Handling Missing Data.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "6aa82e9d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import missingno as msno\n",
    "import scipy.stats as stats\n",
    "\n",
    "from sklearn.impute import IterativeImputer, SimpleImputer,KNNImputer\n",
    "from sklearn.tree import DecisionTreeRegressor\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler,MinMaxScaler,LabelEncoder,OrdinalEncoder,OneHotEncoder\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.experimental import enable_iterative_imputer\n",
    "\n",
    "from sklearn.metrics import accuracy_score,classification_report,roc_curve, roc_auc_score,plot_confusion_matrix\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9674a4f",
   "metadata": {},
   "source": [
    "# 1. Reading data and checking for missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "426c012e",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"train.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "d8318892",
   "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>Unnamed: 0</th>\n",
       "      <th>ID</th>\n",
       "      <th>ncap rating</th>\n",
       "      <th>is power door locks</th>\n",
       "      <th>policyholder age</th>\n",
       "      <th>is parking camera</th>\n",
       "      <th>rear brakes type</th>\n",
       "      <th>is adjustable steering</th>\n",
       "      <th>is tpms</th>\n",
       "      <th>is driver seat height adjustable</th>\n",
       "      <th>...</th>\n",
       "      <th>is parking sensors</th>\n",
       "      <th>power</th>\n",
       "      <th>gross weight</th>\n",
       "      <th>is day night rear view mirror</th>\n",
       "      <th>model</th>\n",
       "      <th>gear box</th>\n",
       "      <th>airbags</th>\n",
       "      <th>turning radius</th>\n",
       "      <th>area danger level</th>\n",
       "      <th>is claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0.394496</td>\n",
       "      <td>No</td>\n",
       "      <td>Drum</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>55.92bhp@5300rpm</td>\n",
       "      <td>1340.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Model_8</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.7</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0.417364</td>\n",
       "      <td>No</td>\n",
       "      <td>Drum</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>55.92bhp@5300rpm</td>\n",
       "      <td>1340.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Model_8</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.7</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0.438922</td>\n",
       "      <td>No</td>\n",
       "      <td>Drum</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>88.50bhp@6000rpm</td>\n",
       "      <td>1335.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Model_6</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.8</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>0.656276</td>\n",
       "      <td>No</td>\n",
       "      <td>Drum</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>40.36bhp@6000rpm</td>\n",
       "      <td>1185.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Model_1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.6</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>0.462328</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Disc</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Yes</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>113.45bhp@4000rpm</td>\n",
       "      <td>1720.0</td>\n",
       "      <td>No</td>\n",
       "      <td>Model_4</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 46 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  ID  ncap rating is power door locks  policyholder age  \\\n",
       "0           0   0          2.0                 Yes          0.394496   \n",
       "1           1   1          2.0                 Yes          0.417364   \n",
       "2           2   2          2.0                 Yes          0.438922   \n",
       "3           3   3          0.0                  No          0.656276   \n",
       "4           4   4          3.0                 Yes          0.462328   \n",
       "\n",
       "  is parking camera rear brakes type is adjustable steering is tpms  \\\n",
       "0                No             Drum                     No      No   \n",
       "1                No             Drum                     No      No   \n",
       "2                No             Drum                    Yes      No   \n",
       "3                No             Drum                     No      No   \n",
       "4               Yes             Disc                    Yes     Yes   \n",
       "\n",
       "  is driver seat height adjustable  ... is parking sensors              power  \\\n",
       "0                               No  ...                Yes   55.92bhp@5300rpm   \n",
       "1                               No  ...                Yes   55.92bhp@5300rpm   \n",
       "2                              Yes  ...                Yes   88.50bhp@6000rpm   \n",
       "3                               No  ...                Yes   40.36bhp@6000rpm   \n",
       "4                              Yes  ...                Yes  113.45bhp@4000rpm   \n",
       "\n",
       "  gross weight is day night rear view mirror    model gear box airbags  \\\n",
       "0       1340.0                            No  Model_8      5.0     2.0   \n",
       "1       1340.0                            No  Model_8      5.0     2.0   \n",
       "2       1335.0                           Yes  Model_6      5.0     2.0   \n",
       "3       1185.0                            No  Model_1      5.0     2.0   \n",
       "4       1720.0                            No  Model_4      6.0     6.0   \n",
       "\n",
       "  turning radius area danger level  is claim  \n",
       "0            4.7                 4         0  \n",
       "1            4.7                 1         1  \n",
       "2            4.8                 5         0  \n",
       "3            4.6                 2         0  \n",
       "4            5.2                 3         0  \n",
       "\n",
       "[5 rows x 46 columns]"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "d635cfa3",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unnamed: 0                            0\n",
       "ID                                    0\n",
       "ncap rating                          99\n",
       "is power door locks                  94\n",
       "policyholder age                     81\n",
       "is parking camera                    83\n",
       "rear brakes type                     86\n",
       "is adjustable steering               90\n",
       "is tpms                              89\n",
       "is driver seat height adjustable     85\n",
       "segment                              83\n",
       "car age                             105\n",
       "is central locking                   89\n",
       "is rear window wiper                 69\n",
       "height                               96\n",
       "cluster area                         85\n",
       "is ecw                               85\n",
       "fuel type                            89\n",
       "torque                               76\n",
       "engine volume                       103\n",
       "transmission type                    97\n",
       "manufacturer                         87\n",
       "cylinder                             97\n",
       "is rear window washer                89\n",
       "is front fog lights                  86\n",
       "is brake assist                     105\n",
       "is power steering                    92\n",
       "is esc                               78\n",
       "population                           95\n",
       "is rear window defogger              88\n",
       "time period                          82\n",
       "engine type                          97\n",
       "is speed alert                       61\n",
       "steering type                       104\n",
       "length                               88\n",
       "width                                74\n",
       "is parking sensors                   81\n",
       "power                                97\n",
       "gross weight                         97\n",
       "is day night rear view mirror        96\n",
       "model                                86\n",
       "gear box                            116\n",
       "airbags                              86\n",
       "turning radius                       98\n",
       "area danger level                     0\n",
       "is claim                              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "2b7041e3",
   "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>ncap rating</th>\n",
       "      <th>is power door locks</th>\n",
       "      <th>segment</th>\n",
       "      <th>car age</th>\n",
       "      <th>population</th>\n",
       "      <th>time period</th>\n",
       "      <th>is claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C2</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C1</td>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       ncap rating is power door locks segment  car age  population  \\\n",
       "0              2.0                 Yes      B1     0.08      8883.0   \n",
       "1              2.0                 Yes      B1     0.06     34751.0   \n",
       "2              2.0                 Yes      B2     0.03     27026.0   \n",
       "3              0.0                  No       A     0.04      5482.0   \n",
       "4              3.0                 Yes      C2     0.04     17833.0   \n",
       "...            ...                 ...     ...      ...         ...   \n",
       "43587          2.0                 Yes      B2     0.05      8828.0   \n",
       "43588          2.0                 Yes      C1     0.11      5455.0   \n",
       "43589          2.0                 Yes      B2     0.15     27082.0   \n",
       "43590          2.0                 Yes      B2     0.14      8845.0   \n",
       "43591          0.0                  No       A     0.01     27052.0   \n",
       "\n",
       "       time period  is claim  \n",
       "0         0.304554         0  \n",
       "1         1.169197         1  \n",
       "2         1.219797         0  \n",
       "3         0.746151         0  \n",
       "4         0.171981         0  \n",
       "...            ...       ...  \n",
       "43587     0.125440         0  \n",
       "43588     1.177618         0  \n",
       "43589     0.601047         0  \n",
       "43590     0.442004         0  \n",
       "43591     0.813746         0  \n",
       "\n",
       "[43592 rows x 7 columns]"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sample columns to use (3 categorical,3 numerical,1 target):\n",
    "\n",
    "df1 = df[[\"ncap rating\",\"is power door locks\",\"segment\",\"car age\",\"population\",\"time period\",\"is claim\"]]\n",
    "df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "3acd5ff2",
   "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>ncap</th>\n",
       "      <th>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "      <th>claim</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C2</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C1</td>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       ncap doorlocks segment   age      pop      time  claim\n",
       "0       2.0       Yes      B1  0.08   8883.0  0.304554      0\n",
       "1       2.0       Yes      B1  0.06  34751.0  1.169197      1\n",
       "2       2.0       Yes      B2  0.03  27026.0  1.219797      0\n",
       "3       0.0        No       A  0.04   5482.0  0.746151      0\n",
       "4       3.0       Yes      C2  0.04  17833.0  0.171981      0\n",
       "...     ...       ...     ...   ...      ...       ...    ...\n",
       "43587   2.0       Yes      B2  0.05   8828.0  0.125440      0\n",
       "43588   2.0       Yes      C1  0.11   5455.0  1.177618      0\n",
       "43589   2.0       Yes      B2  0.15  27082.0  0.601047      0\n",
       "43590   2.0       Yes      B2  0.14   8845.0  0.442004      0\n",
       "43591   0.0        No       A  0.01  27052.0  0.813746      0\n",
       "\n",
       "[43592 rows x 7 columns]"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.columns = [\"ncap\",\"doorlocks\",\"segment\",\"age\",\"pop\",\"time\",\"claim\"]\n",
    "df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "2bbd3031",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ncap          99\n",
       "doorlocks     94\n",
       "segment       83\n",
       "age          105\n",
       "pop           95\n",
       "time          82\n",
       "claim          0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "3f7b0652",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cats = df1[[\"ncap\",\"doorlocks\",\"segment\"]]\n",
    "df_nums = df1[[\"age\",\"pop\",\"time\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27b526d0",
   "metadata": {},
   "source": [
    "**Although the missing values are not significant in 43592 pieces of data on a per piece basis, we will try to fill them statistically correctly instead of deleting them.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96404ce6",
   "metadata": {},
   "source": [
    "# 2. Correlation between features and nullity correlation check"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "52b5cb23",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(font_scale=1.5)  \n",
    "correlation_matrix = df_nums.corr().astype(float).round(4)\n",
    "plt.figure(figsize=(7,7))\n",
    "ax = sns.heatmap(correlation_matrix, vmax=1, square=True,annot=True,cmap='RdYlGn',annot_kws={'size': 13})\n",
    "plt.title('Correlation matrix between the features')\n",
    "plt.yticks(rotation=0)\n",
    "plt.xticks(rotation=90)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "185c29a4",
   "metadata": {},
   "source": [
    "**When we check the correlation of numerical data, it seems that they do not have a relationship with each other.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "67739fb1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "msno.bar(df1, figsize=(20,6),filter=\"top\",log=True);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "7e11cee7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "msno.matrix(df1);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "eb84c8ca",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Presents a `seaborn` heatmap visualization of nullity correlation in the given DataFrame\n",
    "msno.heatmap(df1);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d81b7c9",
   "metadata": {},
   "source": [
    "**As seen in the nullity correlation graph and the matrix, the missing values are not related to each other and are formed randomly.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9d5aad5",
   "metadata": {},
   "source": [
    "# 3. Handling With Missing Values"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97f978fa",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">3.1. Using the SimpleImputer (Cats-Nums)</span> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "6705f96c",
   "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>ncap</th>\n",
       "      <th>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>2.0</td>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.0</td>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      ncap doorlocks segment\n",
       "0      2.0       Yes      B1\n",
       "1      2.0       Yes      B1\n",
       "2      2.0       Yes      B2\n",
       "3      0.0        No       A\n",
       "4      3.0       Yes      C2\n",
       "...    ...       ...     ...\n",
       "43587  2.0       Yes      B2\n",
       "43588  2.0       Yes      C1\n",
       "43589  2.0       Yes      B2\n",
       "43590  2.0       Yes      B2\n",
       "43591  0.0        No       A\n",
       "\n",
       "[43592 rows x 3 columns]"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# For the categorical features, we will use \"most frequent\" strategy.\n",
    "\n",
    "freq_cats = SimpleImputer(missing_values=np.nan, strategy='most_frequent')\n",
    "df_cats_imp = pd.DataFrame(freq_cats.fit_transform(df_cats))\n",
    "df_cats_imp.columns=df_cats.columns\n",
    "df_cats_imp.index=df_cats.index\n",
    "df_cats_imp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "79ce1ab1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ncap         0\n",
       "doorlocks    0\n",
       "segment      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cats_imp.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "cefc0233",
   "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>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        age      pop      time\n",
       "0      0.08   8883.0  0.304554\n",
       "1      0.06  34751.0  1.169197\n",
       "2      0.03  27026.0  1.219797\n",
       "3      0.04   5482.0  0.746151\n",
       "4      0.04  17833.0  0.171981\n",
       "...     ...      ...       ...\n",
       "43587  0.05   8828.0  0.125440\n",
       "43588  0.11   5455.0  1.177618\n",
       "43589  0.15  27082.0  0.601047\n",
       "43590  0.14   8845.0  0.442004\n",
       "43591  0.01  27052.0  0.813746\n",
       "\n",
       "[43592 rows x 3 columns]"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# For the numeric features, we will use \"mean\" strategy.\n",
    "\n",
    "mean_nums = SimpleImputer(missing_values=np.nan, strategy='mean')\n",
    "df_nums_imp = pd.DataFrame(freq_cats.fit_transform(df_nums))\n",
    "df_nums_imp.columns=df_nums.columns\n",
    "df_nums_imp.index=df_nums.index\n",
    "df_nums_imp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "d1955ee2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age     0\n",
       "pop     0\n",
       "time    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_nums_imp.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "36f96857",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">3.2. Using the IterativeImputer (Nums only)</span> "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6816fc15",
   "metadata": {},
   "source": [
    "**Estimators: BayesianRidge(),DecisionTreeRegressor(max_features='sqrt', random_state=0),ExtraTreesRegressor(n_estimators=10, random_state=0),KNeighborsRegressor(n_neighbors=15)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "ff65214c",
   "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>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        age      pop      time\n",
       "0      0.08   8883.0  0.304554\n",
       "1      0.06  34751.0  1.169197\n",
       "2      0.03  27026.0  1.219797\n",
       "3      0.04   5482.0  0.746151\n",
       "4      0.04  17833.0  0.171981\n",
       "...     ...      ...       ...\n",
       "43587  0.05   8828.0  0.125440\n",
       "43588  0.11   5455.0  1.177618\n",
       "43589  0.15  27082.0  0.601047\n",
       "43590  0.14   8845.0  0.442004\n",
       "43591  0.01  27052.0  0.813746\n",
       "\n",
       "[43592 rows x 3 columns]"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Num variables:\n",
    "\n",
    "estimator = DecisionTreeRegressor(random_state=0)\n",
    "df_filled_nums = IterativeImputer(estimator=estimator).fit_transform(df_nums)\n",
    "df_filled_nums = pd.DataFrame(df_filled_nums, columns=df_nums.columns)\n",
    "df_filled_nums"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "54bdb66f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age     0\n",
       "pop     0\n",
       "time    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_filled_nums.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "57c8e272",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">3.3. Using the KNNImputer (Nums only)</span> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "5a6c037f",
   "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>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        age      pop      time\n",
       "0      0.08   8883.0  0.304554\n",
       "1      0.06  34751.0  1.169197\n",
       "2      0.03  27026.0  1.219797\n",
       "3      0.04   5482.0  0.746151\n",
       "4      0.04  17833.0  0.171981\n",
       "...     ...      ...       ...\n",
       "43587  0.05   8828.0  0.125440\n",
       "43588  0.11   5455.0  1.177618\n",
       "43589  0.15  27082.0  0.601047\n",
       "43590  0.14   8845.0  0.442004\n",
       "43591  0.01  27052.0  0.813746\n",
       "\n",
       "[43592 rows x 3 columns]"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Num variables:\n",
    "\n",
    "imputer = KNNImputer(n_neighbors=10,weights='uniform')\n",
    "df_filled_nums_KNN = imputer.fit_transform(df_nums)\n",
    "df_filled_nums_KNN = pd.DataFrame(df_filled_nums_KNN, columns=df_nums.columns)\n",
    "df_filled_nums_KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "1dc34343",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "age     0\n",
       "pop     0\n",
       "time    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_filled_nums_KNN.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f671c040",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">3.4. Predictions For Categorical Missing Values (ML)</span> "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "111a2bfa",
   "metadata": {},
   "source": [
    "**We will set up a classification ML for 1 categorical variable with missing values and try to predict them.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "8f7651d3",
   "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>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>ncap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Yes</td>\n",
       "      <td>C2</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>Yes</td>\n",
       "      <td>C1</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      doorlocks segment  ncap\n",
       "0           Yes      B1   2.0\n",
       "1           Yes      B1   2.0\n",
       "2           Yes      B2   2.0\n",
       "3            No       A   0.0\n",
       "4           Yes      C2   3.0\n",
       "...         ...     ...   ...\n",
       "43587       Yes      B2   2.0\n",
       "43588       Yes      C1   2.0\n",
       "43589       Yes      B2   2.0\n",
       "43590       Yes      B2   2.0\n",
       "43591        No       A   0.0\n",
       "\n",
       "[43592 rows x 3 columns]"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Prediction for \"ncap\"\n",
    "# First, filling the other 2 vars with most frequent values.\n",
    "\n",
    "cats_2 = SimpleImputer(missing_values=np.nan, strategy='most_frequent')\n",
    "ncap_ml = pd.DataFrame(cats_2.fit_transform(df_cats[[\"doorlocks\",\"segment\"]]))\n",
    "ncap_ml.columns=[\"doorlocks\",\"segment\"]\n",
    "ncap_ml = pd.concat([ncap_ml, df_cats.ncap], axis = 1) \n",
    "ncap_ml"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "785319f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "doorlocks     0\n",
       "segment       0\n",
       "ncap         99\n",
       "dtype: int64"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ncap_ml.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "6428fd28",
   "metadata": {},
   "outputs": [],
   "source": [
    "ncap_ml = pd.concat([ncap_ml, df_filled_nums], axis = 1) \n",
    "\n",
    "df_with_ncap = ncap_ml[ncap_ml.ncap.notna()]\n",
    "df_no_ncap = ncap_ml[ncap_ml.ncap.isna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "801be600",
   "metadata": {},
   "outputs": [],
   "source": [
    "class MultiColumnLabelEncoder:\n",
    "    def __init__(self, columns=None):\n",
    "        self.columns = columns \n",
    "        \n",
    "    def fit(self, X, y=None):\n",
    "        self.encoders = {}\n",
    "        columns = X.columns if self.columns is None else self.columns\n",
    "        for col in columns:\n",
    "            self.encoders[col] = LabelEncoder().fit(X[col])\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X):\n",
    "        output = X.copy()\n",
    "        columns = X.columns if self.columns is None else self.columns\n",
    "        for col in columns:\n",
    "            output[col] = self.encoders[col].transform(X[col])\n",
    "        return output\n",
    "\n",
    "    def fit_transform(self, X, y=None):\n",
    "        return self.fit(X,y).transform(X)\n",
    "\n",
    "    def inverse_transform(self, X):\n",
    "        output = X.copy()\n",
    "        columns = X.columns if self.columns is None else self.columns\n",
    "        for col in columns:\n",
    "            output[col] = self.encoders[col].inverse_transform(X[col])\n",
    "        return output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "825f1444",
   "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>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>ncap</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43493 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       doorlocks  segment  ncap   age      pop      time\n",
       "0              1        1   2.0  0.08   8883.0  0.304554\n",
       "1              1        1   2.0  0.06  34751.0  1.169197\n",
       "2              1        2   2.0  0.03  27026.0  1.219797\n",
       "3              0        0   0.0  0.04   5482.0  0.746151\n",
       "4              1        4   3.0  0.04  17833.0  0.171981\n",
       "...          ...      ...   ...   ...      ...       ...\n",
       "43587          1        2   2.0  0.05   8828.0  0.125440\n",
       "43588          1        3   2.0  0.11   5455.0  1.177618\n",
       "43589          1        2   2.0  0.15  27082.0  0.601047\n",
       "43590          1        2   2.0  0.14   8845.0  0.442004\n",
       "43591          0        0   0.0  0.01  27052.0  0.813746\n",
       "\n",
       "[43493 rows x 6 columns]"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multi = MultiColumnLabelEncoder(columns=['doorlocks','segment'])\n",
    "encoded = multi.fit_transform(df_with_ncap)\n",
    "encoded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "6339f6d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ACCURACY: TRAIN=0.9253 TEST=0.8847\n"
     ]
    }
   ],
   "source": [
    "X = encoded.drop(\"ncap\", axis=1)\n",
    "Y = encoded.ncap\n",
    "\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=0)\n",
    "\n",
    "forest = RandomForestClassifier(n_estimators=45, max_depth=25, random_state=False, \n",
    "                                max_features=0.6, min_samples_leaf=3, n_jobs=-1)\n",
    "\n",
    "forest.fit(X_train, Y_train)\n",
    "\n",
    "y_pred_train = forest.predict(X_train)\n",
    "y_pred = forest.predict(X_test)\n",
    "\n",
    "y_pred_proba = forest.predict_proba(X_test)\n",
    "\n",
    "accuracy_train = accuracy_score(Y_train, y_pred_train)\n",
    "accuracy_test = accuracy_score(Y_test, y_pred)\n",
    "\n",
    "print(\"ACCURACY: TRAIN=%.4f TEST=%.4f\" % (accuracy_train,accuracy_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "97907408",
   "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>ncap</th>\n",
       "      <th>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17879.0</td>\n",
       "      <td>0.527331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>731</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>1.133107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1316</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.04</td>\n",
       "      <td>21640.0</td>\n",
       "      <td>1.170905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1622</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0.09</td>\n",
       "      <td>7829.0</td>\n",
       "      <td>0.844531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1969</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0.02</td>\n",
       "      <td>7819.0</td>\n",
       "      <td>1.302173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42624</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5421.0</td>\n",
       "      <td>0.471655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42788</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0.18</td>\n",
       "      <td>34796.0</td>\n",
       "      <td>0.557700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42879</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.07</td>\n",
       "      <td>27032.0</td>\n",
       "      <td>1.154294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42955</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27079.0</td>\n",
       "      <td>0.837517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43573</th>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0.05</td>\n",
       "      <td>73439.0</td>\n",
       "      <td>0.346634</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>99 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       ncap  doorlocks  segment   age      pop      time\n",
       "126     0.0          0        0  0.04  17879.0  0.527331\n",
       "731     0.0          0        0  0.01   8828.0  1.133107\n",
       "1316    2.0          1        1  0.04  21640.0  1.170905\n",
       "1622    2.0          1        2  0.09   7829.0  0.844531\n",
       "1969    2.0          1        2  0.02   7819.0  1.302173\n",
       "...     ...        ...      ...   ...      ...       ...\n",
       "42624   3.0          1        4  0.04   5421.0  0.471655\n",
       "42788   2.0          1        2  0.18  34796.0  0.557700\n",
       "42879   2.0          1        0  0.07  27032.0  1.154294\n",
       "42955   0.0          0        0  0.01  27079.0  0.837517\n",
       "43573   3.0          1        4  0.05  73439.0  0.346634\n",
       "\n",
       "[99 rows x 6 columns]"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multi2 = MultiColumnLabelEncoder(columns=['doorlocks','segment'])\n",
    "encoded2 = multi2.fit_transform(df_no_ncap)\n",
    "\n",
    "encoded2 = encoded2.drop('ncap', axis=1)\n",
    "prediction = forest.predict(encoded2)\n",
    "\n",
    "encoded2.insert(0, 'ncap', prediction)\n",
    "encoded2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "263ffc4f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ncap', 'doorlocks', 'segment', 'age', 'pop', 'time'], dtype='object')"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoded2.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "7f69897d",
   "metadata": {},
   "outputs": [],
   "source": [
    "encoded2 = encoded2[['doorlocks', 'segment', 'ncap','age', 'pop', 'time']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "7f2d70cc",
   "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>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>ncap</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17879.0</td>\n",
       "      <td>0.527331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>731</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>1.133107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1316</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>21640.0</td>\n",
       "      <td>1.170905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1622</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.09</td>\n",
       "      <td>7829.0</td>\n",
       "      <td>0.844531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1969</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.02</td>\n",
       "      <td>7819.0</td>\n",
       "      <td>1.302173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42624</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5421.0</td>\n",
       "      <td>0.471655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42788</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.18</td>\n",
       "      <td>34796.0</td>\n",
       "      <td>0.557700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42879</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.07</td>\n",
       "      <td>27032.0</td>\n",
       "      <td>1.154294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42955</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27079.0</td>\n",
       "      <td>0.837517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43573</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>73439.0</td>\n",
       "      <td>0.346634</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>99 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       doorlocks  segment  ncap   age      pop      time\n",
       "126            0        0   0.0  0.04  17879.0  0.527331\n",
       "731            0        0   0.0  0.01   8828.0  1.133107\n",
       "1316           1        1   2.0  0.04  21640.0  1.170905\n",
       "1622           1        2   2.0  0.09   7829.0  0.844531\n",
       "1969           1        2   2.0  0.02   7819.0  1.302173\n",
       "...          ...      ...   ...   ...      ...       ...\n",
       "42624          1        4   3.0  0.04   5421.0  0.471655\n",
       "42788          1        2   2.0  0.18  34796.0  0.557700\n",
       "42879          1        0   2.0  0.07  27032.0  1.154294\n",
       "42955          0        0   0.0  0.01  27079.0  0.837517\n",
       "43573          1        4   3.0  0.05  73439.0  0.346634\n",
       "\n",
       "[99 rows x 6 columns]"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoded2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "015296ef",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "ncap_fulled = pd.concat([encoded, encoded2]) \n",
    "ncap_fulled = ncap_fulled.sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "dcaa7975",
   "metadata": {
    "scrolled": false
   },
   "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>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>ncap</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       doorlocks  segment  ncap   age      pop      time\n",
       "0              1        1   2.0  0.08   8883.0  0.304554\n",
       "1              1        1   2.0  0.06  34751.0  1.169197\n",
       "2              1        2   2.0  0.03  27026.0  1.219797\n",
       "3              0        0   0.0  0.04   5482.0  0.746151\n",
       "4              1        4   3.0  0.04  17833.0  0.171981\n",
       "...          ...      ...   ...   ...      ...       ...\n",
       "43587          1        2   2.0  0.05   8828.0  0.125440\n",
       "43588          1        3   2.0  0.11   5455.0  1.177618\n",
       "43589          1        2   2.0  0.15  27082.0  0.601047\n",
       "43590          1        2   2.0  0.14   8845.0  0.442004\n",
       "43591          0        0   0.0  0.01  27052.0  0.813746\n",
       "\n",
       "[43592 rows x 6 columns]"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ncap_fulled"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d06d5c0b",
   "metadata": {},
   "source": [
    "### Reverse Transform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "ea97e1ce",
   "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>doorlocks</th>\n",
       "      <th>segment</th>\n",
       "      <th>ncap</th>\n",
       "      <th>age</th>\n",
       "      <th>pop</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.08</td>\n",
       "      <td>8883.0</td>\n",
       "      <td>0.304554</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.06</td>\n",
       "      <td>34751.0</td>\n",
       "      <td>1.169197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.03</td>\n",
       "      <td>27026.0</td>\n",
       "      <td>1.219797</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>5482.0</td>\n",
       "      <td>0.746151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Yes</td>\n",
       "      <td>C2</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.04</td>\n",
       "      <td>17833.0</td>\n",
       "      <td>0.171981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43587</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.05</td>\n",
       "      <td>8828.0</td>\n",
       "      <td>0.125440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43588</th>\n",
       "      <td>Yes</td>\n",
       "      <td>C1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>5455.0</td>\n",
       "      <td>1.177618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43589</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>27082.0</td>\n",
       "      <td>0.601047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43590</th>\n",
       "      <td>Yes</td>\n",
       "      <td>B2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>8845.0</td>\n",
       "      <td>0.442004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43591</th>\n",
       "      <td>No</td>\n",
       "      <td>A</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>27052.0</td>\n",
       "      <td>0.813746</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>43592 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      doorlocks segment  ncap   age      pop      time\n",
       "0           Yes      B1   2.0  0.08   8883.0  0.304554\n",
       "1           Yes      B1   2.0  0.06  34751.0  1.169197\n",
       "2           Yes      B2   2.0  0.03  27026.0  1.219797\n",
       "3            No       A   0.0  0.04   5482.0  0.746151\n",
       "4           Yes      C2   3.0  0.04  17833.0  0.171981\n",
       "...         ...     ...   ...   ...      ...       ...\n",
       "43587       Yes      B2   2.0  0.05   8828.0  0.125440\n",
       "43588       Yes      C1   2.0  0.11   5455.0  1.177618\n",
       "43589       Yes      B2   2.0  0.15  27082.0  0.601047\n",
       "43590       Yes      B2   2.0  0.14   8845.0  0.442004\n",
       "43591        No       A   0.0  0.01  27052.0  0.813746\n",
       "\n",
       "[43592 rows x 6 columns]"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inv = multi.inverse_transform(ncap_fulled)\n",
    "inv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "be5f4cf7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "doorlocks    0\n",
       "segment      0\n",
       "ncap         0\n",
       "age          0\n",
       "pop          0\n",
       "time         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inv.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "977fe078",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}