuoWPHjhg4cGC+Q1LndsN1d3dHVlYWhg8fjrCwMBgZGRUq95huKnrOWFtbw8jIKE/Drn79+ti4cWOhj5MrMTERnp6eeOONNxAZGSl9j3uRnZ0d7OzsUK9ePVhZWeHtt9/GtGnTNLqiF2Tx4sUwMzPTuAGybt06ODo64o8//kDbtm01yq9YsQJ9+/bVaIgmJCTg4sWLGDp0KIYOHQpA/e9CRDAyMsKePXvg5eVV5OsvC/S6MaVQKNCvXz+sX78eDx48gJ2dncZw3EePHkVQUJB0R+rZs2e4evUq3Nzc8j2mjY2NxpDbz549w6VLl9C0aVMAQNOmTUFESExMRMeOHQsd64gRI9C/f/8CyxT1zltmZiYyMzPzjHhiaGgo3fmYO3cuJk2apLG9YcOGCA8Ph5+fHwDAw8MDKSkpOH78uNS39fjx49LIZ6UtOzsb6enpxS7DClbR80VXuY/xc3//Vq9ejdTUVGn7vXv34O3tjfXr10vvdnh4eGDOnDm4c+eONLxrVFQUFAoF3nzzzUKfu02bNtiyZQtSUlKkGyK5Q9E6OTlJZTZs2IDs7Gzp7mB8fDxMTU3z7cue++9UlH8Tpl1Fz6vC1EOA+t3dbt26gYiwd+/ePDf40tLSACDPHW4DAwON4xRGXFwcPD090b9//3ynCnlZdnY2nj9/jqysLBgZGRUq95huKnrOmJiYoEWLFrh8+bLG+vj4+CL/biUkJKBjx45o0KABfvjhBxgZvfprfW4+FeU7VVpamtYcf/F4uU6dOoXY2Fh89dVXGusdHBzw119/aazLvWHy448/aryXrG/0ujEFqB8Xd+rUCdevX4e/v7/GH2I3Nzf8+OOP8PPzg7GxMWbPno1nz54VeDxPT0+sWrUKPXv2hI2NDebOnYvnz59rHDMgIABDhgxBREQEmjVrhuTkZBw6dAguLi7o06eP1uOWxqPiSpUqoX379pg6dSrMzc3h5OSEw4cPY+3atdLdAwcHB62J7ujoCBcXFwDquyFdu3bF+++/Lw2n+f7776NHjx6oW7dukWKKi4tDRkYGHjx4gJSUFMTExABQP10C1N1BnJ2dpeP+/vvvCA8Pl4ZFB9QNwFatWsHFxQXp6enYs2cPIiMjizykNMurIucLoJ7v5tatW9Kw/VeuXIGlpaV01+7KlSvYtm0bOnfuDBsbG9y5cwfz58+HQqFAjx49ACBP94TcL1q1a9eWGk5eXl5o0KABgoKCEBERgYcPHyIsLAwhISGoVKlSoeMdNGgQPv30UwwdOhSzZs3Co0ePMH78eLzzzjvSYDIjR47EokWLMH78eIwZMwY3btzAzJkzMWrUKOkO5aRJk+Dr64uaNWvi/v37+PTTT5Gamorg4OBifZ5MrSLnVWHqoadPn8LLywtPnjzBjh07kJqaKt2QsLKygomJCTw8PGBlZYWhQ4dixowZUKlUWLFiBa5duyblXmFcuHABnp6e6NixIz766CMkJiZK23KfCkdGRkKpVKJhw4YwMTHB6dOn8eGHH+Kdd96RphQoTO4x3VXknAGAyZMno3///nj77bfh6emJgwcPYuPGjdixY4dUJjExEYmJiVIjPi4uDo8ePULNmjVhZWWFe/fuoUOHDqhevTq++uorPHjwQNrXxsYGhoaG2L17Nx4+fIg333wT5ubmuHDhAsLCwtC6dWup909h+Pj44Msvv8Qnn3widfP76KOP4OjomOcGYe5UAi/PNWpsbJxnTqlq1apBoVDkWa935H5pq7CLtpcYidQvdTs5OREAio2N1dh248YN6tSpE5mampKDgwMtXLiQfHx8KDg4WCrz8kuMjx8/poEDB1KlSpWoevXqtHjx4jxlMjIyaObMmeTs7EzGxsZka2tLvr6+dPr0aW0hloj8XtJLSEigIUOGUPXq1UmpVFLdunVp4cKFlJ2dne+xoGXQhuTkZAoICCALCwuysLCggIAA+vfff/PsN3PmzALjzP23eHnJ9eWXX5K7uzuZmppSpUqVqGnTprR48WKNF3+nTp1KderUIaVSSVWqVCEPD498X9rXdi1EZfMFxde5cL5oz5fVq1dr/f3M/b2+desWde3alWxsbMjY2Jhq1KhBgwYNoosXL+Z7rvwGbbh58yb5+PiQSqUiKysrGjt2LD179izPfqtXry7wWi5dukRdunQhlUpF1atXp1GjRtGTJ080yhw/fpw8PDxIqVRSrVq1aPr06RpDpQ8YMIDs7e3J2NiYqlevTn369KELFy4U+lpycV5xXulSDx08eFBr3uGlQWCio6PJy8uLrKysyMLCglq2bJnnhXwnJyeNz05bjPmdK9eGDRuoadOmZG5uTmZmZuTu7k5z587N81J+YXIvNyYegIJzRpuCBlhYvXo1ubq6klKppIYNG+b5npPf73JunZFffYYXBquIioqi1q1bU+XKlUmpVJKrqytNnjyZkpOTNc5VmO93P/zwAzVr1ozMzMzI2tqaevTokaceefLkCZmZmdHnn39e7M9HnwagkD2Awi75JWRFUdh5YkrLtWvXSAihMRdVWcCNKc4XbeTOl8I4cOAAKZVKunr1qtyhSLgxxXlVELnzKjU1lZRKZZFGxHwduDHFOZMfuXOmMMrq9zt9akzp7TxTFdHFixdhbm6OL7744rWfe8+ePQgKCir2fB8lZcSIEXn63DP2IjnzpTD27NmDKVOmSN1t5datWzc0aNBA7jBYGSdnXh08eBCtWrWCv7//az+3NrlzEN26dUvuUFgZpg91UVn6fqePeSWI6NWlygBHR8d7t2/ftpc7DrkkJycjOTkZgHokmNzRjiqq+/fv48mTJwDUI56ZmZlpbHd0dEy4fft2dTliKws4Xzhfiuru3bv477//AKjfqcx9d+RFnFecV5xX/68wnwfnDOcM50zRvOozK4s5pfcDUFQUcs+7U9ZUq1aNXwJm+eJ8KbrXNToi01+cV5r482Cvwr8jRaePnxl382OMMcYYY4wxHXBjijHGGGOMMcZ0wI0pxhhjjDHGGNOB3rwzlZGRccfR0VHuMJieyMjIuCN3DHLifGGlgfOK84oVDecM5wwrWWUxp/RmND/GGGOMMcYYK0u4mx9jjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmA25MMcYYY4wxxpgOuDHFGGOMMcYYYzrgxhRjjDHGGGOM6YAbU4wxxhhjjDGmg/8D44BfzgHD1sQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualizing tree\n",
    "_, ax = plt.subplots(figsize=(15,10)) # Resize figure\n",
    "tree.plot_tree(dt_classifier.best_estimator_, feature_names=X_train.columns, fontsize=14)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We will now try to see if this can be improved upon by using [Cost Complexity Pruning](https://scikit-learn.org/stable/auto_examples/tree/plot_cost_complexity_pruning.html#sphx-glr-auto-examples-tree-plot-cost-complexity-pruning-py) using original Decision Tree which had overfit - \n",
    "\n",
    "Basic Concept - \n",
    "- The `cost_complexity_pruning_path()` function of the `DecisionTreeClassifier()` provides a list of Alpha values, `ccp_alpha`, for each pruned tree at the weakest link (one that least minimizes or does not minimize the misclassificaiton rate from parent to leaf node).\n",
    "- When &#945; $= 0$, the tree is unpruned and as &#945; &#8594; &#8734;, the tree size $T$ &#8594; $1$. \n",
    "- We then plot the traning v/s testing set accuracies for each `ccp_alpha` and see how the accuracy changes with each value. We can then use cross validation around a set of values where training and testing accuracy are good as well as close enough to check which one gives us the best performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf = tree.DecisionTreeClassifier(random_state = 42)\n",
    "path = clf.cost_complexity_pruning_path(X_train, y_train)\n",
    "ccp_alphas, impurities = path.ccp_alphas, path.impurities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Total Impurity vs effective alpha for training set')"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize = (8,8))\n",
    "ax.plot(ccp_alphas[:-1], impurities[:-1], marker=\"o\", drawstyle=\"steps-post\")\n",
    "ax.set_xlabel(\"effective alpha\")\n",
    "ax.set_ylabel(\"total impurity of leaves\")\n",
    "ax.set_title(\"Total Impurity vs effective alpha for training set\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of nodes in the last tree is: 1 with ccp_alpha: 0.05294659854356043\n"
     ]
    }
   ],
   "source": [
    "clfs = []\n",
    "ccp_alphas = ccp_alphas[ccp_alphas>=0]\n",
    "for ccp_alpha in ccp_alphas:\n",
    "    clf = tree.DecisionTreeClassifier(random_state=0, ccp_alpha=ccp_alpha)\n",
    "    clf.fit(X_train, y_train)\n",
    "    clfs.append(clf)\n",
    "print(\n",
    "    \"Number of nodes in the last tree is: {} with ccp_alpha: {}\".format(\n",
    "        clfs[-1].tree_.node_count, ccp_alphas[-1]\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZUAAAEgCAYAAAB8XO79AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAviUlEQVR4nO3deZgcVb3/8fdnJglkD2ExhJDFQMCAEDZBDBH0gggIuFwUAeXCRXADryyCCAxcRIkKeuGHcQ8EEFBE8bIoqGhQLpDIIsoWIAkQkhBCEkKAbN/fH6cmNJ3unu6Z6p7JzOf1PPXM9KlTp86p7q5v16lTVYoIzMzM8tDU2RUwM7Puw0HFzMxy46BiZma5cVAxM7PcOKiYmVluHFTMzCw3DirWZUgKSUfXe5kayv53SU9JWiNpaj3WkSdJd0n6cY3LtEiaVa86NUpXa4ek0dlnc2Jn16XRHFQASVtJekPSPEm9Ors+1vkkNQM/BW4ARgKn5Fz+iGyns2+OxX4E+HKNy3wb2CvHOnRZklZLOrYO5c6S1FKU/CywJXBv3uurB0l35vXDyUElOR74X2AJ8KHOrUoiqXdn16GH2xIYANwaEc9HxNL2FNLR91FSUxbg2hQRiyNiWS3lR8TyiFjUvtpZORGxJiLmR8Sqzq5Lw0VEj55IgXUOKZh8BbitRJ4tgJ8BC4DXgceB4wrmjwV+CSwGVgAPA4dk844FVheVNwIIYN/s9b7Z64OBu7N1fBbYBLgamAu8lq33VEBF5X0cmJkt9xJwW7bssaRA2a8o/7nAk8XlZPP2B9YAI0qsYwUwKHv9VeBp4A3gReB3QN8K2/mTpF9tS4FFwC3AuKI8ARxd9PoU4EbgVeB54JQSy3wOmAa8AjwHnFXruovyH5uVWzi1vlcHZdv6DWAhcAXQv2DZqcCdwBeB2cDaUtulRPmzs/QWYFa2vR8DVgPvAHbN3teFwHLgfuDAojLvAn5c/Bo4B5hP+nxeBQwoyNMCzCp+DRyWrf/VrJxti9Z1JPAU6TP3N+CQrB0TK2zXatowG7gA+F5W3wXApUCvgjwbA9/P3s+Xs/+/UdiOEuueXbzNC+btBvw+q9OLwK+AUUXf1xuzz87rpM/96QXbuPi9HJ1N67ZHwesjSD9gV2TlHFtUzzFZXV4nHe18vvh9LdG23sAlpM/+G8ALwHVFeT4BPJiVOzvL37/gM1vu817T9zwiHFRIO/L5QC9gOLASGF0wvy/wKPB34N+AtwMHAJ/I5g/LPvh3AhNJAeYw4KBs/rFUH1QeIwW3MVmeYcCZpC/jGODo7IP/HwVl/QewirTjGA/sRNoRb5bV/WXg0wX5m7IP1VfKbI+m7MP5laL0W4Frs/8/AizL6joSmAB8qdKHLavnh7LtswtwMymw9SnIUyqoLCbtoMdl7VoNHFaUZwFwQlb257O099ey7qK69gX2yMo5NHsf+mTbdjVpJ7c98EFSwJ9WsOzUbNvcBOwMvBNoLrGOXbLyP5KVv3mW3kLa4fwZ2DNr98DsM3IssEOWdiHpszquoMy7WD+oLCmo7wHZ9vzvgjwtrB9UXgVuJ+1sdyYF0ekFeXYjBcsLge2Aw0mBqK2gUk0bZpM+s2cC25J2wquA4wvyXEoKTIdl7fp2ts0rBZXNs/fulGx7D8vSx5O+U+dnZb0T+AXwBLBxludm0vd7Aik47Accmc0bCjyT1WFYNjVTPqg8nbVpG+CirE7jsjwi7fjvBd6Vre9WUvCsFFS+TPrO7kv6Pu4BfKlg/rHZNj2GtP+aRPrhOy2bPxj4C3B9QRv60I7veYSDCsBvgO8UvL4duLDg9fGk6D6izPL/TQpK/cvMP5bqg8oxVdT3e8AdBa/nApdXyP8/wN0Frz9A+iJvUWGZbwKPFLx+W/bh/0D2+r+yL13vDmz3oVmb31OQViqoTCta7lreuoML4H+K8jwKfKOWdZfI07oTmFiQNg24ryjfYaQd7Kjs9VTSjnxAubJLfQYK0luy8kZWsQ0fAs4ueH0X6weVh4qW+T5wT9H6ioPKarIgl6V9PKtT6072msL3IEs7qXh7Vfk5KG7DbODmojy3AT/P/u9P+j6eUJRnBhWCSpZnNesfGUxl/V/1G5EC++EFdWypUO6s4vnFn5+C118uyNNMOro+MXu9f5Znm6LP6goqB5XvAX+kRM9DwTY9qShtUrauTbLXdwJTi/K063veo8+pSNqKdKQytSD5SuC4ghP2uwH/iojnyhSzG/C3iHg1hyrdV1S/JklnSnpQ0iJJy0lf3lHZ/C2ArUmHy+X8AHiPpHdkr08gfWkXVljmSmAHSbtmr48i/TK8M3t9A+mQe46kqZKOkTSwUsMkTZB0k6RnJL1CCoa0tqWCe4pe/5X0S7fQg0Wv55ECYUfXXWwH0i+6Qn8m/cIcX5D2aEQsr7HsQgsiYm5hgqTNJV0h6TFJS7LPwg603YaHil6/ZduUMS8iXixaRqRuYEht/b+iZYrfp/XU0IYHK9R5LGmn/7eiPHe3tf4y9gA+LGl560TqQt6YdKQE8F3gq5LulXSxpEntXBcUtC0i1pC+V61tGw8siohZBXkWk7q9K/kZ6QhrlqQpkj4qqQ+kbU7avpcUtfG2bNltKpRb8/ccfKL+eNKvhQeykSGrSb9GtyS/E/ZrS6SVO3lbHJhOBc4iHW3sTzr8/DHp0LQqEfFP0hfuhCwIHQr8sI1lHiX98vtUlvQp4OrsS0BEPE/qKjiO9KU4B3hc0talypPUjxT4gtQV9S7e7F6qui0VrCxuAtlnuwHrLqWjPzBKLT8V2Ac4I/s7gbSDaqsNZbdNjctQtFxQu6lU14b21Lm9mkjf+QlF0zjSd42I+BlpxzyFtG+4TdLV7VxfW22rebtGxIOk7vHTsvK/BzwoaVBB2afw1vbtTAqa/6hQbk3f81Y9NqhIaiIFlYtY/wP1c+AzWdaZwHhJI8oUNRPYW1L/MvMXAs2SCn8d7lomb7FJwO0R8dOIeCD7BdP664nsaOM5Ul95JT8gBYbPkE5231HFuq8EjsyOVnYmneBdJyLeiIjbI+IM0q+kfqS+9VLeQerTPjsi7sqC1iakX79tKR7uujfwryqWy2Pdxf5Jek8KvZe0I/hnjWW17lyqGtmVrfeKiLg5Iv5BOhn79hrXmZd/Ae8uSqtmWHIebXiKtO32Lkp/TxXLrmT97T2DdK7sqYiYVTS93JopIl6IiJ9FxKdI+42jsp12uXLb41/A5pLGtiZI2oQU4CqKNIrvpog4Gdid9Ll/b0QsIJ3w365E+2ZFxOuV2lDj9xzowUGFdJJ1a+AHEfFI4UT6RXWApNGkADMHuFnSv0kaI+n9kj6elXMFaTv+RtJ7svmHSPpgNv8+Ur/pNyVtK+lA0uirajwO7CtpP0njJF1IOnlb6HzgREnnSHqHpB0kfUHSZgV5fpn9PYfUN1vNr6Gfk3a+PwH+nm0XACQdL+kESTtLGkXqHhtI+Z39HNLokS9KGivp/aRfU9XU45CsPdtK+iKpf/87VSyXx7qLfQvYVdKlkrbP3svLgGuKu6uqsIh0gvgAScOynUclj5N2ZO+UNIH0/uSxI2uPS0hdqhdkn8tDSUfVUHm7drgNWTfzFOBCSYdK2k7SZNKAgbY8A+wnaXjB9+Mi0g74aknvyr6/+0n6nqS3A0i6XNJB2ednB9IJ7GdJ3+vWct8jaaSkzbIfrO1xJ6m7cpqkPSTtTDqKWk2F7SrpdElHZd/9MaQjizWk8yEAZwMnSzpb0o7ZNjtc0g+Kts1uWRs3k9S7Hd9zoGcHlc8A95bZGfyRNErmPyNiBenX6CPAdaSTwP+PNEKIiHiBNOrrFdJIjX8CXyf7JZz1iR5J+iX3MGnHfkaVdfxvUp/9b0h91puQusLWiYgfkwYDfIzUlfAXUsBcXZDnddKHs4l0QV+bIuIl0tDbCRQdpZBGkvwH6UTwo6TRJ5+JiD+UKWsRaeTa/qTt823SoXqprsFiF5BG3T1EGt54RkTcVE0bclh3cVkPk7oPJ2X1mUbaRie1o6y1pJFqR5CONh9oY5H/IL1/9wG/Jg0oub/W9eYhImaSdjBHkbpPzgK+ls1+vdxy5NeGM7Plp2VlDSF9J9tyKukc6GzS8NjWrt69Sdck/Y60w/wR6fu9JFtOpPMqj5C+X/2BDxb8ODsvq8PjWbkj29EmsvI+TOr+nE4aenxbVm6l7bqM9B28h/R+fBj4aEQ8npU7jfQ5O4S0ve4nDch4vqCM75B+6DyUteE91Pg9b6XqfrTahk7SDaRRHB/u7LpUS1LriLj29l9bg0j6FOmE8aYRsaSTq9NtZCfGnwO+FhGXdXZ9quFbknRzWbfKu0i/Xt7fydWxbkLSacCfSEf0ewAXA79wQOmYrCtxNenIYAvSUVCQRmJtEBxUur8HgE2ByRFRPBzWrL12InUnDSWdX7iatAO0julHOuc6mtQNNpN0rcuCzqxULdz9ZWZmuenJJ+rNzCxnDipmZpabHn1OJRtdZGZmNYqIkhcQ9+igAuBzSmZmtZHK35DC3V9mZpYbBxUzM8uNg4qZmeXGQcXMzHLjoGJmZrnp8aO/2mP0mbes+3/2Nw/uxJqYWVvWrl3LvHnzWLVqVWdXZYPRu3dvhg8fTlNT7ccdPfo2LZKqfLTImwoDSisHFrOu67nnnmPQoEEMGjSo7cwGwLJly1i2bBkjRpR+NqGkstepuPurBqUCSqV0M+t8q1atckCp0aBBg9p9ZOegYmZmuXFQMTOz3DiomJl1MzvssAPXXHNNp6zbo7/MzLqIfffdl3333ZeWlpYOlfPPf/4znwq1g4OKmVkbZsxezOyXVjBms37sNmpop9Vj5cqV9OnTp9PWXw13f5mZVXDebx7hmJ/cx7m/eYSjf3wf5/3mkbqs56STTmL69OlcdNFFDBgwgGHDhtHS0sJ73/tezjrrLLbccksmTpwIwAknnMDIkSMZMGAA2267LZdffvlbyho9ejRTp04FYPbs2Uji6quvZqeddmLgwIHsvffePPbYY3Vph4OKmfUoa9YGC5a9XtX0+3/O5/oZz/LaqjWsWLmG11at4foZz3LHv+ZXXcaatdVdCzdlyhT22WcfvvrVr7J8+XLmz58PwF//+lcGDhzI7NmzueuuuwDYa6+9mDlzJsuWLeOyyy7j1FNP5Y477qhY/rRp07jjjjt48cUXGTZsGJ///Oc7tB3LcfeXmfUoi5a/wZ4X/aHdy7++ai0nXDWz6vz3fvX9vG3Qxu1e3/DhwznrrLPe8gyT448/ft3/Bx54IAceeCB33nkn+++/f9lyzjvvPN72trcBcNxxx3HkkUe2u06VOKiYWY+y2YCNuPer768q70PPLuHk6x7g9VVr16Vt3LuJy47chZ1GDKl6fR0xevTotwSUiODCCy/k2muvZd68eUhixYoVDB1a+VzP8OHD1/0/YMAAli9f3qF6leOgYmY9SnOTqj5yOGCHYXx89625YcZzSBABH999a/YfP6wudSt1r63itJ///Odcfvnl/P73v+ed73wnTU1NHHbYYV3mKbYOKmZmFZx/2I4cOmE4zyyq/+ivYcOG8cQTT1TMs3TpUnr16sVmm21GRHDTTTdxxx13cMQRR9StXrXwiXozszbsNmooH9ttRN2HE5966qk88sgjDBkypOzNHI899lgmTZrE+PHjGTZsGLfddhuHHXZYXetVC9+luIb2V7px5I2ffXenjl83s9KeeeYZxowZ09nV2OBU2m6+S3EDXPaHJzu7CmZmnc5BJSd/fmJRZ1fBzKzTOajUoNLDuHpuJ6KZ2ZsaGlQkNUm6SNICScsl3S5pVIX8h0l6QNIrkp6WdGLR/KmSVmVltU4X17MNgzZqLpk+eGMPpDMza/SRyhnAkcAkYBgwF/itpPXqIWlP4DrgLGAw8GngEkkfLsp6TUQMKJi+Us8GPHz+gSXTH2r5QD1Xa2a2QWh0UDkJmBwRj0fEclKQ2Q6YWCLvR4A/RMTtEbE2IqYDNwFfaFx11+dHCpuZldewoCJpMDAKmNGaFhFLgFnAhBKLNAHFQ9YE7FKUdriklyQ9JWmKpM1zq7SZmdWkkUcqg7K/S4rSlxTMK3QzsL+kQyT1krQfcDgwsCDPZcD2wGbAAcBo4GYV3ignI6lFUhROHWiLmZmV0Migsiz7O7gofUjBvHWy7q5jga8DC4FzgR8BiwryzIyI+ZE8BfwnsBewbYnyWiJChVPHm2RmZoUaFlQiYikwB9i9NS3rEhsLPFhmmWsjYueIGBoR+wFbA3+qsJrWW4k6YJiZdYJGn6ifApwuaZyk/sDFwBPA3cUZs+HHe0hqljRA0snA/sAF2fyNJX0sC0xkQ5N/BMwEfHm7mW1w8ng+faupU6cyevToXMqqRaODymTgBlIQWQiMAQ6NiLWS9smuMxmZ5W0GrgBeBl4ADgHeGxGtz8BsAk4BnpH0KvAX4DngkIh48+EHOdu0X++S6Zv179rPjTazDpj7f/DgtTD33s6uSZfX0KCSDQ0+KyK2iIj+EfGBiJidzZueXWcyN3u9KiL2iIhBETEwIg6IiAcKyloREftkXWP9I2JURJwYEfPr2YaXVqwqmb7o1ZX1XK2ZdZZbT4dpH4ZbToNph6fXdVDqGfUAV111FTvvvDODBw9mhx124Lrrrlu3zNy5cznooIMYOnQogwcPZscdd2T69OlMnz6dk046iblz5zJgwAAGDBjAr3/967rUu5gvAzeznmXtGli+sLq8z/8d/j4NVr/2ZtoD0+Dt+8LwXasrY8AW0FT6ThyFpkyZwmOPPfaWLrCpU6fS0tLCjTfeyC677MLf/vY3Dj74YEaMGMHEiRM566yz2GqrrZg3bx59+vThySefpE+fPowZM4YpU6bQ0tLC7Nmzq6tnThxUzKxnWb4QLtm+/cuveg2u+2T1+b/8GAzasl2ruuSSSzjnnHPYbbfdAJg4cSKf/OQnmTp1KhMnTqRPnz7Mnz+fp556ivHjx7Pddtu1az15clAxs55lwBZpR1+N5/8ONx7/1iOV3n3hoz+p7UilnZ588klOOeUUTj311HVpq1evZtKkSQB8+9vf5sILL+QjH/kIL7/8Mh/60If45je/yeabd9414A4qZtazNDVXf+Qw6GDY9Rh44GrWPaR+l2Ng+/J3LO9Q1YqeRz9s2DC+/vWv88lPlj4y2nTTTbn00ku59NJLef755zn66KM57bTTuPLKK0s+774RHFTMzCo56Fuw48dg8VMwdCyM3LNuqyp+Rv2XvvQlzj//fMaNG8euu+7KqlWrePjhh2lqamK33XbjuuuuY4899mDMmDEMHDiQjTbaiObm5nVlvfjii7z88stssskmdatzMT9PxcysLSP3hAmfrGtAgfWfUX/KKadw7rnncuKJJzJ06FC22morTj/9dF599VUAHnroId73vvcxcOBAxo4dy5AhQ5g8eTIA73vf+zj44IPZZpttGDJkCDfffHNd697Kz6ivsf2V7kZc6SFeZtY5/Iz69vEz6s3MrNM5qJiZWW4cVMzMLDcOKmZmlhsHFTMzy42Dipl1e2vX1u3G5d1SR7aXg4qZdWtDhgxh4cKF9OTLJ2oRESxcuJAhQ4a0a3lfUW9m3dqQIUOYP38+zzzzDJIfCtuWiKBv374OKmZmpUhiyy3bd5dgq527v8zMLDcOKmZmlhsHFTMzy42DipmZ5cZBxczMcuOgYmZmuXFQMTOz3DQ0qEhqknSRpAWSlku6XdKoCvkPk/SApFckPS3pxKL5fSX9QNJiScskXS9paP1bYmZmpTT6SOUM4EhgEjAMmAv8VtJ69ZC0J3AdcBYwGPg0cImkDxdkuxTYDdgRGAUMAK6qZwPMzKy8RgeVk4DJEfF4RCwnBZntgIkl8n4E+ENE3B4RayNiOnAT8AVIRymkQHNORMyLiJeB04CDJY1sRGPMzOytGhZUJA0mHU3MaE2LiCXALGBCiUWagOIb9QjYJft/HLBxUXmPAivKlGdmZnXWyCOVQdnfJUXpSwrmFboZ2F/SIZJ6SdoPOBwY2J7yJLVIisKp1gaYmVlljQwqy7K/g4vShxTMWyfr7joW+DqwEDgX+BGwqJ3ltUSECqfam2BmZpU0LKhExFJgDrB7a1rWJTYWeLDMMtdGxM4RMTQi9gO2Bv6UzX4CeL2ovO2BfsBD9WiDmZlV1ugT9VOA0yWNk9QfuJgUHO4uzpgNP95DUrOkAZJOBvYHLgCIiNeAK4ELJG0paRPgW8CtETGnUQ0yM7M3NTqoTAZuIAWRhcAY4NCIWCtpn+zaldaRW83AFcDLwAvAIcB7I+KxgvL+i3SU8y/SUdBrwDGNaIiZma1PPfkRm5Ki1vaPPvOWsvNmf/PgjlbJzKzLk0S589K+TYuZmeXGQcXMzHLjoGJmZrlxUDEzs9w4qJiZWW4cVMzMLDcOKmZmlhsHFTMzy42DipmZ5abqoJI9FMvMzKysWo5UXpB0uaSd6lYbMzPboNUSVE4GdgIelHSfpOOzOw2bmZkBNQSViLgqIiYB44G/ABcB8yRNkbRrvSpoZmYbjppP1EfEYxFxGjACOJv0dMb7Jc2UdFTO9TMzsw1Ir/YsJOkDwAnAocDjpMf8jgAul7RfRPxnflU0M7MNRdVBRdJWwHHZ9Dbgl8B+EfHXgjy/AO4Cum1QEVDqCSx+4L2ZWW1HKnNIj/79H+DKiFhcIs/jwH15VKyrKvdIr577qDMzszfVElTeHxF/rpQhIpYB+3WsSmZmtqGq5UT9XEnjihMlbStpdH5VMjOzDVUtQeWnwHtKpO8N/CSf6piZ2YaslqCyC/DXEun3AL5OxczMagoqvYGNSqRvBPTJpzpmZrYhqyWozCRd6FjsOODBPCpjZmYbtlpGf10A3C5pG+COLO0A4IPAQXlXzMzMNjy13PvrTuBAYBPg4mwaAhwUEXdUWHQdSU2SLpK0QNJySbdLGlUh/1GS/iFpmaTnJH1X0kYF86dKWpWV1TpdXG2bzMwsXzXdpiULLHd2YH1nAEcCk4DngUuA30qaEBFrCzNK2hm4Cvg4cCMwErgdeAU4pyDrNRFxbAfqZGZmOWn0kx9PAiZHxOMRsZwUZLYDJpbI+3ZgcUT8MpI5wC3AhIbV1szMalLLkx97SzpP0uOSXpe0pnCqYvnBwChgRmtaRCwBZlE6UPwOeEbSJyQ1SxoLfAj4VVG+wyW9JOmp7Db8m5dZf4ukKJyqariZmVWtliOVFuAzwPdJt7r6Gumix8XAF6tYflD2d0lR+pKCeetExIqs/CnAG6Tgcw+pS6zVZcD2wGakQQOjgZslrXd/x4hoiQgVTlXU2czMalBLUPkEcGJEfBdYDdwQEZ8BLiRdVd+WZdnfwUXpQwrmrSPp08A3SEcnfYDhpOBxZWueiJgZEfOz7rGnSHdH3gvYtvpmmZlZXmoJKlsCD2f/v8qbRxe/BQ5pa+GIWEq60/HurWlZl9hYSl/nsjvwp4iYHhFrI+IF4IfAYRVW03qy30chZmadoJagMg/YIvt/NmkEF8COpCOXakwBTpc0Lnu+/cWk2+nfXSLvdGBfSe9WsgWp+20mgKSNJX0sC0xkQ5N/lM1/soZ21aSpTLgql25m1pPUElT+yJtHCT8BJku6B7gW+EWVZUwGbiAFkYXAGODQiFgraZ/sOpORABFxA+mCy5+Rusf+QTpCOqag7qeQTua/CvwFeA44pHh4cp6izOn9culmZj2Jooa9oSRFtoCkfwf2IT2Y6wcRUe3RSpdR0JyqjT7zlrLzZn/z4I5Wycysy5NEucFOVV38KKk36ahhCum8CBHxC6o/Quk2ejXB6hLHQb0afcWPmVkXVNWuMCJWUd2w4W5vbZmOtXLpZmY9SS2/r/9MdUOHu7VyscMxxcystnt/XQNcLGkMcD/ppPk6EfG3PCtmZmYbnlqCytXZ3wtLzAuguePVMTOzDVktQWVM3WphZmbdQtVBJbtLsJmZWVlVBxVJn6o0PyKuqjTfzMy6v1q6v35S9LqJdI+ttdnkoGJm1sPV8jjh3oUT0BvYA/grsF+9KmhmZhuOdl8Hnt05eCZwNnB5flXq2srdN9L3kzQzy+dxwi8C43IoZ4Owce/Sm6xvH4+oNjOr5UT98OIk0oOzWoB/5Vinrq3c/Sd9m2Izs5pO1D/H+rtUkW4weURuNeriVpW5ydfKNb5Ri5lZLUGl+GT8WtIzUWZFxJr8qtS1NTeJ1WvXPyppbvJtis3Marn48c/1rMiGYnDf3ix8ZeV66UP69u6E2piZdS1V/7yWdKyk9bq5JB3R1oWR3ckm/fqUSXdQMTOrpc/mK8DiEumLgDPzqU7Xt8PwwaXTtyqdbmbWk9QSVEYDs0qkP53N6xFWlTkhv2qNR3+ZmdUSVJZS+k7FY4Hl+VSn65v1YummPrnwlQbXxMys66klqNwGfEvSlq0J2bUrFwO35l2xrmpVqQfUA6t9pGJmVlNQOQPoDzwlaYakGaTusP7ZvB5h0fI3yqS/3uCamJl1PbUMKX5R0i7AUcCuWfIVwM8j4rV6VK4rWvra6pLpS1aUTjcz60lqumIvIl6PiJ9ExOez6ae1BBRJTZIukrRA0nJJt0saVSH/UZL+IWmZpOckfVfSRgXz+0r6gaTFWZ7rJQ2tpU218g0lzczKq+U6lTMlHV8i/XhJ1XZ/nQEcCUwChgFzgd9KWq8eknYmPaPlfGAw8B7gA8DXCrJdCuwG7AiMAgZQ5+e6bDqg9HUq5dLNzHqSWo5UPgM8XiL9UeDEKss4CZgcEY9HxHJSkNkOmFgi79uBxRHxy0jmALcAEyAdpQCfBs6JiHkR8TJwGnCwpJE1tKsm47ccWDJ9h+GD6rVKM7MNRi1BZTjpppLF5gFbtbWwpMGko4kZrWkRsYR0sn9CiUV+Bzwj6ROSmiWNBT4E/CqbPw7YuKi8R4EVZcrLxfI3St/m7JXXfU7FzKyWoLIQeGeJ9J2Al6pYvvWn/JKi9CUF89aJiBWkRxhPAd4gBZ97eLN7q6byJLVIisKpijqvZ8BGpcc2DNy4lntzmpl1T7UElV8Bl2YjwACQtCvwHeCXVSy/LPtbfD+TIQXz1pH0aeAbpKOTPqQjpc2AK9tTXkS0RIQKpyrqvJ53jih9O5YdfZsWM7OagsrZpO6vmZIWSVpE6nqaB3y1rYUjYinp2Su7t6ZlXWJjgQdLLLI78KeImJ49uvgF4IfAYdn8J4DXi8rbHugHPFRDu2rywNwlJdMffHZpvVZpZrbBqOU6lVeBfSW9jzTiCmBmRPyxhvVNAU6X9EfgedLV+E8Ad5fIOx34vqR3A/8HbE4aLDAzq89rkq4ELpD0ECnAfAu4NTupXxf3zS51T02495lqegDNzLq3mq5TkbQJ8DagGegLTJR0rqRzqyxiMnADKYgsJN1L7NCIWCtpn+zalZEAEXEDcAHwM1J31j+AV4FjCsr7L9JRzr9IR0GvFc3PXbnbsfg2LWZmoKjy2eqS9gBuJ13nNwh4EdiCNNrqhYgYV69K1oukqLb9rcadfSsrSwQQAZO23YwXl7/BNlsMZNK4zRmzWT92G1X9tZgzZi9m9ksr2lyu2nxmZvUgiXLnpWsJKneRuqpOIt2xeGdgJXAtcGlE3JRLbRuoPUFl1/N/z+LXVlWVt2/vZo7YfQTnH7Zjm3nP+80j3DDjOSSIoOxy1eYzM6uXSkGllnGwE4DPZl1Va4E+EfG0pK8APwU2uKDSHv037lV1UHlt1Rquu/9Z3r75ALYbVvqiSYDH57/Cdfc/yxsFd0AutVy1+czMKtm4dzM7jxiMlP8Npmo5UnkJ2DMiZkl6EvhcRNwhaVvgoYjol3vt6qw9Rypfvv5BfvXA8zUt06dZ9Gouf/pq9Zq1JbvUiperNp+ZWTkR6QfvnV+exDZbtO/HaF5HKg+TjlZmkUZjfTW7Z9cJlL59S7f0yT1H1hRU+vZu5ur/fFeb50iO+cl9vLZqTcXlqs1nZlbOqjVr2fbs21i5uj6Di2r5eft1oPVeJOeQTtLfBuwDnJxzvbqs3UcP5dPvLntj5bfo27uZj+++dZs7/N1HD+WI3UfQt3cz/fo0l12u2nxmZp2l6u6vkgun28y/XHMfUhfRnu6vVjPnLOb4K+9nyYrV9GkWB+64JcteW8nCV9o/+mvmnMU8s6jtUV3V5jMzK9Z6pHLryfswvp03ws1l9Fd31JGgYma2Iap3UPHZXTMzy42DipmZ5cb3a++IXxwHs/4Azb1gzL7w+hJYvgA22w62fT8MHQsj96y+vLn/B4ufbnu5avOZmTWYg0p7nb8JxJsXIfLPgrv/L/hHet27H+xyNBz0rbbLu/V0eOBq0g1fovxy1eYzM+sEDirtccGmbw0o5axaAX+/CobtCG+rcCuV+Y+kfKtffzOt1HLV5jMzK0Nrg3F6Nl0FWY/ye/Lop3aN/vrd1+Cey9qztgrzKtWhcLlq85mZlZP2I7OOuIttxu/SRt7S8rqi3gAeubG2/L37wTG/bvscybQPpyObSstVm8/MrIxVK9+g90VboDVv1KV8j/6q1RvLq8/bux/sckzbO/yRe6VzI737QZ/+5ZerNp+ZWSfxkUqtVr5Sft42/wavLIDNt4dt3lfb6KyDvgU7fgwWP1V5uWrzmZl1AgeVmlU4r3F0jV1jxUbuWV2QqDafmVmDufvLzMxy46CSG4++MjNzUMlNzx2abWbWykHFzMxy46BiZma5cVAxM7PcNDSoSGqSdJGkBZKWS7pdUsln80o6KstTOK2W9FBBnhZJa4ry/LxxLTIz2zDd+/RLzJyzOPdyG32kcgZwJDAJGAbMBX4rab16RMQ1ETGgdQI2ARYB04qyTi/MFxFH1rkNZmYbrAtveRSA62c8y9E/vo/zfvNIruU3OqicBEyOiMcjYjkpyGwHTKxi2Y8Cg4Cf1rF+Zmbd1ozZi/nV358DYNLa+3nH6ke5YcZzuR6xNCyoSBoMjAJmtKZFxBJgFjChiiI+B1wfEcWt313Si5LmSLpW0picqmxm1q3MfmkFZzVdCcBne93M1X2+wdlNP+WZRSvaWLJ6jTxSGZT9XVKUvqRgXkmSdgT2Aa4omvVLYAdgC+DdwGrgTkkDSpTRIikKp5pbYGa2Adtx9b84nD8D0E8r6ac3+Ah3seOaR3NbRyODyrLs7+Ci9CEF88r5HDAjIu4vTIyIRyJiTiTzgOOBLYG9iwuIiJaIUOHUrlaYmW2gtt9oEc3Nb93tNzc3sX2fF3NbR8OCSkQsBeYAu7emZV1iY4EHyy0naSBwNOsfpZRcTTY5YJiZFRv6djZqfuvucaNmpTue56TRJ+qnAKdLGiepP3Ax8ARwd4VljgFWAdcVz5D0cUmbZ/9vAfwIeBH4W94VNzPb4DXgmUyNvvX9ZFL3191A/+zvoRGxVtI+wG3A+IiYW7DMZ4GpEfFaifKOAi7PAtTLwF+Af4uICg89MTPrwer8TCY/o77W9rcUnxIqnLe0YxUyM9sAVHpGvW/TYmZmuXFQMTOz3DiomJlZbhxUzMwsNw4qZmaWGwcVMzPLjYOKmZnlxkHFzMxy46BiZma5cVAxM7PcOKiYmVluHFTMzCw3DipmZpYbBxUzM8uNg4qZmeXGQcXMzHLjoGJmZrlxUDEzs9w4qJiZWW4cVMzMLDcOKmZmlhsHFTMzy42DipmZ5aahQUVSk6SLJC2QtFzS7ZJGlcl7VJancFot6aH2lGdmZvXX6COVM4AjgUnAMGAu8FtJ69UjIq6JiAGtE7AJsAiY1p7yzMys/hq98z0JmBwRj0fEclJQ2A6YWMWyHwUGAT/NqTwzM8tZw4KKpMHAKGBGa1pELAFmAROqKOJzwPURsTin8szMLGeNPFIZlP1dUpS+pGBeSZJ2BPYBrmhveZJaJEXhVFWtzcysao0MKsuyv4OL0ocUzCvnc8CMiLi/veVFREtEqHCqqtZmZla1hgWViFgKzAF2b03LurDGAg+WW07SQOBo3nqU0u7yzMysfhp9on4KcLqkcZL6AxcDTwB3V1jmGGAVcF1O5ZmZWZ30avD6JpO6q+4G+md/D42ItZL2AW4DxkfE3IJlPgtMjYjXaimvLrVvKe5pMzOzQoroueerJUVN7W8rqLQs7ViFzMw2AJIod17aFwmamVluHFTy4qMUMzMHFTMzy4+DipmZ5cZBxczMcuOgYmZmuXFQMTOz3Dio1KJ549Lpvfo1th5mZl2Ug0otzllQOv1rLzS2HmZmXZSDSq1alr55ZNKrn69PMTMr4Nu09OD2m5m1h2/TYmZmDeGgYmZmuXFQMTOz3DiomJlZbhxUzMwsN41+8mOXI5UcwGBmZu3Qo4cUd0Q2HLnbR6Se0M6e0EboGe3sCW2Ert1Od3+ZmVluHFTMzCw3DipmZpYbB5X2O7+zK9AgPaGdPaGN0DPa2RPaCF24nT5Rb2ZmufGRipmZ5cZBxczMcuOgAkhqknSRpAWSlku6XdKoCvl3kfQ3SSskzZV0ctH8vpJ+IGmxpGWSrpc0tP4tqawO7TxZ0r3Z/Nl1b0AV8myjpI2y9/FJSa9k878jqW9jWlNeHd7L/ydpTvZ5XSjpRkmj696QCvJuY0G+XpLulxSd3casPnm/l1MlrcrKap0urn9LMhHR4yfgTOAZYDtgAPBD4GGgqUTeQcAC4DxgY2AvYDHwsYI8U4AZwHBgE+AW4H+7YTs/BnwUOBuY3dnty7uNQH/g68D2QDMwBngAuKw7tTPLswMwsKDdlwD3dac2FuQ9F/gdEMDobvheTgWmdlp7OnuDdoUJmA18tuD1EOANYFKJvMcC8wrfcOBi4I/Z/32B14APFsx/R/YBHtld2lki7+zOfh/r2caC+V8AHu7O7cyCyreBpd2tjcCuwCxgJ7pOUMm1nXRyUOnx3V+SBgOjSEcWAETEEtIHb0KJRXYGHoiItQVpMwryjiP9gigs71FgRZnyGqIO7exyGtTG9wMPdbCqHVKvdkr6rKSlwHLgFNKv4U5RjzZK2gi4EvgcsCzvOrdHHT+zh0t6SdJTkqZI2jzPelfS44MK6XASYElR+pKCecX5K+WttbxGybudXVFd2yjpdOA9pO6+zlSXdkbE9yNiMLAV6TqIBzpWzQ6pRxv/G7g3In7f4drlpx7tvIzUZbsZcAAwGrhZDbp7bo+/SzFv/mIZXJQ+hNK/ZpYBb6uQt7C8RVWU1yh5t7MrqlsbJZ0GnAq8LyLmdqiWHVfX9zIi5kn6IfC0pNERsahUvjrLtY2S9gaOIHV7dSW5v5cRMbNg3lOS/hN4FtgWeKIDda1Kjz9SiYilwBxg99a07JB0LPBgiUUeAnaRVLjtdivI+wTwelF52wP96MRukzq0s8upVxsltQBfAt4bEY/kWef2aNB72Zt0bmVEB6vbLnVo4wGknfHTkhYBf8/S/y7prFwrX4MGvZetXWWNuatxZ5+k6goTafTFU6TzIf1Jo7faGn1xDrAR8C7gJeDfC/JMAe4DtiSN/votcEs3bGcv0vmjE0hfjI2zSd2ojd8mjcx5e2e/f/VqZ/YZPQ7YNHu9NfAb0gnkPt2kjYNIAbJ12ot0on4vYFA3ei83Jo3KHJy9HkUafTqjVHl1aU9nbsyuMpGO2L4BLAReJQ03HJ3N24d04nJkQf5dgHtIo7yeBU4uKq8vaVjgy6TD0huAod2wnS3ZF7N4Gt0d2ph9IYM0Emd54dSd3ktS98kdpO7aV4HngGuAbbpLG0uUPbqzP6t1ei/7AdNJw4xfJf3Y+wEwrFHt8b2/zMwsNz3+nIqZmeXHQcXMzHLjoGJmZrlxUDEzs9w4qJiZWW4cVMzMLDcOKmYNJGm2pK/VexmzzuKgYmZmuXFQMTOz3DiomOVI0v6S7lJ6lPRSSX+W9K4K+WdL+rqkH2eP8l2UPVq2+LvZR9L3snIXSLpUUq+Ccmpar1m9OKiY5WsAcAXwbmBv4EngdkmbVljmi6Sn+e0B/BfpAVlfLJHnBWDP7P8vAJ/u4HrNcud7f5nVUXbE8RLwhYi4RtJs4McRcWE2fzbwbETsU7DMRcAxEbF1QZ6HI+LQgjy3AUsi4shq1luPtpmV4iMVsxxJGiNpmqRZkpaR7lLd+sjYcu4pev1XYISkwqf5PViUZx4FD2tq53rNcucnP5rl639Jt5D/POm25CuBu4E+HSx3ZdHr4K0/Cuu1XrOaOKiY5SQ7fzEeOCgifpeljQC2aGPRvYpe7w08HxFVPbq5A+s1y52Dill+XgZeBE6Q9BSwKTCZ9DClSiZkjyy+lvRY2VNIT/ar93rNcudzKmY5iYi1wL+Tni/+MDAV+C5p1FYll5HOfczI/r8c+F4D1muWO4/+MutExaPBzDZ0PlIxM7PcOKiYmVlu3P1lZma58ZGKmZnlxkHFzMxy46BiZma5cVAxM7PcOKiYmVluHFTMzCw3/x9qd5KvwJuaEgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_scores = [clf.score(X_train, y_train) for clf in clfs]\n",
    "test_scores = [clf.score(X_test, y_test) for clf in clfs]\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax.set_xlabel(\"alpha\")\n",
    "ax.set_ylabel(\"accuracy\")\n",
    "ax.set_title(\"Accuracy vs alpha for training and testing sets\")\n",
    "ax.plot(ccp_alphas, train_scores, marker=\"o\", label=\"train\", drawstyle=\"steps-post\")\n",
    "ax.plot(ccp_alphas, test_scores, marker=\"o\", label=\"test\", drawstyle=\"steps-post\")\n",
    "ax.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Using Cost Complexity pruninig as well, we see that the accuracy is not improved significantly. We will move ahead with Ensemble Learning now."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 2. Random Forests - Bagging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will first fit a default Random Forest and then use CV (Randomized Search)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(oob_score=True, random_state=42)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "clf_rf = RandomForestClassifier(random_state=42, oob_score = True)\n",
    "\n",
    "clf_rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9803877703206563"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7336464857341684"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- As we can see, the model seems to be overfitting. This is because we have not applied any hyperparameter tuninig. \n",
    "- We will now try hyperparameter tuninig and see if we get any improvement. \n",
    "- First, let's check how OOB error drops with increase in trees."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check oob error\n",
    "n_estimators = 500\n",
    "\n",
    "oob_errors = []\n",
    "\n",
    "for i in range(1, n_estimators + 1, 10):\n",
    "    clf_rf.set_params(n_estimators=i)\n",
    "    clf_rf.fit(X_train, y_train)\n",
    "    oob_errors.append([i, 1 - clf_rf.oob_score_])\n",
    "\n",
    "xs, ys = zip(*oob_errors)  # * to unpack by 1 dim\n",
    "\n",
    "plt.figure(figsize=(8,4))\n",
    "sns.lineplot(x = xs, y = ys)\n",
    "plt.xlabel(\"Number of Trees\")\n",
    "plt.ylabel(\"OOB Error\")\n",
    "plt.title(\"OOB Error Plot\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Even for 500 trees, we are not getting any significant improvement. \n",
    "- Let's try some other stuff."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=RandomForestClassifier(random_state=42),\n",
       "             param_grid={'max_features': ['auto', 'sqrt', 'log2'],\n",
       "                         'n_estimators': [200, 400, 600]},\n",
       "             return_train_score=True, scoring='accuracy')"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "params = {'max_features': ['auto', 'sqrt', 'log2'],\n",
    "          'n_estimators': [200, 400, 600]\n",
    "         }\n",
    "\n",
    "clf_rf_cv = GridSearchCV(estimator = RandomForestClassifier(random_state = 42),\n",
    "                         param_grid = params,\n",
    "                         cv = 3,\n",
    "                         return_train_score = True,\n",
    "                         scoring = 'accuracy'\n",
    "                        )\n",
    "\n",
    "clf_rf_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7449664429530202"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9803877703206563"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf_cv.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7360821155184412"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf_cv.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It seems that tuninig hyperparameters has not solved the issue of overfitting using Random Forests either. Let's check the feature importance chart nonetheless."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.30172669, 0.02641152, 0.02290648, 0.02071453, 0.0339138 ,\n",
       "       0.0188256 , 0.0696447 , 0.04391115, 0.46194553])"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf_cv.best_estimator_.feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_scores = pd.Series(clf_rf_cv.best_estimator_.feature_importances_, index=X_train.columns).sort_values(ascending=False)\n",
    "\n",
    "sns.barplot(x=feature_scores, y=feature_scores.index)\n",
    "plt.xlabel('Gini Importance Score')\n",
    "plt.ylabel('Features')\n",
    "plt.title(\"Feature Importance\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Let's try and build a Random Forest usiong the two important features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['log_training_hours', 'city_development_index'], dtype='object')"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_scores.index[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(oob_score=True, random_state=42)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf = RandomForestClassifier(random_state=42, oob_score = True)\n",
    "clf_rf.fit(X_train[feature_scores.index[:2]], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.839821029082774"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf.score(X_train[feature_scores.index[:2]], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7413013221990258"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_rf.score(X_test[feature_scores.index[:2]], y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- A little improvement but still negligible. Let's try some boosting methods now."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 3. AdaBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AdaBoostClassifier(random_state=42)"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "\n",
    "clf_adaBoost = AdaBoostClassifier(random_state=42)\n",
    "\n",
    "clf_adaBoost.fit(X_train, y_train)\n",
    "\n",
    "# if you use base_estimator = DecisionTreeClassifier(), AdaBoost will grow the trees to it's fullest (default DT settings); \n",
    "# so unless using a specific DT, use base_estimator = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.784041759880686"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7694850382741824"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'algorithm': 'SAMME.R',\n",
       " 'base_estimator': None,\n",
       " 'learning_rate': 1.0,\n",
       " 'n_estimators': 50,\n",
       " 'random_state': 42}"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost.get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 35 candidates, totalling 105 fits\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.2s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.5s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.5s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.4s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   0.9s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   0.9s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   1.0s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   2.0s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   2.0s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   1.9s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   2.9s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   2.9s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   3.0s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   3.1s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   3.0s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   3.0s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.2s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.2s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.4s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.4s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.4s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.9s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.9s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.9s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   2.1s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   2.0s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   1.9s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   3.0s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   3.0s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   3.0s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.2s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.2s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.2s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.4s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.5s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.4s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   1.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   1.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   2.0s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   2.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   3.0s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   3.0s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.2s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   1.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   1.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   1.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   2.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   3.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   3.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   2.9s\n",
      "[CV] END .................learning_rate=1.5, n_estimators=50; total time=   0.2s\n",
      "[CV] END .................learning_rate=1.5, n_estimators=50; total time=   0.2s\n",
      "[CV] END .................learning_rate=1.5, n_estimators=50; total time=   0.2s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=100; total time=   0.4s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=100; total time=   0.4s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=100; total time=   0.4s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=200; total time=   0.9s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=400; total time=   1.9s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=400; total time=   2.0s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=400; total time=   2.0s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=600; total time=   3.0s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=600; total time=   3.0s\n",
      "[CV] END ................learning_rate=1.5, n_estimators=600; total time=   3.0s\n",
      "[CV] END ...................learning_rate=2, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...................learning_rate=2, n_estimators=50; total time=   0.2s\n",
      "[CV] END ...................learning_rate=2, n_estimators=50; total time=   0.2s\n",
      "[CV] END ..................learning_rate=2, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=2, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=2, n_estimators=100; total time=   0.4s\n",
      "[CV] END ..................learning_rate=2, n_estimators=200; total time=   0.9s\n",
      "[CV] END ..................learning_rate=2, n_estimators=200; total time=   0.9s\n",
      "[CV] END ..................learning_rate=2, n_estimators=200; total time=   1.0s\n",
      "[CV] END ..................learning_rate=2, n_estimators=400; total time=   2.0s\n",
      "[CV] END ..................learning_rate=2, n_estimators=400; total time=   1.9s\n",
      "[CV] END ..................learning_rate=2, n_estimators=400; total time=   2.0s\n",
      "[CV] END ..................learning_rate=2, n_estimators=600; total time=   3.1s\n",
      "[CV] END ..................learning_rate=2, n_estimators=600; total time=   3.1s\n",
      "[CV] END ..................learning_rate=2, n_estimators=600; total time=   3.0s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=AdaBoostClassifier(random_state=42),\n",
       "             param_grid={'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1, 1.5, 2],\n",
       "                         'n_estimators': [50, 100, 200, 400, 600]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "params = {'n_estimators': [50, 100, 200, 400, 600],\n",
    "          'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1, 1.5, 2]\n",
    "         }\n",
    "\n",
    "clf_adaBoost_cv = GridSearchCV(estimator = AdaBoostClassifier(random_state=42),\n",
    "                               param_grid = params,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_adaBoost_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.786875466070097"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.786875466070097"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost_cv.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7717466945024356"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost_cv.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'algorithm': 'SAMME.R',\n",
       " 'base_estimator': None,\n",
       " 'learning_rate': 0.0001,\n",
       " 'n_estimators': 50,\n",
       " 'random_state': 42}"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_adaBoost_cv.best_estimator_.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- AdaBoost is giving us a traing and test accuracy of ~78%. While not great, at least our model is not overfitting."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 4. Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(random_state=42)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "gbm_clf = GradientBoostingClassifier(random_state=42)\n",
    "gbm_clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7903057419835944"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gbm_clf.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7726165622825331"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gbm_clf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 50 candidates, totalling 150 fits\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.0001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.5; total time=   0.4s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.0001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.5; total time=   0.7s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.5; total time=   2.6s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.5; total time=   2.6s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.0001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.001, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.001, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.5; total time=   2.4s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.5; total time=   2.4s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.001, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END .learning_rate=0.01, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END .learning_rate=0.01, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END .learning_rate=0.01, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END learning_rate=0.01, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.01, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.01, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.01, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.01, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.5; total time=   1.8s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.5; total time=   1.7s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.5; total time=   1.7s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.75; total time=   2.0s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.75; total time=   2.0s\n",
      "[CV] END learning_rate=0.01, n_estimators=400, subsample=0.75; total time=   2.0s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.01, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END ..learning_rate=0.1, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END ..learning_rate=0.1, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END ..learning_rate=0.1, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END .learning_rate=0.1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END .learning_rate=0.1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END .learning_rate=0.1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END .learning_rate=0.1, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END .learning_rate=0.1, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END .learning_rate=0.1, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END learning_rate=0.1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END learning_rate=0.1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END .learning_rate=0.1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END .learning_rate=0.1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END .learning_rate=0.1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END learning_rate=0.1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END learning_rate=0.1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END .learning_rate=0.1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END .learning_rate=0.1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END .learning_rate=0.1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END learning_rate=0.1, n_estimators=400, subsample=0.75; total time=   2.0s\n",
      "[CV] END learning_rate=0.1, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END learning_rate=0.1, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END .learning_rate=0.1, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END .learning_rate=0.1, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END .learning_rate=0.1, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END learning_rate=0.1, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.1, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END learning_rate=0.1, n_estimators=600, subsample=0.75; total time=   3.1s\n",
      "[CV] END ....learning_rate=1, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END ....learning_rate=1, n_estimators=50, subsample=0.5; total time=   0.2s\n",
      "[CV] END ....learning_rate=1, n_estimators=50, subsample=0.5; total time=   0.1s\n",
      "[CV] END ...learning_rate=1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END ...learning_rate=1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END ...learning_rate=1, n_estimators=50, subsample=0.75; total time=   0.2s\n",
      "[CV] END ...learning_rate=1, n_estimators=100, subsample=0.5; total time=   0.4s\n",
      "[CV] END ...learning_rate=1, n_estimators=100, subsample=0.5; total time=   0.4s\n",
      "[CV] END ...learning_rate=1, n_estimators=100, subsample=0.5; total time=   0.3s\n",
      "[CV] END ..learning_rate=1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END ..learning_rate=1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END ..learning_rate=1, n_estimators=100, subsample=0.75; total time=   0.4s\n",
      "[CV] END ...learning_rate=1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END ...learning_rate=1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END ...learning_rate=1, n_estimators=200, subsample=0.5; total time=   0.8s\n",
      "[CV] END ..learning_rate=1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=200, subsample=0.75; total time=   0.9s\n",
      "[CV] END ...learning_rate=1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END ...learning_rate=1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END ...learning_rate=1, n_estimators=400, subsample=0.5; total time=   1.6s\n",
      "[CV] END ..learning_rate=1, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=400, subsample=0.75; total time=   1.9s\n",
      "[CV] END ...learning_rate=1, n_estimators=600, subsample=0.5; total time=   2.4s\n",
      "[CV] END ...learning_rate=1, n_estimators=600, subsample=0.5; total time=   2.5s\n",
      "[CV] END ...learning_rate=1, n_estimators=600, subsample=0.5; total time=   2.4s\n",
      "[CV] END ..learning_rate=1, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=600, subsample=0.75; total time=   2.9s\n",
      "[CV] END ..learning_rate=1, n_estimators=600, subsample=0.75; total time=   2.9s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=GradientBoostingClassifier(random_state=42),\n",
       "             param_grid={'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1],\n",
       "                         'n_estimators': [50, 100, 200, 400, 600],\n",
       "                         'subsample': [0.5, 0.75]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "params = {'n_estimators': [50, 100, 200, 400, 600],\n",
    "          'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1],\n",
    "          'subsample' : [0.5, 0.75]\n",
    "         }\n",
    "\n",
    "clf_gbm_cv = GridSearchCV(estimator = GradientBoostingClassifier(random_state=42),\n",
    "                               param_grid = params,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_gbm_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradientBoostingClassifier(learning_rate=0.01, n_estimators=400,\n",
       "                           random_state=42, subsample=0.5)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_gbm_cv.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7859060402684565"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_gbm_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7879194630872484"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_gbm_cv.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7705288796102993"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_gbm_cv.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 5. XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "              colsample_bynode=1, colsample_bytree=1, enable_categorical=False,\n",
       "              eval_metric='logloss', gamma=0, gpu_id=-1, importance_type=None,\n",
       "              interaction_constraints='', learning_rate=0.300000012,\n",
       "              max_delta_step=0, max_depth=6, min_child_weight=1, missing=nan,\n",
       "              monotone_constraints='()', n_estimators=100, n_jobs=8,\n",
       "              num_parallel_tree=1, predictor='auto', random_state=42,\n",
       "              reg_alpha=0, reg_lambda=1, scale_pos_weight=1, subsample=1,\n",
       "              tree_method='exact', validate_parameters=1, verbosity=None)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from xgboost import XGBClassifier\n",
    "\n",
    "clf_xgb = XGBClassifier(random_state = 42, eval_metric = 'logloss')\n",
    "clf_xgb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8413870246085011"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7627000695894224"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We get a decent model with XGBoost as well. Let's see if we can better this with CV and some hyperparameter tuning.\n",
    "- We will now do a multistep CV tuning the individual learner hyperparameters first and then the model hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 12 candidates, totalling 36 fits\n",
      "[CV] END ....................max_depth=3, min_child_weight=1; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=1; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=1; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=3; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=3; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=3; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=5; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=5; total time=   0.1s\n",
      "[CV] END ....................max_depth=3, min_child_weight=5; total time=   0.1s\n",
      "[CV] END ....................max_depth=5, min_child_weight=1; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=1; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=1; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=3; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=3; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=3; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=5; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=5; total time=   0.2s\n",
      "[CV] END ....................max_depth=5, min_child_weight=5; total time=   0.2s\n",
      "[CV] END ....................max_depth=7, min_child_weight=1; total time=   0.4s\n",
      "[CV] END ....................max_depth=7, min_child_weight=1; total time=   0.5s\n",
      "[CV] END ....................max_depth=7, min_child_weight=1; total time=   0.4s\n",
      "[CV] END ....................max_depth=7, min_child_weight=3; total time=   0.3s\n",
      "[CV] END ....................max_depth=7, min_child_weight=3; total time=   0.3s\n",
      "[CV] END ....................max_depth=7, min_child_weight=3; total time=   0.3s\n",
      "[CV] END ....................max_depth=7, min_child_weight=5; total time=   0.3s\n",
      "[CV] END ....................max_depth=7, min_child_weight=5; total time=   0.3s\n",
      "[CV] END ....................max_depth=7, min_child_weight=5; total time=   0.3s\n",
      "[CV] END ....................max_depth=9, min_child_weight=1; total time=   0.6s\n",
      "[CV] END ....................max_depth=9, min_child_weight=1; total time=   0.6s\n",
      "[CV] END ....................max_depth=9, min_child_weight=1; total time=   0.5s\n",
      "[CV] END ....................max_depth=9, min_child_weight=3; total time=   0.4s\n",
      "[CV] END ....................max_depth=9, min_child_weight=3; total time=   0.4s\n",
      "[CV] END ....................max_depth=9, min_child_weight=3; total time=   0.4s\n",
      "[CV] END ....................max_depth=9, min_child_weight=5; total time=   0.4s\n",
      "[CV] END ....................max_depth=9, min_child_weight=5; total time=   0.5s\n",
      "[CV] END ....................max_depth=9, min_child_weight=5; total time=   0.4s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3,\n",
       "             estimator=XGBClassifier(base_score=None, booster=None,\n",
       "                                     colsample_bylevel=None,\n",
       "                                     colsample_bynode=None,\n",
       "                                     colsample_bytree=None,\n",
       "                                     enable_categorical=False,\n",
       "                                     eval_metric='logloss', gamma=None,\n",
       "                                     gpu_id=None, importance_type=None,\n",
       "                                     interaction_constraints=None,\n",
       "                                     learning_rate=None, max_delta_step=None,\n",
       "                                     max_depth=None, min_child_weight=None,\n",
       "                                     missing=nan, mono...constraints=None,\n",
       "                                     n_estimators=100, n_jobs=None,\n",
       "                                     num_parallel_tree=None, predictor=None,\n",
       "                                     random_state=42, reg_alpha=None,\n",
       "                                     reg_lambda=None, scale_pos_weight=None,\n",
       "                                     subsample=None, tree_method=None,\n",
       "                                     validate_parameters=None, verbosity=None),\n",
       "             param_grid={'max_depth': range(3, 10, 2),\n",
       "                         'min_child_weight': range(1, 6, 2)},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "param_test1 = {\n",
    "    'max_depth':range(3,10,2),\n",
    "    'min_child_weight':range(1,6,2)\n",
    "}\n",
    "\n",
    "clf_xgb_cv = GridSearchCV(estimator = XGBClassifier(random_state = 42, eval_metric = 'logloss'),\n",
    "                               param_grid = param_test1,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_xgb_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7782252050708426"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 3, 'min_child_weight': 3}"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 20 candidates, totalling 60 fits\n",
      "[CV] END ......................gamma=0.0, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.0, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.0, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.0, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.0, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.0, min_child_weight=8; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.0, min_child_weight=12; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.1, min_child_weight=8; total time=   0.1s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=10; total time=   0.2s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.1, min_child_weight=12; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.2, min_child_weight=8; total time=   0.1s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=12; total time=   0.2s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.2, min_child_weight=12; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.3, min_child_weight=8; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.3, min_child_weight=12; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=6; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=8; total time=   0.1s\n",
      "[CV] END ......................gamma=0.4, min_child_weight=8; total time=   0.1s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=10; total time=   0.1s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=12; total time=   0.1s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=12; total time=   0.2s\n",
      "[CV] END .....................gamma=0.4, min_child_weight=12; total time=   0.1s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3,\n",
       "             estimator=XGBClassifier(base_score=None, booster=None,\n",
       "                                     colsample_bylevel=None,\n",
       "                                     colsample_bynode=None,\n",
       "                                     colsample_bytree=None,\n",
       "                                     enable_categorical=False,\n",
       "                                     eval_metric='logloss', gamma=None,\n",
       "                                     gpu_id=None, importance_type=None,\n",
       "                                     interaction_constraints=None,\n",
       "                                     learning_rate=None, max_delta_step=None,\n",
       "                                     max_depth=3, min_child_weight=3,\n",
       "                                     missing=nan, monotone_constraints=None,\n",
       "                                     n_estimators=100, n_jobs=None,\n",
       "                                     num_parallel_tree=None, predictor=None,\n",
       "                                     random_state=42, reg_alpha=None,\n",
       "                                     reg_lambda=None, scale_pos_weight=None,\n",
       "                                     subsample=None, tree_method=None,\n",
       "                                     validate_parameters=None, verbosity=None),\n",
       "             param_grid={'gamma': [0.0, 0.1, 0.2, 0.3, 0.4],\n",
       "                         'min_child_weight': [6, 8, 10, 12]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_test1 = {\n",
    "    'min_child_weight' : [6,8,10,12],\n",
    "    'gamma' : [0.0, 0.1, 0.2, 0.3, 0.4]\n",
    "}\n",
    "\n",
    "clf_xgb_cv = GridSearchCV(estimator = XGBClassifier(random_state = 42, eval_metric = 'logloss',\n",
    "                                                    max_depth = 3, min_child_weight = 3\n",
    "                                                   ),\n",
    "                               param_grid = param_test1,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_xgb_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7809843400447427"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'gamma': 0.4, 'min_child_weight': 8}"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 9 candidates, totalling 27 fits\n",
      "[CV] END ................colsample_bytree=0.5, subsample=0.5; total time=   0.2s\n",
      "[CV] END ................colsample_bytree=0.5, subsample=0.5; total time=   0.1s\n",
      "[CV] END ................colsample_bytree=0.5, subsample=0.5; total time=   0.1s\n",
      "[CV] END ...............colsample_bytree=0.5, subsample=0.75; total time=   0.1s\n",
      "[CV] END ...............colsample_bytree=0.5, subsample=0.75; total time=   0.2s\n",
      "[CV] END ...............colsample_bytree=0.5, subsample=0.75; total time=   0.1s\n",
      "[CV] END ..................colsample_bytree=0.5, subsample=1; total time=   0.1s\n",
      "[CV] END ..................colsample_bytree=0.5, subsample=1; total time=   0.1s\n",
      "[CV] END ..................colsample_bytree=0.5, subsample=1; total time=   0.1s\n",
      "[CV] END ...............colsample_bytree=0.75, subsample=0.5; total time=   0.2s\n",
      "[CV] END ...............colsample_bytree=0.75, subsample=0.5; total time=   0.2s\n",
      "[CV] END ...............colsample_bytree=0.75, subsample=0.5; total time=   0.2s\n",
      "[CV] END ..............colsample_bytree=0.75, subsample=0.75; total time=   0.1s\n",
      "[CV] END ..............colsample_bytree=0.75, subsample=0.75; total time=   0.1s\n",
      "[CV] END ..............colsample_bytree=0.75, subsample=0.75; total time=   0.2s\n",
      "[CV] END .................colsample_bytree=0.75, subsample=1; total time=   0.1s\n",
      "[CV] END .................colsample_bytree=0.75, subsample=1; total time=   0.1s\n",
      "[CV] END .................colsample_bytree=0.75, subsample=1; total time=   0.1s\n",
      "[CV] END ..................colsample_bytree=1, subsample=0.5; total time=   0.2s\n",
      "[CV] END ..................colsample_bytree=1, subsample=0.5; total time=   0.2s\n",
      "[CV] END ..................colsample_bytree=1, subsample=0.5; total time=   0.2s\n",
      "[CV] END .................colsample_bytree=1, subsample=0.75; total time=   0.2s\n",
      "[CV] END .................colsample_bytree=1, subsample=0.75; total time=   0.2s\n",
      "[CV] END .................colsample_bytree=1, subsample=0.75; total time=   0.2s\n",
      "[CV] END ....................colsample_bytree=1, subsample=1; total time=   0.1s\n",
      "[CV] END ....................colsample_bytree=1, subsample=1; total time=   0.1s\n",
      "[CV] END ....................colsample_bytree=1, subsample=1; total time=   0.1s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3,\n",
       "             estimator=XGBClassifier(base_score=None, booster=None,\n",
       "                                     colsample_bylevel=None,\n",
       "                                     colsample_bynode=None,\n",
       "                                     colsample_bytree=None,\n",
       "                                     enable_categorical=False,\n",
       "                                     eval_metric='logloss', gamma=0.4,\n",
       "                                     gpu_id=None, importance_type=None,\n",
       "                                     interaction_constraints=None,\n",
       "                                     learning_rate=None, max_delta_step=None,\n",
       "                                     max_depth=3, min_child_weight=8,\n",
       "                                     missing=nan, monotone_constraints=None,\n",
       "                                     n_estimators=100, n_jobs=None,\n",
       "                                     num_parallel_tree=None, predictor=None,\n",
       "                                     random_state=42, reg_alpha=None,\n",
       "                                     reg_lambda=None, scale_pos_weight=None,\n",
       "                                     subsample=None, tree_method=None,\n",
       "                                     validate_parameters=None, verbosity=None),\n",
       "             param_grid={'colsample_bytree': [0.5, 0.75, 1],\n",
       "                         'subsample': [0.5, 0.75, 1]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_test1 = {\n",
    "    'subsample':[0.5, 0.75, 1],\n",
    "    'colsample_bytree':[0.5, 0.75, 1]\n",
    "}\n",
    "\n",
    "clf_xgb_cv = GridSearchCV(estimator = XGBClassifier(random_state = 42, eval_metric = 'logloss',\n",
    "                                                    max_depth = 3, min_child_weight = 8, \n",
    "                                                    gamma = 0.4\n",
    "                                                   ),\n",
    "                               param_grid = param_test1,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_xgb_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7809843400447427"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'colsample_bytree': 1, 'subsample': 1}"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 36 candidates, totalling 108 fits\n",
      "[CV] END ..................reg_alpha=1e-05, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ..................reg_alpha=1e-05, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ..................reg_alpha=1e-05, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=1e-05, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=1e-05, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=1e-05, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=0.1; total time=   0.2s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=1e-05, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1e-05, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=0.01, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=0.01, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ...................reg_alpha=0.01, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.01, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.01, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.01, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=1; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=1; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=1; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=100; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=100; total time=   0.2s\n",
      "[CV] END .....................reg_alpha=0.01, reg_lambda=100; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=0; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=0; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0.01, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=0.1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=0.1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0.1, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=0.1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=1, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=1, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=1, reg_lambda=100; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ..........................reg_alpha=1, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=100, reg_lambda=1e-05; total time=   0.0s\n",
      "[CV] END ....................reg_alpha=100, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ....................reg_alpha=100, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=100, reg_lambda=0.01; total time=   0.0s\n",
      "[CV] END .....................reg_alpha=100, reg_lambda=0.01; total time=   0.1s\n",
      "[CV] END .....................reg_alpha=100, reg_lambda=0.01; total time=   0.2s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=0.1; total time=   0.0s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=0.1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=1; total time=   0.0s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=1; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=100, reg_lambda=100; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=0; total time=   0.0s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ........................reg_alpha=100, reg_lambda=0; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END ......................reg_alpha=0, reg_lambda=1e-05; total time=   0.1s\n",
      "[CV] END .......................reg_alpha=0, reg_lambda=0.01; total time=   0.2s\n",
      "[CV] END .......................reg_alpha=0, reg_lambda=0.01; total time=   0.2s\n",
      "[CV] END .......................reg_alpha=0, reg_lambda=0.01; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=0.1; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=0.1; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=0.1; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=1; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=1; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=1; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=100; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=100; total time=   0.2s\n",
      "[CV] END ........................reg_alpha=0, reg_lambda=100; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=0; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=0; total time=   0.2s\n",
      "[CV] END ..........................reg_alpha=0, reg_lambda=0; total time=   0.3s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3,\n",
       "             estimator=XGBClassifier(base_score=None, booster=None,\n",
       "                                     colsample_bylevel=None,\n",
       "                                     colsample_bynode=None,\n",
       "                                     colsample_bytree=None,\n",
       "                                     enable_categorical=False,\n",
       "                                     eval_metric='logloss', gamma=0.4,\n",
       "                                     gpu_id=None, importance_type=None,\n",
       "                                     interaction_constraints=None,\n",
       "                                     learning_rate=None, max_delta_step=None,\n",
       "                                     max_depth=3, min_child_weight=8,\n",
       "                                     missing=nan, monotone_co...\n",
       "                                     n_estimators=100, n_jobs=None,\n",
       "                                     num_parallel_tree=None, predictor=None,\n",
       "                                     random_state=42, reg_alpha=None,\n",
       "                                     reg_lambda=None, scale_pos_weight=None,\n",
       "                                     subsample=None, tree_method=None,\n",
       "                                     validate_parameters=None, verbosity=None),\n",
       "             param_grid={'reg_alpha': [1e-05, 0.01, 0.1, 1, 100, 0],\n",
       "                         'reg_lambda': [1e-05, 0.01, 0.1, 1, 100, 0]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_test1 = {\n",
    "    'reg_alpha' : [1e-5, 1e-2, 0.1, 1, 100, 0],\n",
    "    'reg_lambda' : [1e-5, 1e-2, 0.1, 1, 100, 0],\n",
    "}\n",
    "\n",
    "clf_xgb_cv = GridSearchCV(estimator = XGBClassifier(random_state = 42, eval_metric = 'logloss',\n",
    "                                                    max_depth = 3, min_child_weight = 8, \n",
    "                                                    gamma = 0.4\n",
    "                                                   ),\n",
    "                               param_grid = param_test1,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_xgb_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.786875466070097"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'reg_alpha': 100, 'reg_lambda': 1e-05}"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 25 candidates, totalling 75 fits\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.0s\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.0s\n",
      "[CV] END ..............learning_rate=0.0001, n_estimators=50; total time=   0.0s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.0s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.1s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=100; total time=   0.1s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   0.2s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   0.2s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=200; total time=   0.2s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   0.4s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   0.5s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=400; total time=   0.6s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   0.7s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   0.8s\n",
      "[CV] END .............learning_rate=0.0001, n_estimators=600; total time=   0.8s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.0s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.0s\n",
      "[CV] END ...............learning_rate=0.001, n_estimators=50; total time=   0.0s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.0s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.1s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=100; total time=   0.1s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=200; total time=   0.2s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   0.4s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   0.5s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=400; total time=   0.5s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   0.8s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   1.0s\n",
      "[CV] END ..............learning_rate=0.001, n_estimators=600; total time=   1.0s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.0s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.0s\n",
      "[CV] END ................learning_rate=0.01, n_estimators=50; total time=   0.0s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.1s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.1s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=100; total time=   0.1s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.3s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=200; total time=   0.3s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   0.6s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   0.6s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=400; total time=   0.6s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   1.2s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   1.0s\n",
      "[CV] END ...............learning_rate=0.01, n_estimators=600; total time=   1.3s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.0s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.1s\n",
      "[CV] END .................learning_rate=0.1, n_estimators=50; total time=   0.0s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.1s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.1s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=100; total time=   0.2s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.5s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.5s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=200; total time=   0.5s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   1.1s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   0.9s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=400; total time=   0.8s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   1.3s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   1.4s\n",
      "[CV] END ................learning_rate=0.1, n_estimators=600; total time=   1.5s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.0s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.0s\n",
      "[CV] END ...................learning_rate=1, n_estimators=50; total time=   0.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=100; total time=   0.0s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   0.2s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   0.1s\n",
      "[CV] END ..................learning_rate=1, n_estimators=200; total time=   0.2s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   0.3s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   0.4s\n",
      "[CV] END ..................learning_rate=1, n_estimators=400; total time=   0.5s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   0.7s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   0.6s\n",
      "[CV] END ..................learning_rate=1, n_estimators=600; total time=   0.6s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3,\n",
       "             estimator=XGBClassifier(base_score=None, booster=None,\n",
       "                                     colsample_bylevel=None,\n",
       "                                     colsample_bynode=None,\n",
       "                                     colsample_bytree=None,\n",
       "                                     enable_categorical=False,\n",
       "                                     eval_metric='logloss', gamma=0.4,\n",
       "                                     gpu_id=None, importance_type=None,\n",
       "                                     interaction_constraints=None,\n",
       "                                     learning_rate=None, max_delta_step=None,\n",
       "                                     max_depth=3, min_child_weight=8,\n",
       "                                     missing=nan, monotone_co...\n",
       "                                     n_estimators=100, n_jobs=None,\n",
       "                                     num_parallel_tree=None, predictor=None,\n",
       "                                     random_state=42, reg_alpha=100,\n",
       "                                     reg_lambda=1e-05, scale_pos_weight=None,\n",
       "                                     subsample=None, tree_method=None,\n",
       "                                     validate_parameters=None, verbosity=None),\n",
       "             param_grid={'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1],\n",
       "                         'n_estimators': [50, 100, 200, 400, 600]},\n",
       "             return_train_score=True, scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "param_test1 = {\n",
    "    'n_estimators': [50, 100, 200, 400, 600],\n",
    "    'learning_rate': [0.0001, 0.001, 0.01, 0.1, 1]\n",
    "}\n",
    "\n",
    "clf_xgb_cv = GridSearchCV(estimator = XGBClassifier(random_state = 42, eval_metric = 'logloss',\n",
    "                                                    max_depth = 3, min_child_weight = 8, \n",
    "                                                    gamma = 0.4, reg_alpha = 100, reg_lambda = 1e-05\n",
    "                                                   ),\n",
    "                               param_grid = param_test1,\n",
    "                               cv = 3,\n",
    "                               return_train_score = True,\n",
    "                               scoring = 'accuracy',\n",
    "                               verbose = 2\n",
    "                              )\n",
    "\n",
    "clf_xgb_cv.fit(X_train, y_train) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.786875466070097"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'learning_rate': 0.01, 'n_estimators': 100}"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We now have final model built for XGBoost. Let's compare train and test accuracies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.786875466070097"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7717466945024356"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_xgb_cv.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 6. KNN Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We have repeatedly seen that the two most important features in the dataset are `city_development_index` and `log_training_hours`.\n",
    "- We will first build a kNN classifier using all featues and then using only the two important features.\n",
    "- Also, since this is a distance based algorithm, it is important to scale the features first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.35129741, 0.        , 1.        , 0.        , 1.        ,\n",
       "        1.        , 0.5       , 0.5       , 0.63414285],\n",
       "       [0.94211577, 1.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 1.        , 1.        , 0.66548514],\n",
       "       [0.35129741, 0.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 0.25      , 0.5       , 0.84324928],\n",
       "       [0.46706587, 1.        , 0.        , 1.        , 0.5       ,\n",
       "        1.        , 0.        , 0.        , 0.65439054],\n",
       "       [0.35129741, 1.        , 0.        , 1.        , 0.        ,\n",
       "        0.        , 0.5       , 0.        , 0.72536137]])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Niormalizing Data - MinMaxScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "norm = MinMaxScaler().fit(X_train)\n",
    "\n",
    "# transform training data\n",
    "X_train_norm = norm.transform(X_train)\n",
    "\n",
    "# transform testing dataabs\n",
    "X_test_norm = norm.transform(X_test)\n",
    "\n",
    "X_train_norm[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier()"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "clf_knn = KNeighborsClassifier()\n",
    "clf_knn.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7419972164231037"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-Now, let's try using only the top two features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.35129741, 0.63414285],\n",
       "       [0.94211577, 0.66548514],\n",
       "       [0.35129741, 0.84324928],\n",
       "       ...,\n",
       "       [0.88822355, 0.69502734],\n",
       "       [0.35129741, 0.4766264 ],\n",
       "       [0.83433134, 0.67250271]])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_norm[:, [0,-1]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsClassifier()"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn = KNeighborsClassifier()\n",
    "clf_knn.fit(X_train_norm[:, [0,-1]], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7350382741823243"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn.score(X_test_norm[:, [0,-1]], y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- We see there is no improvement. Let's continue with the earlier approach with all features and check with some hyperparameter tuning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 40 candidates, totalling 120 fits\n",
      "[CV] END ................n_neighbors=3, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=3, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=3, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ................n_neighbors=3, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=3, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=3, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=3, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=6, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=6, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=9, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=9, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=9, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ................n_neighbors=9, p=2, weights=uniform; total time=   0.1s\n",
      "[CV] END ...............n_neighbors=9, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=9, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=9, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=12, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=12, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=15, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=15, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=18, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=18, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=21, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=21, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=24, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=24, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=27, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=27, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=1, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=1, weights=uniform; total time=   0.1s\n",
      "[CV] END ..............n_neighbors=30, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=30, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=30, p=1, weights=distance; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ...............n_neighbors=30, p=2, weights=uniform; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=30, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=30, p=2, weights=distance; total time=   0.0s\n",
      "[CV] END ..............n_neighbors=30, p=2, weights=distance; total time=   0.0s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=KNeighborsClassifier(),\n",
       "             param_grid={'n_neighbors': array([ 3,  6,  9, 12, 15, 18, 21, 24, 27, 30]),\n",
       "                         'p': [1, 2], 'weights': ['uniform', 'distance']},\n",
       "             scoring='accuracy', verbose=2)"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "params = {'n_neighbors': np.arange(3, 31, 3),\n",
    "          'weights': ['uniform', 'distance'],\n",
    "          'p' : [1, 2]      # p = 1, Manhattan; p = 2, Euclidean distance\n",
    "         }\n",
    "\n",
    "clf_knn_cv = GridSearchCV(estimator = KNeighborsClassifier(),\n",
    "                          param_grid = params,\n",
    "                          cv = 3,\n",
    "                          scoring = 'accuracy',\n",
    "                          verbose = 2\n",
    "                         )\n",
    "\n",
    "clf_knn_cv.fit(X_train_norm[:, [0,-1]], y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_neighbors': 27, 'p': 2, 'weights': 'uniform'}"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7787472035794183"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7684411969380655"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_knn_cv.score(X_test_norm[:, [0,-1]], y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 7. Naive Bayes Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MultinomialNB()"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.naive_bayes import MultinomialNB\n",
    "\n",
    "clf_nb = MultinomialNB()\n",
    "clf_nb.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'alpha': 1.0, 'class_prior': None, 'fit_prior': True}"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_nb.get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7517524235645041"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_nb.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7480862908837856"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_nb.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 5 candidates, totalling 15 fits\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=MultinomialNB(),\n",
       "             param_grid={'alpha': [0.1, 1, 10, 100, 1000]}, scoring='accuracy',\n",
       "             verbose=1)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grid Search to find alpha\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "params = {'alpha' : [0.1, 1, 10, 100, 1000]}\n",
    "\n",
    "clf_nb_cv = GridSearchCV(estimator = MultinomialNB(),\n",
    "                          param_grid = params,\n",
    "                          cv = 3,\n",
    "                          scoring = 'accuracy',\n",
    "                          verbose = 1\n",
    "                         )\n",
    "\n",
    "clf_nb_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'alpha': 0.1}"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_nb_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7517524235645041"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_nb_cv.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 8. Gaussian Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GaussianNB()"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "\n",
    "clk_gaussNB = GaussianNB()\n",
    "clk_gaussNB.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.767710663683818"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clk_gaussNB.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.755741127348643"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clk_gaussNB.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 9 candidates, totalling 27 fits\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=3, estimator=GaussianNB(),\n",
       "             param_grid={'var_smoothing': [0, 1e-20, 1e-15, 1e-09, 1e-06, 0.001,\n",
       "                                           0.1, 1, 10]},\n",
       "             scoring='accuracy', verbose=1)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# CV to find smooting parameter\n",
    "params = {'var_smoothing' : [0, 1e-20, 1e-15, 1e-9, 1e-6, 0.001, 0.1, 1, 10]}\n",
    "\n",
    "clk_gaussNB_cv = GridSearchCV(estimator = GaussianNB(),\n",
    "                          param_grid = params,\n",
    "                          cv = 3,\n",
    "                          scoring = 'accuracy',\n",
    "                          verbose = 1\n",
    "                         )\n",
    "\n",
    "clk_gaussNB_cv.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'var_smoothing': 0}"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clk_gaussNB_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.767710663683818"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clk_gaussNB_cv.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.755741127348643"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clk_gaussNB_cv.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 9. SVM Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Since SVM is a distance based algorithm, it is important to first normalize the features (mean = 0 and SD = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.35129741, 0.        , 1.        , 0.        , 1.        ,\n",
       "        1.        , 0.5       , 0.5       , 0.63414285],\n",
       "       [0.94211577, 1.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 1.        , 1.        , 0.66548514],\n",
       "       [0.35129741, 0.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 0.25      , 0.5       , 0.84324928],\n",
       "       [0.46706587, 1.        , 0.        , 1.        , 0.5       ,\n",
       "        1.        , 0.        , 0.        , 0.65439054],\n",
       "       [0.35129741, 1.        , 0.        , 1.        , 0.        ,\n",
       "        0.        , 0.5       , 0.        , 0.72536137]])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Niormalizing Data - MinMaxScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "norm = MinMaxScaler().fit(X_train)\n",
    "\n",
    "# transform training data\n",
    "X_train_norm = norm.transform(X_train)\n",
    "\n",
    "# transform testing dataabs\n",
    "X_test_norm = norm.transform(X_test)\n",
    "\n",
    "X_train_norm[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Applying SVM in SKLearn can be done in two ways - \n",
    "    1. If simple SVC (Linear Classifier) is to be applied, use `sklearn.svm.LinearSVC()`\n",
    "    2. If kernels are to be used, i.e. SVM, then use `sklearn.svm.SVC()`\n",
    "    3. Another algorithm in the SVM family is the `sklearn.svm.NuSVC()` with hyperparameter `nu`\n",
    "    \n",
    "- Note that Linear SVC can also be applied using `SVC()` but use `LinearSVC()` because - \n",
    "    1. it has more flexibility in the choice of penalties (no penalty for Linear SVC in `SVC()`) and loss functions and \n",
    "    2. it should scale better to large numbers of samples."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 1. `LinearSVC()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearSVC(random_state=42)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Linear SVC\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "clf_lin_svc = LinearSVC(random_state=42)\n",
    "clf_lin_svc.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7718866517524235"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_lin_svc.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7576548364648573"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_lin_svc.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 2. `SVC(kernel = 'linear')`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(kernel='linear', random_state=42)"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM Linear\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_svc_lin = SVC(random_state=42, kernel='linear')\n",
    "clf_svc_lin.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7517524235645041"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_lin.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7480862908837856"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_lin.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 3. `SVC(kernel = 'poly')`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(kernel='poly', random_state=42)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM Poly\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_svc_poly = SVC(random_state=42, kernel='poly')\n",
    "clf_svc_poly.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7746457867263237"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_poly.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7559151009046625"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_poly.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 4. `SVC(kernel = 'rbf')`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(random_state=42)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM Radial Basis Function\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_svc_rbf = SVC(random_state=42, kernel='rbf')\n",
    "clf_svc_rbf.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7802386278896346"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_rbf.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7604384133611691"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_rbf.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 5. `SVC(kernel = 'sigmoid')`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(kernel='sigmoid', random_state=42)"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM Sigmoid (not used since we already have Logistic Regression)\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_svc_sigmoid = SVC(random_state=42, kernel='sigmoid')\n",
    "clf_svc_sigmoid.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5866517524235645"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_sigmoid.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5882045929018789"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svc_sigmoid.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### 6. Hyperparameter Tuning for SVM - `C` and `gamma` for RBF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 5 candidates, totalling 10 fits\n",
      "[CV 1/2] END .................C=0.1, kernel=rbf;, score=0.754 total time=   4.7s\n",
      "[CV 2/2] END .................C=0.1, kernel=rbf;, score=0.752 total time=   4.1s\n",
      "[CV 1/2] END ...................C=1, kernel=rbf;, score=0.766 total time=   4.2s\n",
      "[CV 2/2] END ...................C=1, kernel=rbf;, score=0.770 total time=   4.1s\n",
      "[CV 1/2] END ..................C=10, kernel=rbf;, score=0.765 total time=   4.7s\n",
      "[CV 2/2] END ..................C=10, kernel=rbf;, score=0.770 total time=   4.9s\n",
      "[CV 1/2] END .................C=100, kernel=rbf;, score=0.759 total time=   8.7s\n",
      "[CV 2/2] END .................C=100, kernel=rbf;, score=0.763 total time=   8.2s\n",
      "[CV 1/2] END ................C=1000, kernel=rbf;, score=0.753 total time=  31.2s\n",
      "[CV 2/2] END ................C=1000, kernel=rbf;, score=0.759 total time=  29.1s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=2, estimator=SVC(random_state=42),\n",
       "             param_grid={'C': [0.1, 1, 10, 100, 1000], 'kernel': ['rbf']},\n",
       "             scoring='accuracy', verbose=3)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "params = {'C' : [0.1, 1, 10, 100, 1000],\n",
    "          'kernel' : ['rbf']\n",
    "          # 'degree' : [2, 3, 4, 5, 10],    # only used for poly kernel\n",
    "          # 'gamma' : [1, 0.1, 0.01, 0.001, 0.0001]      # scaling for poly, rbf and sigmoid\n",
    "         }\n",
    "\n",
    "clf_svm_cv = GridSearchCV(estimator = SVC(random_state=42),\n",
    "                                param_grid = params,\n",
    "                                scoring = 'accuracy',\n",
    "                                verbose = 3,\n",
    "                                cv = 2\n",
    "                               )\n",
    "\n",
    "clf_svm_cv.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 1, 'kernel': 'rbf'}"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 5 candidates, totalling 10 fits\n",
      "[CV 1/2] END ..........C=1, gamma=1, kernel=rbf;, score=0.764 total time=   4.3s\n",
      "[CV 2/2] END ..........C=1, gamma=1, kernel=rbf;, score=0.770 total time=   4.4s\n",
      "[CV 1/2] END ........C=1, gamma=0.1, kernel=rbf;, score=0.757 total time=   3.9s\n",
      "[CV 2/2] END ........C=1, gamma=0.1, kernel=rbf;, score=0.752 total time=   4.2s\n",
      "[CV 1/2] END .......C=1, gamma=0.01, kernel=rbf;, score=0.752 total time=   4.0s\n",
      "[CV 2/2] END .......C=1, gamma=0.01, kernel=rbf;, score=0.752 total time=   4.2s\n",
      "[CV 1/2] END ......C=1, gamma=0.001, kernel=rbf;, score=0.752 total time=   3.8s\n",
      "[CV 2/2] END ......C=1, gamma=0.001, kernel=rbf;, score=0.752 total time=   3.7s\n",
      "[CV 1/2] END .....C=1, gamma=0.0001, kernel=rbf;, score=0.752 total time=   3.9s\n",
      "[CV 2/2] END .....C=1, gamma=0.0001, kernel=rbf;, score=0.752 total time=   3.7s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=2, estimator=SVC(random_state=42),\n",
       "             param_grid={'C': [1], 'gamma': [1, 0.1, 0.01, 0.001, 0.0001],\n",
       "                         'kernel': ['rbf']},\n",
       "             scoring='accuracy', verbose=3)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "params = {'C' : [1],\n",
    "          'kernel' : ['rbf'],\n",
    "          'gamma' : [1, 0.1, 0.01, 0.001, 0.0001]      # scaling for poly, rbf and sigmoid\n",
    "         }\n",
    "\n",
    "clf_svm_cv = GridSearchCV(estimator = SVC(random_state=42),\n",
    "                                param_grid = params,\n",
    "                                scoring = 'accuracy',\n",
    "                                verbose = 3,\n",
    "                                cv = 2\n",
    "                               )\n",
    "\n",
    "clf_svm_cv.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'C': 1, 'gamma': 1, 'kernel': 'rbf'}"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7819537658463833"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm_cv.score(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.761134307585247"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_svm_cv.score(X_test_norm, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "## 10. Dimensionality Reduction Techniques"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.35129741, 0.        , 1.        , 0.        , 1.        ,\n",
       "        1.        , 0.5       , 0.5       , 0.63414285],\n",
       "       [0.94211577, 1.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 1.        , 1.        , 0.66548514],\n",
       "       [0.35129741, 0.        , 1.        , 0.        , 0.5       ,\n",
       "        1.        , 0.25      , 0.5       , 0.84324928],\n",
       "       [0.46706587, 1.        , 0.        , 1.        , 0.5       ,\n",
       "        1.        , 0.        , 0.        , 0.65439054],\n",
       "       [0.35129741, 1.        , 0.        , 1.        , 0.        ,\n",
       "        0.        , 0.5       , 0.        , 0.72536137]])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Niormalizing Data - MinMaxScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "norm = MinMaxScaler().fit(X_train)\n",
    "\n",
    "# transform training data\n",
    "X_train_norm = norm.transform(X_train)\n",
    "\n",
    "# transform testing dataabs\n",
    "X_test_norm = norm.transform(X_test)\n",
    "\n",
    "X_train_norm[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>PC1</th>\n",
       "      <th>PC2</th>\n",
       "      <th>PC3</th>\n",
       "      <th>PC4</th>\n",
       "      <th>PC5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.220554</td>\n",
       "      <td>0.847956</td>\n",
       "      <td>0.225623</td>\n",
       "      <td>-0.093747</td>\n",
       "      <td>-0.296108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.692023</td>\n",
       "      <td>-0.219279</td>\n",
       "      <td>-0.007677</td>\n",
       "      <td>0.009308</td>\n",
       "      <td>0.261753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.011803</td>\n",
       "      <td>0.735750</td>\n",
       "      <td>0.317277</td>\n",
       "      <td>0.021734</td>\n",
       "      <td>-0.468813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.980493</td>\n",
       "      <td>-0.078406</td>\n",
       "      <td>-0.918492</td>\n",
       "      <td>-0.230819</td>\n",
       "      <td>-0.405041</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.359308</td>\n",
       "      <td>-0.654462</td>\n",
       "      <td>-0.181374</td>\n",
       "      <td>0.139470</td>\n",
       "      <td>-0.078288</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        PC1       PC2       PC3       PC4       PC5\n",
       "0 -0.220554  0.847956  0.225623 -0.093747 -0.296108\n",
       "1 -0.692023 -0.219279 -0.007677  0.009308  0.261753\n",
       "2 -0.011803  0.735750  0.317277  0.021734 -0.468813\n",
       "3  0.980493 -0.078406 -0.918492 -0.230819 -0.405041\n",
       "4  1.359308 -0.654462 -0.181374  0.139470 -0.078288"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "\n",
    "pca = PCA(n_components=5)\n",
    "pca = pca.fit(X_train_norm)\n",
    "pca_data = pca.transform(X_train_norm)\n",
    "\n",
    "pca_df = pd.DataFrame(pca_data, columns = ['PC1', 'PC2', 'PC3', 'PC4', 'PC5'])\n",
    "pca_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.30407464, 0.17479721, 0.15521218, 0.0993118 , 0.0924038 ])"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scree plot\n",
    "plt.figure(figsize = (8,6))\n",
    "sns.lineplot(x = np.arange(pca.n_components_) + 1, y = pca.explained_variance_ratio_, marker = 'o', color = 'red')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Let's now try to build a model on this data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(random_state=42)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM w/ Radial Basis Function\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_pca_svc_rbf = SVC(random_state=42, kernel='rbf')\n",
    "clf_pca_svc_rbf.fit(pca_df, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7560775540641312"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_pca_svc_rbf.score(pca_df, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# LDA\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis \n",
    "\n",
    "lda = LinearDiscriminantAnalysis()\n",
    "lda = lda.fit(X=X_train_norm, y = y_train)\n",
    "lda_data = lda.transform(X_train_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(-0.1, 0.4)]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,2))\n",
    "fig = sns.scatterplot(x = lda_data.reshape([13410]), y = np.zeros_like(lda_data.reshape([13410])), hue = y_train, \n",
    "                s = 60, palette = 'inferno', alpha = 0.5, linewidth = 0.1)\n",
    "sns.despine(left = True)\n",
    "fig.set(yticklabels=[], ylabel = None, xlabel = 'LD1')\n",
    "fig.tick_params(left = False)\n",
    "fig.set(ylim=(-0.1,0.4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(random_state=42)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVM w/ Radial Basis Function\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "clf_lda_svc_rbf = SVC(random_state=42, kernel='rbf')\n",
    "clf_lda_svc_rbf.fit(lda_data, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7762117822520507"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_lda_svc_rbf.score(lda_data, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "## 11. Using Improtant Features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- In Random Forests' Feature Importance Plot, we have al;ready seen that the two features - `city_development_index` and `log_training_hours` are the most important ones. We will try to analyse these features and see if any improvement in separability can be obtained."
   ]
  },
  {
   "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>city_development_index</th>\n",
       "      <th>gender</th>\n",
       "      <th>relevent_experience</th>\n",
       "      <th>enrolled_university</th>\n",
       "      <th>education_level</th>\n",
       "      <th>major_discipline</th>\n",
       "      <th>experience</th>\n",
       "      <th>last_new_job</th>\n",
       "      <th>target</th>\n",
       "      <th>log_training_hours</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.583519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.850148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.418841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.951244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.079442</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index  gender  relevent_experience  enrolled_university  \\\n",
       "0                   0.920       1                    1                    0   \n",
       "1                   0.776       1                    0                    0   \n",
       "2                   0.624       0                    0                    1   \n",
       "3                   0.789       0                    0                    0   \n",
       "4                   0.767       1                    1                    0   \n",
       "\n",
       "   education_level  major_discipline  experience  last_new_job  target  \\\n",
       "0                1                 1           3             1     1.0   \n",
       "1                1                 1           4             2     0.0   \n",
       "2                1                 1           1             0     0.0   \n",
       "3                1                 0           4             0     1.0   \n",
       "4                2                 1           3             2     0.0   \n",
       "\n",
       "   log_training_hours  \n",
       "0            3.583519  \n",
       "1            3.850148  \n",
       "2            4.418841  \n",
       "3            3.951244  \n",
       "4            2.079442  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>city_development_index</th>\n",
       "      <th>log_training_hours</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.920</td>\n",
       "      <td>3.583519</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.776</td>\n",
       "      <td>3.850148</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.624</td>\n",
       "      <td>4.418841</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.789</td>\n",
       "      <td>3.951244</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.767</td>\n",
       "      <td>2.079442</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   city_development_index  log_training_hours  target\n",
       "0                   0.920            3.583519     1.0\n",
       "1                   0.776            3.850148     0.0\n",
       "2                   0.624            4.418841     0.0\n",
       "3                   0.789            3.951244     1.0\n",
       "4                   0.767            2.079442     0.0"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_reduced = df[['city_development_index', 'log_training_hours', 'target']]\n",
    "df_reduced.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train = (13410, 2)\n",
      "X_test = (5748, 2)\n",
      "y_train = (13410,)\n",
      "y_test = (5748,)\n"
     ]
    }
   ],
   "source": [
    "# train test split\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X = df_reduced.drop('target', axis = 1)\n",
    "y = df_reduced['target']\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)\n",
    "\n",
    "print('X_train =', X_train.shape)\n",
    "print('X_test =', X_test.shape)\n",
    "print('y_train =', y_train.shape)\n",
    "print('y_test =', y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='city_development_index', ylabel='log_training_hours'>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (12,8))\n",
    "sns.scatterplot(x = df['city_development_index'], y = np.exp(df['log_training_hours']), hue = df['target'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.35129741, 0.63414285],\n",
       "       [0.94211577, 0.66548514],\n",
       "       [0.35129741, 0.84324928],\n",
       "       [0.46706587, 0.65439054],\n",
       "       [0.35129741, 0.72536137]])"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Niormalizing Data - MinMaxScaler\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "norm = MinMaxScaler().fit(X_train)\n",
    "\n",
    "# transform training data\n",
    "X_train_norm = norm.transform(X_train)\n",
    "\n",
    "# transform testing dataabs\n",
    "X_test_norm = norm.transform(X_test)\n",
    "\n",
    "X_train_norm[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 3 candidates, totalling 6 fits\n",
      "[CV 1/2] END ................C=0.1, kernel=poly;, score=0.755 total time=  35.9s\n",
      "[CV 2/2] END ................C=0.1, kernel=poly;, score=0.752 total time= 1.3min\n",
      "[CV 1/2] END ..................C=1, kernel=poly;, score=0.754 total time= 2.0min\n",
      "[CV 2/2] END ..................C=1, kernel=poly;, score=0.752 total time= 2.9min\n",
      "[CV 1/2] END .................C=10, kernel=poly;, score=0.755 total time= 3.3min\n",
      "[CV 2/2] END .................C=10, kernel=poly;, score=0.752 total time= 3.2min\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=2, estimator=SVC(random_state=42),\n",
       "             param_grid={'C': [0.1, 1, 10], 'kernel': ['poly']},\n",
       "             scoring='accuracy', verbose=3)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Poly SVM with different degrees\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "params = {'C' : [0.1, 1, 10],\n",
    "          'kernel' : ['poly'],\n",
    "          # 'degree' : [2, 3, 4, 5, 10]    # only used for poly kernel\n",
    "         }\n",
    "\n",
    "clf_svm_cv = GridSearchCV(estimator = SVC(random_state=42),\n",
    "                                param_grid = params,\n",
    "                                scoring = 'accuracy',\n",
    "                                verbose = 3,\n",
    "                                cv = 2\n",
    "                               )\n",
    "\n",
    "clf_svm_cv.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- `C` doesn't have much effect so we will use default value i.e. `C = 1`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 5 candidates, totalling 10 fits\n",
      "[CV 1/2] END .............degree=2, kernel=poly;, score=0.752 total time=  13.1s\n",
      "[CV 2/2] END .............degree=2, kernel=poly;, score=0.752 total time=   6.5s\n",
      "[CV 1/2] END .............degree=3, kernel=poly;, score=0.754 total time= 2.0min\n",
      "[CV 2/2] END .............degree=3, kernel=poly;, score=0.752 total time= 2.8min\n",
      "[CV 1/2] END .............degree=4, kernel=poly;, score=0.755 total time= 5.4min\n",
      "[CV 2/2] END .............degree=4, kernel=poly;, score=0.754 total time= 5.3min\n",
      "[CV 1/2] END .............degree=5, kernel=poly;, score=0.755 total time= 8.1min\n",
      "[CV 2/2] END .............degree=5, kernel=poly;, score=0.754 total time= 5.0min\n",
      "[CV 1/2] END ............degree=10, kernel=poly;, score=0.779 total time=47.6min\n",
      "[CV 2/2] END ............degree=10, kernel=poly;, score=0.770 total time=24.7min\n"
     ]
    }
   ],
   "source": [
    "# Poly SVM with different degrees\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "params = {'kernel' : ['poly'],\n",
    "          'degree' : [2, 3, 4, 5, 10]    # only used for poly kernel\n",
    "         }\n",
    "\n",
    "clf_svm_cv = GridSearchCV(estimator = SVC(random_state=42),\n",
    "                                param_grid = params,\n",
    "                                scoring = 'accuracy',\n",
    "                                verbose = 3,\n",
    "                                cv = 2\n",
    "                               )\n",
    "\n",
    "clf_svm_cv.fit(X_train_norm, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- As we can see, higher order SVM polynomial kernel is not improving the accuracy to any significant extent.\n",
    "- Although, 10th order poly kernel did show an ~2.5-3 % improvement in the performance, the time it took to run those models is not worth it unless we are using a distributed system or GPUs."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# THE END"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}