Costa_Rican_Households_Classification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append('/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import datasets\n",
    "df_train = pd.read_csv('train.csv')\n",
    "df_test = pd.read_csv('test.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test_or = pd.read_csv('test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "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>Id</th>\n",
       "      <th>v2a1</th>\n",
       "      <th>hacdor</th>\n",
       "      <th>rooms</th>\n",
       "      <th>hacapo</th>\n",
       "      <th>v14a</th>\n",
       "      <th>refrig</th>\n",
       "      <th>v18q</th>\n",
       "      <th>v18q1</th>\n",
       "      <th>r4h1</th>\n",
       "      <th>...</th>\n",
       "      <th>SQBescolari</th>\n",
       "      <th>SQBage</th>\n",
       "      <th>SQBhogar_total</th>\n",
       "      <th>SQBedjefe</th>\n",
       "      <th>SQBhogar_nin</th>\n",
       "      <th>SQBovercrowding</th>\n",
       "      <th>SQBdependency</th>\n",
       "      <th>SQBmeaned</th>\n",
       "      <th>agesq</th>\n",
       "      <th>Target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_279628684</td>\n",
       "      <td>190000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>100</td>\n",
       "      <td>1849</td>\n",
       "      <td>1</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1849</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_f29eb3ddd</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>144</td>\n",
       "      <td>4489</td>\n",
       "      <td>1</td>\n",
       "      <td>144</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>64.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>4489</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_68de51c94</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>121</td>\n",
       "      <td>8464</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>64.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>8464</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ID_d671db89c</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>81</td>\n",
       "      <td>289</td>\n",
       "      <td>16</td>\n",
       "      <td>121</td>\n",
       "      <td>4</td>\n",
       "      <td>1.777778</td>\n",
       "      <td>1.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>289</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ID_d56d6f5f5</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>121</td>\n",
       "      <td>1369</td>\n",
       "      <td>16</td>\n",
       "      <td>121</td>\n",
       "      <td>4</td>\n",
       "      <td>1.777778</td>\n",
       "      <td>1.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>1369</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 143 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id      v2a1  hacdor  rooms  hacapo  v14a  refrig  v18q  v18q1  \\\n",
       "0  ID_279628684  190000.0       0      3       0     1       1     0    NaN   \n",
       "1  ID_f29eb3ddd  135000.0       0      4       0     1       1     1    1.0   \n",
       "2  ID_68de51c94       NaN       0      8       0     1       1     0    NaN   \n",
       "3  ID_d671db89c  180000.0       0      5       0     1       1     1    1.0   \n",
       "4  ID_d56d6f5f5  180000.0       0      5       0     1       1     1    1.0   \n",
       "\n",
       "   r4h1   ...    SQBescolari  SQBage  SQBhogar_total  SQBedjefe  SQBhogar_nin  \\\n",
       "0     0   ...            100    1849               1        100             0   \n",
       "1     0   ...            144    4489               1        144             0   \n",
       "2     0   ...            121    8464               1          0             0   \n",
       "3     0   ...             81     289              16        121             4   \n",
       "4     0   ...            121    1369              16        121             4   \n",
       "\n",
       "   SQBovercrowding  SQBdependency  SQBmeaned  agesq  Target  \n",
       "0         1.000000            0.0      100.0   1849       4  \n",
       "1         1.000000           64.0      144.0   4489       4  \n",
       "2         0.250000           64.0      121.0   8464       4  \n",
       "3         1.777778            1.0      121.0    289       4  \n",
       "4         1.777778            1.0      121.0   1369       4  \n",
       "\n",
       "[5 rows x 143 columns]"
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "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>Id</th>\n",
       "      <th>v2a1</th>\n",
       "      <th>hacdor</th>\n",
       "      <th>rooms</th>\n",
       "      <th>hacapo</th>\n",
       "      <th>v14a</th>\n",
       "      <th>refrig</th>\n",
       "      <th>v18q</th>\n",
       "      <th>v18q1</th>\n",
       "      <th>r4h1</th>\n",
       "      <th>...</th>\n",
       "      <th>age</th>\n",
       "      <th>SQBescolari</th>\n",
       "      <th>SQBage</th>\n",
       "      <th>SQBhogar_total</th>\n",
       "      <th>SQBedjefe</th>\n",
       "      <th>SQBhogar_nin</th>\n",
       "      <th>SQBovercrowding</th>\n",
       "      <th>SQBdependency</th>\n",
       "      <th>SQBmeaned</th>\n",
       "      <th>agesq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_2f6873615</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>272.25</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_1c78846d2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>256</td>\n",
       "      <td>1681</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>272.25</td>\n",
       "      <td>1681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_e5442cf6a</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>41</td>\n",
       "      <td>289</td>\n",
       "      <td>1681</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>272.25</td>\n",
       "      <td>1681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ID_a8db26a79</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>59</td>\n",
       "      <td>256</td>\n",
       "      <td>3481</td>\n",
       "      <td>1</td>\n",
       "      <td>256</td>\n",
       "      <td>0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>256.00</td>\n",
       "      <td>3481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ID_a62966799</td>\n",
       "      <td>175000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>18</td>\n",
       "      <td>121</td>\n",
       "      <td>324</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.25</td>\n",
       "      <td>64.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>324</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 142 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id      v2a1  hacdor  rooms  hacapo  v14a  refrig  v18q  v18q1  \\\n",
       "0  ID_2f6873615       NaN       0      5       0     1       1     0    NaN   \n",
       "1  ID_1c78846d2       NaN       0      5       0     1       1     0    NaN   \n",
       "2  ID_e5442cf6a       NaN       0      5       0     1       1     0    NaN   \n",
       "3  ID_a8db26a79       NaN       0     14       0     1       1     1    1.0   \n",
       "4  ID_a62966799  175000.0       0      4       0     1       1     1    1.0   \n",
       "\n",
       "   r4h1  ...    age  SQBescolari  SQBage  SQBhogar_total  SQBedjefe  \\\n",
       "0     1  ...      4            0      16               9          0   \n",
       "1     1  ...     41          256    1681               9          0   \n",
       "2     1  ...     41          289    1681               9          0   \n",
       "3     0  ...     59          256    3481               1        256   \n",
       "4     0  ...     18          121     324               1          0   \n",
       "\n",
       "   SQBhogar_nin  SQBovercrowding  SQBdependency  SQBmeaned  agesq  \n",
       "0             1             2.25           0.25     272.25     16  \n",
       "1             1             2.25           0.25     272.25   1681  \n",
       "2             1             2.25           0.25     272.25   1681  \n",
       "3             0             1.00           0.00     256.00   3481  \n",
       "4             1             0.25          64.00        NaN    324  \n",
       "\n",
       "[5 rows x 142 columns]"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 9557 entries, 0 to 9556\n",
      "Columns: 143 entries, Id to Target\n",
      "dtypes: float64(8), int64(130), object(5)\n",
      "memory usage: 10.4+ MB\n"
     ]
    }
   ],
   "source": [
    "df_train.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory Data Analysis & Data Prep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "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": [
    "#Exploring the distribution of the target variable\n",
    "sns.countplot(x='Target', data = df_train)\n",
    "labels = ['extreme poverty','moderate poverty','vulnerable','non vulnerable']\n",
    "plt.xticks(ticks = [0,1,2,3], labels = labels, rotation = 'vertical');\n",
    "\n",
    "#Unequal distribution of target classes, to be address by balancing class weights in the models used eventually"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "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>Id</th>\n",
       "      <th>idhogar</th>\n",
       "      <th>dependency</th>\n",
       "      <th>edjefe</th>\n",
       "      <th>edjefa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_279628684</td>\n",
       "      <td>21eb7fcc1</td>\n",
       "      <td>no</td>\n",
       "      <td>10</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_f29eb3ddd</td>\n",
       "      <td>0e5d7a658</td>\n",
       "      <td>8</td>\n",
       "      <td>12</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_68de51c94</td>\n",
       "      <td>2c7317ea8</td>\n",
       "      <td>8</td>\n",
       "      <td>no</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ID_d671db89c</td>\n",
       "      <td>2b58d945f</td>\n",
       "      <td>yes</td>\n",
       "      <td>11</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ID_d56d6f5f5</td>\n",
       "      <td>2b58d945f</td>\n",
       "      <td>yes</td>\n",
       "      <td>11</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ID_ec05b1a7b</td>\n",
       "      <td>2b58d945f</td>\n",
       "      <td>yes</td>\n",
       "      <td>11</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ID_e9e0c1100</td>\n",
       "      <td>2b58d945f</td>\n",
       "      <td>yes</td>\n",
       "      <td>11</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ID_3e04e571e</td>\n",
       "      <td>d6dae86b7</td>\n",
       "      <td>yes</td>\n",
       "      <td>9</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ID_1284f8aad</td>\n",
       "      <td>d6dae86b7</td>\n",
       "      <td>yes</td>\n",
       "      <td>9</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ID_51f52fdd2</td>\n",
       "      <td>d6dae86b7</td>\n",
       "      <td>yes</td>\n",
       "      <td>9</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ID_db44f5c59</td>\n",
       "      <td>d6dae86b7</td>\n",
       "      <td>yes</td>\n",
       "      <td>9</td>\n",
       "      <td>no</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>ID_de822510c</td>\n",
       "      <td>bb2094100</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>ID_d94071d7c</td>\n",
       "      <td>bb2094100</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>ID_064b57869</td>\n",
       "      <td>c51f9c774</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ID_5c837d8a4</td>\n",
       "      <td>c51f9c774</td>\n",
       "      <td>yes</td>\n",
       "      <td>no</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Id    idhogar dependency edjefe edjefa\n",
       "0   ID_279628684  21eb7fcc1         no     10     no\n",
       "1   ID_f29eb3ddd  0e5d7a658          8     12     no\n",
       "2   ID_68de51c94  2c7317ea8          8     no     11\n",
       "3   ID_d671db89c  2b58d945f        yes     11     no\n",
       "4   ID_d56d6f5f5  2b58d945f        yes     11     no\n",
       "5   ID_ec05b1a7b  2b58d945f        yes     11     no\n",
       "6   ID_e9e0c1100  2b58d945f        yes     11     no\n",
       "7   ID_3e04e571e  d6dae86b7        yes      9     no\n",
       "8   ID_1284f8aad  d6dae86b7        yes      9     no\n",
       "9   ID_51f52fdd2  d6dae86b7        yes      9     no\n",
       "10  ID_db44f5c59  d6dae86b7        yes      9     no\n",
       "11  ID_de822510c  bb2094100        yes     no     11\n",
       "12  ID_d94071d7c  bb2094100        yes     no     11\n",
       "13  ID_064b57869  c51f9c774        yes     no      4\n",
       "14  ID_5c837d8a4  c51f9c774        yes     no      4"
      ]
     },
     "execution_count": 164,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Exploring data stored as object type\n",
    "objects = df_train.select_dtypes('object')\n",
    "objects.head(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two of the features are IDs. \n",
    "\n",
    "Dependency is the ratio of number of dependents to the number of individuals in the working age group and is calculated. \n",
    "\n",
    "edjefe is the years of education of male head of household\n",
    "\n",
    "edjefa is the years of education of female head of household\n",
    "\n",
    "edjefe and edjefa are present for different observations depending on whether the head of household is male or female. \n",
    "\n",
    "In these columns yes is replacing the value 1, and no is replacing the value 0. We replace these values back.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Check if dependency value is really corresponding to 0 and 1\n",
    "df_train['check'] = (df_train['hogar_nin'] + df_train['hogar_mayor'])/(df_train['hogar_total'] - (df_train['hogar_nin'] + df_train['hogar_mayor']))\n",
    "df_train.loc[:,['Id', 'check', 'dependency']]\n",
    "\n",
    "#Yes it is, replace values\n",
    "df_train.drop('check', axis = 1, inplace = True)\n",
    "\n",
    "df_train['dependency'].replace({'yes': 1, 'no': 0}, inplace = True)\n",
    "\n",
    "#Replace nos\n",
    "df_train['edjefe'].replace({'no': 0}, inplace = True)\n",
    "df_train['edjefa'].replace({'no': 0}, inplace = True)\n",
    "\n",
    "#Replace years of education for heads of household, with the figure on education\n",
    "df_train.loc[(df_train['edjefa'] == \"yes\"), \"edjefa\"] = df_train.loc[(df_train['edjefa'] == \"yes\"), \"escolari\"]\n",
    "df_train.loc[(df_train['edjefe'] == \"yes\"), \"edjefe\"] = df_train.loc[(df_train['edjefe'] == \"yes\"), \"escolari\"]\n",
    "\n",
    "#Change object type to float\n",
    "df_train['dependency'] = df_train['dependency'].astype('float64')\n",
    "df_train['edjefe'] = df_train['edjefe'].astype('float64')\n",
    "df_train['edjefa'] = df_train['edjefa'].astype('float64')\n",
    "\n",
    "#Create one feature for education of head of household\n",
    "a = df_train[['edjefa','edjefe']]\n",
    "df_train['edjef'] = a.max(axis=1) \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "### REPEAT FOR TEST DATA ###\n",
    "\n",
    "#Replace dependency values\n",
    "df_test['dependency'].replace({'yes': 1, 'no': 0}, inplace = True)\n",
    "\n",
    "#Replace nos\n",
    "df_test['edjefe'].replace({'no': 0}, inplace = True)\n",
    "df_test['edjefa'].replace({'no': 0}, inplace = True)\n",
    "\n",
    "#Replace years of education for all, with the figure on years of schooling\n",
    "df_test.loc[(df_test['edjefa'] == \"yes\"), \"edjefa\"] = df_test.loc[(df_test['edjefa'] == \"yes\"), \"escolari\"]\n",
    "df_test.loc[(df_test['edjefe'] == \"yes\") , \"edjefe\"] = df_test.loc[(df_test['edjefe'] == \"yes\"), \"escolari\"]\n",
    "\n",
    "df_test['dependency'] = df_test['dependency'].astype('float64')\n",
    "df_test['edjefe'] = df_test['edjefe'].astype('float64')\n",
    "df_test['edjefa'] = df_test['edjefa'].astype('float64')\n",
    "\n",
    "#Create one feature for education of head of household\n",
    "df_test['edjef'] = np.max(df_test[['edjefa','edjefe']], axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [],
   "source": [
    "#If household has no water, assume no access to toilet\n",
    "df_train.loc[(df_train.v14a ==  1) & (df_train.sanitario1 ==  1) & (df_train.abastaguano == 0), \"v14a\"] = 0\n",
    "df_train.loc[(df_train.v14a ==  1) & (df_train.sanitario1 ==  1) & (df_train.abastaguano == 0), \"sanitario1\"] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [],
   "source": [
    "#If household has no water, assume no access to toilet\n",
    "df_test.loc[(df_test.v14a ==  1) & (df_test.sanitario1 ==  1) & (df_test.abastaguano == 0), \"v14a\"] = 0\n",
    "df_test.loc[(df_test.v14a ==  1) & (df_test.sanitario1 ==  1) & (df_test.abastaguano == 0), \"sanitario1\"] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {
    "scrolled": true
   },
   "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>Id</th>\n",
       "      <th>dependency</th>\n",
       "      <th>edjefe</th>\n",
       "      <th>edjefa</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_279628684</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_f29eb3ddd</td>\n",
       "      <td>8.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_68de51c94</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ID_d671db89c</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ID_d56d6f5f5</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id  dependency  edjefe  edjefa\n",
       "0  ID_279628684         0.0    10.0     0.0\n",
       "1  ID_f29eb3ddd         8.0    12.0     0.0\n",
       "2  ID_68de51c94         8.0     0.0    11.0\n",
       "3  ID_d671db89c         1.0    11.0     0.0\n",
       "4  ID_d56d6f5f5         1.0    11.0     0.0"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check for successful replacement\n",
    "df_train.loc[:,['Id', 'dependency', 'edjefe', 'edjefa']].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 9557 entries, 0 to 9556\n",
      "Columns: 144 entries, Id to edjef\n",
      "dtypes: float64(12), int64(130), object(2)\n",
      "memory usage: 10.5+ MB\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 23856 entries, 0 to 23855\n",
      "Columns: 143 entries, Id to edjef\n",
      "dtypes: float64(12), int64(129), object(2)\n",
      "memory usage: 26.0+ MB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(None, None)"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.info(), df_test.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Relationship btw number of people in the household, and the target\n",
    "plt.figure(figsize=(10,6))\n",
    "labels = ['extreme poverty','moderate poverty','vulnerable','non vulnerable']\n",
    "sns.boxplot(x='Target', y ='r4t3', data = df_train)\n",
    "locs, labels2 = plt.xticks()\n",
    "plt.xticks(ticks = locs, labels = labels, rotation = 'vertical');\n",
    "#No pattern\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Relationship btw average schooling of people in the household, and the target categories\n",
    "plt.figure(figsize=(10,6))\n",
    "labels = ['extreme poverty','moderate poverty','vulnerable','non vulnerable']\n",
    "sns.boxplot(x='Target', y ='meaneduc', data = df_train)\n",
    "locs, labels2 = plt.xticks()\n",
    "plt.xticks(ticks = locs, labels = labels, rotation = 'vertical');\n",
    "#More education seems linked to improved economic status\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "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>v2a1</th>\n",
       "      <th>v18q1</th>\n",
       "      <th>rez_esc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>190000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>135000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>180000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>180000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       v2a1  v18q1  rez_esc\n",
       "0  190000.0    NaN      NaN\n",
       "1  135000.0    1.0      NaN\n",
       "2       NaN    NaN      NaN\n",
       "3  180000.0    1.0      1.0\n",
       "4  180000.0    1.0      NaN"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check for features where > 70% data missing\n",
    "missing = df_train[[column for column in df_train if df_train[column].count()/len(df_train) <= 0.3]]\n",
    "missing.head()\n",
    "#These are Monthly rent payment(>70% missing), number of tablets a household owns (>70% missing), and years behind in school (>80% missing)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Plotting monthly rent payment\n",
    "plt.figure(figsize=(10,6))\n",
    "df_train.groupby('Target')['v2a1'].mean().plot('bar')\n",
    "labels = ['extreme poverty','moderate poverty','vulnerable','non vulnerable']\n",
    "plt.xticks(ticks = [0,1,2,3], labels = labels, rotation = 'vertical')\n",
    "\n",
    "#Since several households are not on rent, two options - use value 0 or impute values. 0 may skew relationship btw rent and target\n",
    "#Try both methods - first taking value 0 for those not on rent - who own a house for sure\n",
    "df_train.loc[((df_train.tipovivi1 == 1) & df_train['v2a1'].isnull()), 'v2a1'] = 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Replacing rent values in test data\n",
    "df_test.loc[((df_test.tipovivi1 == 1) & df_test['v2a1'].isnull()), 'v2a1'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "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>meaneduc</th>\n",
       "      <th>rez_esc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>meaneduc</th>\n",
       "      <td>1.00000</td>\n",
       "      <td>-0.09102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rez_esc</th>\n",
       "      <td>-0.09102</td>\n",
       "      <td>1.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          meaneduc  rez_esc\n",
       "meaneduc   1.00000 -0.09102\n",
       "rez_esc   -0.09102  1.00000"
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Fill 0 for Nan values in case of number of tablets owned by those who do not own a tablet and drop years behind in school\n",
    "df_train[df_train['v18q'] == 0]['v18q1']\n",
    "df_train.loc[((df_train['v18q'] == 0) & df_train['v18q1'].isnull()), 'v18q1'] = 0\n",
    "df_train.loc[(((df_train['age'] > 19) | (df_train['age'] < 7)) & df_train['rez_esc'].isnull()), 'rez_esc'] = 0\n",
    "df_train.loc[:,['meaneduc', 'rez_esc']].corr()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       0.0\n",
       "1       0.0\n",
       "2       0.0\n",
       "3       1.0\n",
       "4       0.0\n",
       "5       0.0\n",
       "6       0.0\n",
       "7       0.0\n",
       "8       0.0\n",
       "9       0.0\n",
       "10      1.0\n",
       "11      NaN\n",
       "12      0.0\n",
       "13      0.0\n",
       "14      0.0\n",
       "15      NaN\n",
       "16      0.0\n",
       "17      NaN\n",
       "18      0.0\n",
       "19      0.0\n",
       "20      0.0\n",
       "21      0.0\n",
       "22      0.0\n",
       "23      0.0\n",
       "24      0.0\n",
       "25      0.0\n",
       "26      0.0\n",
       "27      0.0\n",
       "28      0.0\n",
       "29      0.0\n",
       "       ... \n",
       "9527    0.0\n",
       "9528    0.0\n",
       "9529    0.0\n",
       "9530    0.0\n",
       "9531    0.0\n",
       "9532    0.0\n",
       "9533    0.0\n",
       "9534    0.0\n",
       "9535    0.0\n",
       "9536    1.0\n",
       "9537    0.0\n",
       "9538    0.0\n",
       "9539    0.0\n",
       "9540    0.0\n",
       "9541    0.0\n",
       "9542    0.0\n",
       "9543    0.0\n",
       "9544    0.0\n",
       "9545    0.0\n",
       "9546    0.0\n",
       "9547    0.0\n",
       "9548    NaN\n",
       "9549    0.0\n",
       "9550    0.0\n",
       "9551    0.0\n",
       "9552    0.0\n",
       "9553    0.0\n",
       "9554    0.0\n",
       "9555    0.0\n",
       "9556    0.0\n",
       "Name: rez_esc, Length: 9557, dtype: float64"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['rez_esc']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Replacing values in the test data\n",
    "df_test.loc[((df_test['v18q'] == 0) & df_test['v18q1'].isnull()), 'v18q1'] = 0\n",
    "df_test.loc[(((df_test['age'] > 19) | (df_test['age'] < 7)) & df_test['rez_esc'].isnull()), 'rez_esc'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "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>Id</th>\n",
       "      <th>v2a1</th>\n",
       "      <th>hacdor</th>\n",
       "      <th>rooms</th>\n",
       "      <th>hacapo</th>\n",
       "      <th>v14a</th>\n",
       "      <th>refrig</th>\n",
       "      <th>v18q</th>\n",
       "      <th>v18q1</th>\n",
       "      <th>r4h1</th>\n",
       "      <th>...</th>\n",
       "      <th>SQBage</th>\n",
       "      <th>SQBhogar_total</th>\n",
       "      <th>SQBedjefe</th>\n",
       "      <th>SQBhogar_nin</th>\n",
       "      <th>SQBovercrowding</th>\n",
       "      <th>SQBdependency</th>\n",
       "      <th>SQBmeaned</th>\n",
       "      <th>agesq</th>\n",
       "      <th>Target</th>\n",
       "      <th>edjef</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_279628684</td>\n",
       "      <td>190000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1849</td>\n",
       "      <td>1</td>\n",
       "      <td>100</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1849</td>\n",
       "      <td>4</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_f29eb3ddd</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>4489</td>\n",
       "      <td>1</td>\n",
       "      <td>144</td>\n",
       "      <td>0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>64.0</td>\n",
       "      <td>144.0</td>\n",
       "      <td>4489</td>\n",
       "      <td>4</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_68de51c94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>8464</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>64.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>8464</td>\n",
       "      <td>4</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ID_d671db89c</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>289</td>\n",
       "      <td>16</td>\n",
       "      <td>121</td>\n",
       "      <td>4</td>\n",
       "      <td>1.777778</td>\n",
       "      <td>1.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>289</td>\n",
       "      <td>4</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ID_d56d6f5f5</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1369</td>\n",
       "      <td>16</td>\n",
       "      <td>121</td>\n",
       "      <td>4</td>\n",
       "      <td>1.777778</td>\n",
       "      <td>1.0</td>\n",
       "      <td>121.0</td>\n",
       "      <td>1369</td>\n",
       "      <td>4</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 144 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id      v2a1  hacdor  rooms  hacapo  v14a  refrig  v18q  v18q1  \\\n",
       "0  ID_279628684  190000.0       0      3       0     1       1     0    0.0   \n",
       "1  ID_f29eb3ddd  135000.0       0      4       0     1       1     1    1.0   \n",
       "2  ID_68de51c94       0.0       0      8       0     1       1     0    0.0   \n",
       "3  ID_d671db89c  180000.0       0      5       0     1       1     1    1.0   \n",
       "4  ID_d56d6f5f5  180000.0       0      5       0     1       1     1    1.0   \n",
       "\n",
       "   r4h1  ...    SQBage  SQBhogar_total  SQBedjefe  SQBhogar_nin  \\\n",
       "0     0  ...      1849               1        100             0   \n",
       "1     0  ...      4489               1        144             0   \n",
       "2     0  ...      8464               1          0             0   \n",
       "3     0  ...       289              16        121             4   \n",
       "4     0  ...      1369              16        121             4   \n",
       "\n",
       "   SQBovercrowding  SQBdependency  SQBmeaned  agesq  Target  edjef  \n",
       "0         1.000000            0.0      100.0   1849       4   10.0  \n",
       "1         1.000000           64.0      144.0   4489       4   12.0  \n",
       "2         0.250000           64.0      121.0   8464       4   11.0  \n",
       "3         1.777778            1.0      121.0    289       4   11.0  \n",
       "4         1.777778            1.0      121.0   1369       4   11.0  \n",
       "\n",
       "[5 rows x 144 columns]"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating household level data - creating household level features for each individual,\n",
    "#and then subsetting data to only include heads of households\n",
    "\n",
    "#Household level per capita features\n",
    "df_train['tabs_per_cap'] = df_train['v18q1']/df_train['tamviv']\n",
    "df_train['bedrooms_per_cap'] = df_train['bedrooms']/df_train['tamviv']\n",
    "df_train['rooms_per_cap'] = df_train['rooms']/df_train['tamviv']\n",
    "df_train['mobiles_per_cap'] = df_train['qmobilephone']/df_train['tamviv']\n",
    "df_train['rent_per_cap'] = df_train['v2a1']/df_train['tamviv']\n",
    "\n",
    "# No toilet, no electricity, no floor, no water service, no kitchen\n",
    "df_train['risk'] = (df_train['sanitario1'] + \n",
    "                         df_train['noelec'] + \n",
    "                         df_train['pisonotiene'] + \n",
    "                         df_train['abastaguano'] + \n",
    "                         df_train['energcocinar1'])\n",
    "\n",
    "#Poorly constructed/unsafe home - walls, roof, floor quality all bad\n",
    "df_train['poor_construction'] = (df_train['epared1'] + \n",
    "                         df_train['etecho1'] + \n",
    "                         df_train['eviv1'])\n",
    "\n",
    "#Tech score - presence of tv, computer, phone, tablet, refrigerator - 1 point for each\n",
    "df_train['tech_score'] = (df_train['computer'] + \n",
    "                         df_train['television'] + \n",
    "                         df_train['mobilephone'] +\n",
    "                         df_train['v18q'] +\n",
    "                         df_train['refrig'])\n",
    "\n",
    "#Number of disabled persons in a household\n",
    "sum_disabled = pd.DataFrame(df_train.groupby(\"idhogar\")[\"dis\"].sum().reset_index(name='disabled_count'))\n",
    "sum_disabled.columns\n",
    "df_train = pd.merge(df_train, sum_disabled[['idhogar', 'disabled_count']], on= 'idhogar')\n",
    "\n",
    "#Number of widowed members in the household - death more likely in poorer families?\n",
    "sum_widowed = pd.DataFrame(df_train.groupby(\"idhogar\")[\"estadocivil6\"].sum().reset_index(name='widowed_count'))\n",
    "sum_widowed.columns\n",
    "df_train = pd.merge(df_train, sum_widowed[['idhogar', 'widowed_count']], on= 'idhogar')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Recreating features in test set\n",
    "#Creating household level data - creating household level features for each individual,\n",
    "#and then subsetting data to only include heads of households\n",
    "\n",
    "#Household level per capita features\n",
    "df_test['tabs_per_cap'] = df_test['v18q1']/df_test['tamviv']\n",
    "df_test['bedrooms_per_cap'] = df_test['bedrooms']/df_test['tamviv']\n",
    "df_test['rooms_per_cap'] = df_test['rooms']/df_test['tamviv']\n",
    "df_test['mobiles_per_cap'] = df_test['qmobilephone']/df_test['tamviv']\n",
    "df_test['rent_per_cap'] = df_test['v2a1']/df_test['tamviv']\n",
    "\n",
    "# No toilet, no electricity, no floor, no water service, no kitchen\n",
    "df_test['risk'] = (df_test['sanitario1'] + \n",
    "                         df_test['noelec'] + \n",
    "                         df_test['pisonotiene'] + \n",
    "                         df_test['abastaguano'] + \n",
    "                         df_test['energcocinar1'])\n",
    "\n",
    "#Poorly constructed/unsafe home - walls, roof, floor quality all bad\n",
    "df_test['poor_construction'] = (df_test['epared1'] + \n",
    "                         df_test['etecho1'] + \n",
    "                         df_test['eviv1'])\n",
    "\n",
    "#Tech score - presence of tv, computer, phone, tablet, refrigerator - 1 point for each\n",
    "df_test['tech_score'] = (df_test['computer'] + \n",
    "                         df_test['television'] + \n",
    "                         df_test['mobilephone'] +\n",
    "                         df_test['v18q'] +\n",
    "                         df_test['refrig'])\n",
    "\n",
    "#Number of disabled persons in a household\n",
    "sum_disabled = pd.DataFrame(df_test.groupby(\"idhogar\")[\"dis\"].sum().reset_index(name='disabled_count'))\n",
    "sum_disabled.columns\n",
    "df_test = pd.merge(df_test, sum_disabled[['idhogar', 'disabled_count']], on= 'idhogar')\n",
    "\n",
    "#Number of widowed members in the household - death more likely in poorer families?\n",
    "sum_widowed = pd.DataFrame(df_test.groupby(\"idhogar\")[\"estadocivil6\"].sum().reset_index(name='widowed_count'))\n",
    "sum_widowed.columns\n",
    "df_test = pd.merge(df_test, sum_widowed[['idhogar', 'widowed_count']], on= 'idhogar')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Impute missing values for the rest of the data if any\n",
    "df_train.interpolate(inplace=True)\n",
    "#Impute missing data in the test set\n",
    "df_test.interpolate(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "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>Id</th>\n",
       "      <th>v2a1</th>\n",
       "      <th>hacdor</th>\n",
       "      <th>rooms</th>\n",
       "      <th>hacapo</th>\n",
       "      <th>v14a</th>\n",
       "      <th>refrig</th>\n",
       "      <th>v18q</th>\n",
       "      <th>v18q1</th>\n",
       "      <th>r4h1</th>\n",
       "      <th>...</th>\n",
       "      <th>tabs_per_cap</th>\n",
       "      <th>bedrooms_per_cap</th>\n",
       "      <th>rooms_per_cap</th>\n",
       "      <th>mobiles_per_cap</th>\n",
       "      <th>rent_per_cap</th>\n",
       "      <th>risk</th>\n",
       "      <th>poor_construction</th>\n",
       "      <th>tech_score</th>\n",
       "      <th>disabled_count</th>\n",
       "      <th>widowed_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ID_279628684</td>\n",
       "      <td>190000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>3.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>190000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ID_f29eb3ddd</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>4.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ID_68de51c94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.00</td>\n",
       "      <td>8.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ID_ec05b1a7b</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.75</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.75</td>\n",
       "      <td>45000.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>ID_1284f8aad</td>\n",
       "      <td>130000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.25</td>\n",
       "      <td>32500.0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 154 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             Id      v2a1  hacdor  rooms  hacapo  v14a  refrig  v18q  v18q1  \\\n",
       "0  ID_279628684  190000.0       0      3       0     1       1     0    0.0   \n",
       "1  ID_f29eb3ddd  135000.0       0      4       0     1       1     1    1.0   \n",
       "2  ID_68de51c94       0.0       0      8       0     1       1     0    0.0   \n",
       "5  ID_ec05b1a7b  180000.0       0      5       0     1       1     1    1.0   \n",
       "8  ID_1284f8aad  130000.0       1      2       0     1       1     0    0.0   \n",
       "\n",
       "   r4h1      ...        tabs_per_cap  bedrooms_per_cap  rooms_per_cap  \\\n",
       "0     0      ...                0.00              1.00           3.00   \n",
       "1     0      ...                1.00              1.00           4.00   \n",
       "2     0      ...                0.00              2.00           8.00   \n",
       "5     0      ...                0.25              0.75           1.25   \n",
       "8     0      ...                0.00              0.25           0.50   \n",
       "\n",
       "   mobiles_per_cap  rent_per_cap  risk  poor_construction  tech_score  \\\n",
       "0             1.00      190000.0     0                  2           2   \n",
       "1             1.00      135000.0     0                  0           3   \n",
       "2             0.00           0.0     0                  0           1   \n",
       "5             0.75       45000.0     0                  0           3   \n",
       "8             0.25       32500.0     0                  2           2   \n",
       "\n",
       "   disabled_count  widowed_count  \n",
       "0               0              0  \n",
       "1               0              0  \n",
       "2               1              1  \n",
       "5               0              0  \n",
       "8               0              0  \n",
       "\n",
       "[5 rows x 154 columns]"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "heads_train = df_train.loc[df_train['parentesco1'] == 1,]\n",
    "heads_train.head()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "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>rez_esc</th>\n",
       "      <th>age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.0</td>\n",
       "      <td>79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.0</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.0</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.0</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.0</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.0</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.0</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.0</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.0</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>0.0</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>0.0</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>0.0</td>\n",
       "      <td>76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>0.0</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>0.0</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.0</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>0.0</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>0.0</td>\n",
       "      <td>71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>0.0</td>\n",
       "      <td>55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>0.0</td>\n",
       "      <td>71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0.0</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>0.0</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>0.0</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>0.0</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9477</th>\n",
       "      <td>0.0</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9480</th>\n",
       "      <td>0.0</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9481</th>\n",
       "      <td>0.0</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9485</th>\n",
       "      <td>0.0</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9487</th>\n",
       "      <td>0.0</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9490</th>\n",
       "      <td>0.0</td>\n",
       "      <td>77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9492</th>\n",
       "      <td>0.0</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9495</th>\n",
       "      <td>0.0</td>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9499</th>\n",
       "      <td>0.0</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9502</th>\n",
       "      <td>0.0</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9505</th>\n",
       "      <td>0.0</td>\n",
       "      <td>70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9506</th>\n",
       "      <td>0.0</td>\n",
       "      <td>82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9507</th>\n",
       "      <td>0.0</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9508</th>\n",
       "      <td>0.0</td>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9510</th>\n",
       "      <td>0.0</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9512</th>\n",
       "      <td>0.0</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9515</th>\n",
       "      <td>0.0</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9517</th>\n",
       "      <td>0.0</td>\n",
       "      <td>60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9519</th>\n",
       "      <td>0.0</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9523</th>\n",
       "      <td>0.0</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9526</th>\n",
       "      <td>0.0</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9528</th>\n",
       "      <td>0.0</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9530</th>\n",
       "      <td>0.0</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9531</th>\n",
       "      <td>0.0</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9534</th>\n",
       "      <td>0.0</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9535</th>\n",
       "      <td>0.0</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9541</th>\n",
       "      <td>0.0</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9545</th>\n",
       "      <td>0.0</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9551</th>\n",
       "      <td>0.0</td>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9552</th>\n",
       "      <td>0.0</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2973 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      rez_esc  age\n",
       "0         0.0   43\n",
       "1         0.0   67\n",
       "2         0.0   92\n",
       "5         0.0   38\n",
       "8         0.0   30\n",
       "12        0.0   34\n",
       "13        0.0   79\n",
       "18        0.0   50\n",
       "20        0.0   26\n",
       "21        0.0   69\n",
       "23        0.0   41\n",
       "26        0.0   44\n",
       "27        0.0   69\n",
       "30        0.0   35\n",
       "33        0.0   56\n",
       "39        0.0   37\n",
       "40        0.0   44\n",
       "43        0.0   76\n",
       "45        0.0   69\n",
       "49        0.0   48\n",
       "53        0.0   42\n",
       "56        0.0   37\n",
       "58        0.0   71\n",
       "59        0.0   30\n",
       "61        0.0   55\n",
       "64        0.0   71\n",
       "68        0.0   29\n",
       "72        0.0   45\n",
       "74        0.0   57\n",
       "77        0.0   41\n",
       "...       ...  ...\n",
       "9477      0.0   37\n",
       "9480      0.0   75\n",
       "9481      0.0   48\n",
       "9485      0.0   19\n",
       "9487      0.0   38\n",
       "9490      0.0   77\n",
       "9492      0.0   29\n",
       "9495      0.0   67\n",
       "9499      0.0   74\n",
       "9502      0.0   21\n",
       "9505      0.0   70\n",
       "9506      0.0   82\n",
       "9507      0.0   62\n",
       "9508      0.0   49\n",
       "9510      0.0   34\n",
       "9512      0.0   62\n",
       "9515      0.0   41\n",
       "9517      0.0   60\n",
       "9519      0.0   35\n",
       "9523      0.0   52\n",
       "9526      0.0   29\n",
       "9528      0.0   31\n",
       "9530      0.0   52\n",
       "9531      0.0   54\n",
       "9534      0.0   74\n",
       "9535      0.0   26\n",
       "9541      0.0   40\n",
       "9545      0.0   45\n",
       "9551      0.0   67\n",
       "9552      0.0   46\n",
       "\n",
       "[2973 rows x 2 columns]"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "null_columns=heads_train.columns[heads_train.isnull().any()]\n",
    "heads_train[null_columns].isnull().sum()\n",
    "heads_train[['rez_esc', 'age']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Training and Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "for column_name in heads_train.columns:\n",
    "    if heads_train[column_name].dtype == object:\n",
    "        heads_train[column_name] = le.fit_transform(heads_train[column_name])\n",
    "    else:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "for column_name in df_test.columns:\n",
    "    if df_test[column_name].dtype == object:\n",
    "        df_test[column_name] = le.fit_transform(df_test[column_name])\n",
    "    else:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2973 entries, 0 to 9552\n",
      "Columns: 154 entries, Id to widowed_count\n",
      "dtypes: float64(17), int64(137)\n",
      "memory usage: 3.5 MB\n"
     ]
    }
   ],
   "source": [
    "heads_train.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import tree\n",
    "from sklearn import model_selection\n",
    "from sklearn import metrics\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = heads_train['Target'].values\n",
    "X = heads_train.drop(labels = ['Target', 'idhogar'], axis = 1).values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y, test_size = 0.3, random_state = 11 )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2081, 152)"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(892, 152)"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2081,)"
      ]
     },
     "execution_count": 211,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(892,)"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [],
   "source": [
    "col_names = df_train.drop(labels = ['Target', 'idhogar'], axis = 1).columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(class_weight='balanced', criterion='entropy',\n",
       "            max_depth=10, max_features=None, max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, presort=False, random_state=7,\n",
       "            splitter='best')"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indiv_tree = tree.DecisionTreeClassifier(criterion = 'entropy', max_depth = 10, random_state= 7, class_weight= 'balanced')\n",
    "indiv_tree.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4786995515695067"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred = indiv_tree.predict(X_test)\n",
    "indiv_tree.score(X_train,y_train)\n",
    "indiv_tree.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.31732675305768965"
      ]
     },
     "execution_count": 227,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score\n",
    "score = f1_score(y_test, pred, average = 'macro')\n",
    "score\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Using cross validation to find appropriate hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import GridSearchCV, cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 150 candidates, totalling 750 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done 412 tasks      | elapsed:    1.9s\n",
      "[Parallel(n_jobs=-1)]: Done 750 out of 750 | elapsed:    3.3s finished\n",
      "/anaconda3/lib/python3.7/site-packages/sklearn/model_selection/_search.py:841: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.\n",
      "  DeprecationWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise-deprecating',\n",
       "       estimator=DecisionTreeClassifier(class_weight='balanced', criterion='entropy',\n",
       "            max_depth=10, max_features=None, max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, presort=False, random_state=7,\n",
       "            splitter='best'),\n",
       "       fit_params=None, iid='warn', n_jobs=-1,\n",
       "       param_grid={'max_depth': range(1, 11), 'max_features': range(4, 19)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring=None, verbose=True)"
      ]
     },
     "execution_count": 229,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_params = {'max_depth': range(1,11),\n",
    "               'max_features': range(4,19)}\n",
    "\n",
    "tree_grid = GridSearchCV(indiv_tree, tree_params,\n",
    "                         cv=5, n_jobs=-1, verbose=True)\n",
    "\n",
    "tree_grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 2, 'max_features': 12}"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5881787602114368"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tree_grid.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5852017937219731"
      ]
     },
     "execution_count": 232,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score\n",
    "accuracy_score(y_test, tree_grid.predict(X_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Improvement in performance of our decision tree by changing model hyperparams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/metrics/classification.py:1143: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.2817045867912712"
      ]
     },
     "execution_count": 233,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "indiv_tree2 = tree.DecisionTreeClassifier(criterion = 'entropy', max_depth = 2, max_features=12, random_state= 7, class_weight= 'balanced')\n",
    "indiv_tree2.fit(X_train, y_train)\n",
    "pred2 = indiv_tree2.predict(X_test)\n",
    "indiv_tree2.score(X_train,y_train)\n",
    "indiv_tree2.score(X_test,y_test)\n",
    "score2 = f1_score(y_test, pred2, average = 'macro')\n",
    "score2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.5881787602114368, 0.5852017937219731)"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Checking for overfitting\n",
    "indiv_tree2.score(X_train,y_train), indiv_tree2.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pydotplus #pip install pydotplus\n",
    "from sklearn.tree import export_graphviz\n",
    "\n",
    "def tree_graph_to_png(tree, feature_names, png_file_to_save):\n",
    "    tree_str = export_graphviz(tree, feature_names=feature_names, \n",
    "                                     filled=True, out_file=None)\n",
    "    graph = pydotplus.graph_from_dot_data(tree_str)  \n",
    "    graph.write_png(png_file_to_save)\n",
    "    \n",
    "tree_graph_to_png(tree=indiv_tree2, feature_names=col_names, png_file_to_save='decisiontree1.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import ensemble"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight='balanced',\n",
       "            criterion='entropy', max_depth=10, max_features='auto',\n",
       "            max_leaf_nodes=None, min_impurity_decrease=0.0,\n",
       "            min_impurity_split=None, min_samples_leaf=1,\n",
       "            min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "            n_estimators=500, n_jobs=None, oob_score=False, random_state=7,\n",
       "            verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = ensemble.RandomForestClassifier(n_estimators = 500, criterion = 'entropy', max_depth = 10, random_state = 7, class_weight= 'balanced')\n",
    "rf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6266816143497758"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_pred = rf.predict(X_test)\n",
    "rf.score(X_train, y_train) \n",
    "rf.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3925530647541115"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_score = f1_score(y_test, rf_pred, average = 'macro')\n",
    "rf_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Using cross-validation to find ideal hyperparams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 8 folds for each of 180 candidates, totalling 1440 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  42 tasks      | elapsed:   23.6s\n",
      "[Parallel(n_jobs=-1)]: Done 192 tasks      | elapsed:  2.0min\n",
      "[Parallel(n_jobs=-1)]: Done 442 tasks      | elapsed:  5.3min\n",
      "[Parallel(n_jobs=-1)]: Done 792 tasks      | elapsed: 11.1min\n",
      "[Parallel(n_jobs=-1)]: Done 1242 tasks      | elapsed: 19.0min\n",
      "[Parallel(n_jobs=-1)]: Done 1440 out of 1440 | elapsed: 23.0min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=8, error_score='raise-deprecating',\n",
       "       estimator=RandomForestClassifier(bootstrap=True, class_weight='balanced',\n",
       "            criterion='entropy', max_depth=10, max_features='auto',\n",
       "            max_leaf_nodes=None, min_impurity_decrease=0.0,\n",
       "            min_impurity_split=None, min_samples_leaf=1,\n",
       "            min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "            n_estimators=500, n_jobs=None, oob_score=False, random_state=7,\n",
       "            verbose=0, warm_start=False),\n",
       "       fit_params=None, iid='warn', n_jobs=-1,\n",
       "       param_grid={'max_depth': range(6, 18), 'max_features': range(4, 19)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring=None, verbose=True)"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_params = {'max_depth': range(6,18),\n",
    "               'max_features': range(4,19)}\n",
    "\n",
    "rf_grid = GridSearchCV(rf, rf_params,\n",
    "                         cv=8, n_jobs=-1, verbose=True)\n",
    "\n",
    "rf_grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 14, 'max_features': 14}"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_grid.best_params_#{'max_depth': 10, 'max_features': 16}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6900528592023066"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_grid.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3050905417250469"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf2 = ensemble.RandomForestClassifier(n_estimators = 500, criterion = 'entropy', max_depth = 14, max_features = 14, random_state = 7, class_weight= 'balanced')\n",
    "rf2.fit(X_train, y_train)\n",
    "rf_pred2 = rf2.predict(X_test)\n",
    "rf2.score(X_train,y_train)\n",
    "rf2.score(X_test,y_test)\n",
    "rf_score2 = f1_score(y_test, rf_pred2, average = 'macro')\n",
    "rf_score2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1.0, 0.6457399103139013)"
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Checking for performance on training and validation\n",
    "rf2.score(X_train,y_train), rf2.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoost Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xgb\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_mod = xgb.XGBClassifier(colsample_bytree = 0.3, learning_rate = 0.1,\n",
    "                max_depth = 6, alpha = 10, n_estimators = 500, class_weight='balanced')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBClassifier(alpha=10, base_score=0.5, booster='gbtree',\n",
       "       class_weight='balanced', colsample_bylevel=1, colsample_bytree=0.3,\n",
       "       gamma=0, learning_rate=0.1, max_delta_step=0, max_depth=6,\n",
       "       min_child_weight=1, missing=None, n_estimators=500, n_jobs=1,\n",
       "       nthread=None, objective='multi:softprob', random_state=0,\n",
       "       reg_alpha=0, reg_lambda=1, scale_pos_weight=1, seed=None,\n",
       "       silent=True, subsample=1)"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb_mod.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_pred = xgb_mod.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1.0, 0.6457399103139013)"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xgb_mod.score(X_train, y_train), xgb_mod.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6457399103139013"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, xgb_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3411967221038378"
      ]
     },
     "execution_count": 251,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Macro f1 score\n",
    "xgb_score = f1_score(y_test, xgb_pred, average = 'macro')\n",
    "xgb_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Creating Submission on Test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission = pd.DataFrame()\n",
    "submission['Id'] = df_test_or['Id']\n",
    "test_X = df_test.drop(labels = ['idhogar'], axis = 1).values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {},
   "outputs": [],
   "source": [
    "submission['Target'] = xgb_mod.predict(test_X)\n",
    "submission.to_csv('submission.csv', index = False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7",
   "language": "python",
   "name": "python3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}