A1_start_SGD.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Description\n",
    "\n",
    "|    Variable            \t|    Definition                         |     \n",
    "|---------------------------|---------------------------------------|\n",
    "|    employee_id         \t|    Unique ID for employee             |     \n",
    "|    department          \t|    Department of employee             |     \n",
    "|    region              \t|    Region of employment (unordered)   |     \n",
    "|    education           \t|    Education Level                    |     \n",
    "|    gender              \t|    Gender of Employee                 |     \n",
    "|    recruitment_channel \t|    Channel of recruitment for employee |    \n",
    "|    no_of_trainings     \t|    no of other trainings completed in previous year on soft skills, technical skills etc. |\n",
    "|    age                 \t|    Age of Employee                         |\n",
    "|    previous_year_rating\t|    Employee Rating for the previous year   |\n",
    "|    length_of_service   \t|    Length of service in years              \n",
    "|    KPIs_met >80%       \t|    if Percent of KPIs(Key performance Indicators) >80% then 1 else 0 |\n",
    "|    awards_won?         \t|    if awards won during previous year then 1 else 0 |\n",
    "|    avg_training_score  \t|    Average score in current training evaluations |\n",
    "|    is_promoted         \t|    (Target) Recommended for promotion      |\n",
    "\n",
    "\n",
    "## EVALUATION METRIC\n",
    "\n",
    "The evaluation metric for this competition is F1 Score.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing Python Modules"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T05:59:54.512424Z",
     "start_time": "2018-09-14T05:59:54.499459Z"
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder, StandardScaler, MinMaxScaler, FunctionTransformer\n",
    "from sklearn.pipeline import Pipeline, make_pipeline, FeatureUnion, make_union"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:23:48.690625Z",
     "start_time": "2018-09-14T04:23:48.685637Z"
    }
   },
   "outputs": [],
   "source": [
    "sns.set_style('whitegrid')\n",
    "# sns.set(context=\"paper\", font=\"monospace\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Loading "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:00:43.163619Z",
     "start_time": "2018-09-14T04:00:42.959164Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Data Size : (54808, 13)\n"
     ]
    },
    {
     "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "      <th>no_of_trainings</th>\n",
       "      <th>age</th>\n",
       "      <th>previous_year_rating</th>\n",
       "      <th>length_of_service</th>\n",
       "      <th>is_kpi_met</th>\n",
       "      <th>has_won_awards</th>\n",
       "      <th>avg_training_score</th>\n",
       "      <th>is_promoted</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>65438</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_7</td>\n",
       "      <td>Master's &amp; above</td>\n",
       "      <td>f</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>35</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65141</th>\n",
       "      <td>Operations</td>\n",
       "      <td>region_22</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7513</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_19</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>34</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2542</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_23</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48945</th>\n",
       "      <td>Technology</td>\n",
       "      <td>region_26</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>73</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    department     region         education gender  \\\n",
       "employee_id                                                          \n",
       "65438        Sales & Marketing   region_7  Master's & above      f   \n",
       "65141               Operations  region_22        Bachelor's      m   \n",
       "7513         Sales & Marketing  region_19        Bachelor's      m   \n",
       "2542         Sales & Marketing  region_23        Bachelor's      m   \n",
       "48945               Technology  region_26        Bachelor's      m   \n",
       "\n",
       "            recruitment_channel  no_of_trainings  age  previous_year_rating  \\\n",
       "employee_id                                                                   \n",
       "65438                  sourcing                1   35                   5.0   \n",
       "65141                     other                1   30                   5.0   \n",
       "7513                   sourcing                1   34                   3.0   \n",
       "2542                      other                2   39                   1.0   \n",
       "48945                     other                1   45                   3.0   \n",
       "\n",
       "             length_of_service  is_kpi_met  has_won_awards  \\\n",
       "employee_id                                                  \n",
       "65438                        8           1               0   \n",
       "65141                        4           0               0   \n",
       "7513                         7           0               0   \n",
       "2542                        10           0               0   \n",
       "48945                        2           0               0   \n",
       "\n",
       "             avg_training_score  is_promoted  \n",
       "employee_id                                   \n",
       "65438                        49            0  \n",
       "65141                        60            0  \n",
       "7513                         50            0  \n",
       "2542                         50            0  \n",
       "48945                        73            0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def load_training_dataset():\n",
    "    return pd.read_csv('data/train_LZdllcl.csv', index_col='employee_id')\n",
    "df = load_training_dataset()\n",
    "df = df.rename(columns={'KPIs_met >80%':'is_kpi_met', 'awards_won?':'has_won_awards'})\n",
    "print('Train Data Size :', df.shape)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:00:54.556714Z",
     "start_time": "2018-09-14T04:00:54.460971Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Data Size : (23490, 12)\n"
     ]
    },
    {
     "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "      <th>no_of_trainings</th>\n",
       "      <th>age</th>\n",
       "      <th>previous_year_rating</th>\n",
       "      <th>length_of_service</th>\n",
       "      <th>is_kpi_met</th>\n",
       "      <th>has_won_awards</th>\n",
       "      <th>avg_training_score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8724</th>\n",
       "      <td>Technology</td>\n",
       "      <td>region_26</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74430</th>\n",
       "      <td>HR</td>\n",
       "      <td>region_4</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>f</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>31</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72255</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_13</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>31</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38562</th>\n",
       "      <td>Procurement</td>\n",
       "      <td>region_2</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>f</td>\n",
       "      <td>other</td>\n",
       "      <td>3</td>\n",
       "      <td>31</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64486</th>\n",
       "      <td>Finance</td>\n",
       "      <td>region_29</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>4.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    department     region   education gender  \\\n",
       "employee_id                                                    \n",
       "8724                Technology  region_26  Bachelor's      m   \n",
       "74430                       HR   region_4  Bachelor's      f   \n",
       "72255        Sales & Marketing  region_13  Bachelor's      m   \n",
       "38562              Procurement   region_2  Bachelor's      f   \n",
       "64486                  Finance  region_29  Bachelor's      m   \n",
       "\n",
       "            recruitment_channel  no_of_trainings  age  previous_year_rating  \\\n",
       "employee_id                                                                   \n",
       "8724                   sourcing                1   24                   NaN   \n",
       "74430                     other                1   31                   3.0   \n",
       "72255                     other                1   31                   1.0   \n",
       "38562                     other                3   31                   2.0   \n",
       "64486                  sourcing                1   30                   4.0   \n",
       "\n",
       "             length_of_service  is_kpi_met  has_won_awards  avg_training_score  \n",
       "employee_id                                                                     \n",
       "8724                         1           1               0                  77  \n",
       "74430                        5           0               0                  51  \n",
       "72255                        4           0               0                  47  \n",
       "38562                        9           0               0                  65  \n",
       "64486                        7           0               0                  61  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test = pd.read_csv('data/test_2umaH9m.csv', index_col='employee_id')\n",
    "test = test.rename(columns={'KPIs_met >80%':'is_kpi_met', 'awards_won?':'has_won_awards'})\n",
    "print('Test Data Size :', test.shape)\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:00:59.103450Z",
     "start_time": "2018-09-14T04:00:59.077519Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "department                 0\n",
       "region                     0\n",
       "education               1034\n",
       "gender                     0\n",
       "recruitment_channel        0\n",
       "no_of_trainings            0\n",
       "age                        0\n",
       "previous_year_rating    1812\n",
       "length_of_service          0\n",
       "is_kpi_met                 0\n",
       "has_won_awards             0\n",
       "avg_training_score         0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-13T19:05:55.603686Z",
     "start_time": "2018-09-13T19:05:55.596707Z"
    }
   },
   "source": [
    "## Data Pre-processing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:01:01.440322Z",
     "start_time": "2018-09-14T04:01:01.383474Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 54808 entries, 65438 to 51526\n",
      "Data columns (total 13 columns):\n",
      "department              54808 non-null object\n",
      "region                  54808 non-null object\n",
      "education               52399 non-null object\n",
      "gender                  54808 non-null object\n",
      "recruitment_channel     54808 non-null object\n",
      "no_of_trainings         54808 non-null int64\n",
      "age                     54808 non-null int64\n",
      "previous_year_rating    50684 non-null float64\n",
      "length_of_service       54808 non-null int64\n",
      "is_kpi_met              54808 non-null int64\n",
      "has_won_awards          54808 non-null int64\n",
      "avg_training_score      54808 non-null int64\n",
      "is_promoted             54808 non-null int64\n",
      "dtypes: float64(1), int64(7), object(5)\n",
      "memory usage: 5.9+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:15:39.676342Z",
     "start_time": "2018-09-14T04:15:39.591568Z"
    }
   },
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>no_of_trainings</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>1.253011</td>\n",
       "      <td>0.609264</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>34.803915</td>\n",
       "      <td>7.660169</td>\n",
       "      <td>20.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>previous_year_rating</th>\n",
       "      <td>50684.0</td>\n",
       "      <td>3.329256</td>\n",
       "      <td>1.259993</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>length_of_service</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>5.865512</td>\n",
       "      <td>4.265094</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_kpi_met</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>0.351974</td>\n",
       "      <td>0.477590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_won_awards</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>0.023172</td>\n",
       "      <td>0.150450</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>avg_training_score</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>63.386750</td>\n",
       "      <td>13.371559</td>\n",
       "      <td>39.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>99.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_promoted</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>0.085170</td>\n",
       "      <td>0.279137</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>iregion</th>\n",
       "      <td>54808.0</td>\n",
       "      <td>14.195045</td>\n",
       "      <td>10.086273</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>34.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        count       mean        std   min   25%   50%   75%  \\\n",
       "no_of_trainings       54808.0   1.253011   0.609264   1.0   1.0   1.0   1.0   \n",
       "age                   54808.0  34.803915   7.660169  20.0  29.0  33.0  39.0   \n",
       "previous_year_rating  50684.0   3.329256   1.259993   1.0   3.0   3.0   4.0   \n",
       "length_of_service     54808.0   5.865512   4.265094   1.0   3.0   5.0   7.0   \n",
       "is_kpi_met            54808.0   0.351974   0.477590   0.0   0.0   0.0   1.0   \n",
       "has_won_awards        54808.0   0.023172   0.150450   0.0   0.0   0.0   0.0   \n",
       "avg_training_score    54808.0  63.386750  13.371559  39.0  51.0  60.0  76.0   \n",
       "is_promoted           54808.0   0.085170   0.279137   0.0   0.0   0.0   0.0   \n",
       "iregion               54808.0  14.195045  10.086273   1.0   4.0  13.0  22.0   \n",
       "\n",
       "                       max  \n",
       "no_of_trainings       10.0  \n",
       "age                   60.0  \n",
       "previous_year_rating   5.0  \n",
       "length_of_service     37.0  \n",
       "is_kpi_met             1.0  \n",
       "has_won_awards         1.0  \n",
       "avg_training_score    99.0  \n",
       "is_promoted            1.0  \n",
       "iregion               34.0  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:01:24.110150Z",
     "start_time": "2018-09-14T04:01:24.064274Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "department                 0\n",
       "region                     0\n",
       "education               2409\n",
       "gender                     0\n",
       "recruitment_channel        0\n",
       "no_of_trainings            0\n",
       "age                        0\n",
       "previous_year_rating    4124\n",
       "length_of_service          0\n",
       "is_kpi_met                 0\n",
       "has_won_awards             0\n",
       "avg_training_score         0\n",
       "is_promoted                0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:01:28.950231Z",
     "start_time": "2018-09-14T04:01:28.918317Z"
    },
    "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "      <th>no_of_trainings</th>\n",
       "      <th>age</th>\n",
       "      <th>previous_year_rating</th>\n",
       "      <th>length_of_service</th>\n",
       "      <th>is_kpi_met</th>\n",
       "      <th>has_won_awards</th>\n",
       "      <th>avg_training_score</th>\n",
       "      <th>is_promoted</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>29934</th>\n",
       "      <td>Technology</td>\n",
       "      <td>region_23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>77</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33332</th>\n",
       "      <td>Operations</td>\n",
       "      <td>region_15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>41</td>\n",
       "      <td>4.0</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>57</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35465</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>f</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>48</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17423</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>3</td>\n",
       "      <td>24</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>48</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66013</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>2</td>\n",
       "      <td>25</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>53</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    department     region education gender  \\\n",
       "employee_id                                                  \n",
       "29934               Technology  region_23       NaN      m   \n",
       "33332               Operations  region_15       NaN      m   \n",
       "35465        Sales & Marketing   region_7       NaN      f   \n",
       "17423        Sales & Marketing   region_2       NaN      m   \n",
       "66013        Sales & Marketing   region_2       NaN      m   \n",
       "\n",
       "            recruitment_channel  no_of_trainings  age  previous_year_rating  \\\n",
       "employee_id                                                                   \n",
       "29934                  sourcing                1   30                   NaN   \n",
       "33332                  sourcing                1   41                   4.0   \n",
       "35465                  sourcing                1   24                   1.0   \n",
       "17423                     other                3   24                   2.0   \n",
       "66013                  sourcing                2   25                   3.0   \n",
       "\n",
       "             length_of_service  is_kpi_met  has_won_awards  \\\n",
       "employee_id                                                  \n",
       "29934                        1           0               0   \n",
       "33332                       11           0               0   \n",
       "35465                        2           0               0   \n",
       "17423                        2           0               0   \n",
       "66013                        2           0               0   \n",
       "\n",
       "             avg_training_score  is_promoted  \n",
       "employee_id                                   \n",
       "29934                        77            0  \n",
       "33332                        57            0  \n",
       "35465                        48            0  \n",
       "17423                        48            0  \n",
       "66013                        53            0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.education.isnull() == True].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:01:32.445907Z",
     "start_time": "2018-09-14T04:01:32.386070Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Is there anyone who is promoted whose education is null?? :  122 folks!\n"
     ]
    },
    {
     "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>no_of_trainings</th>\n",
       "      <td>122.0</td>\n",
       "      <td>1.114754</td>\n",
       "      <td>0.533145</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>122.0</td>\n",
       "      <td>29.172131</td>\n",
       "      <td>6.729352</td>\n",
       "      <td>20.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>previous_year_rating</th>\n",
       "      <td>97.0</td>\n",
       "      <td>4.061856</td>\n",
       "      <td>1.028899</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>length_of_service</th>\n",
       "      <td>122.0</td>\n",
       "      <td>3.950820</td>\n",
       "      <td>3.177536</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_kpi_met</th>\n",
       "      <td>122.0</td>\n",
       "      <td>0.704918</td>\n",
       "      <td>0.457960</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>has_won_awards</th>\n",
       "      <td>122.0</td>\n",
       "      <td>0.163934</td>\n",
       "      <td>0.371743</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>avg_training_score</th>\n",
       "      <td>122.0</td>\n",
       "      <td>68.008197</td>\n",
       "      <td>17.208731</td>\n",
       "      <td>45.0</td>\n",
       "      <td>50.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>98.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>is_promoted</th>\n",
       "      <td>122.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      count       mean        std   min   25%   50%   75%  \\\n",
       "no_of_trainings       122.0   1.114754   0.533145   1.0   1.0   1.0   1.0   \n",
       "age                   122.0  29.172131   6.729352  20.0  24.0  28.0  32.0   \n",
       "previous_year_rating   97.0   4.061856   1.028899   1.0   3.0   4.0   5.0   \n",
       "length_of_service     122.0   3.950820   3.177536   1.0   2.0   3.0   5.0   \n",
       "is_kpi_met            122.0   0.704918   0.457960   0.0   0.0   1.0   1.0   \n",
       "has_won_awards        122.0   0.163934   0.371743   0.0   0.0   0.0   0.0   \n",
       "avg_training_score    122.0  68.008197  17.208731  45.0  50.0  64.0  85.0   \n",
       "is_promoted           122.0   1.000000   0.000000   1.0   1.0   1.0   1.0   \n",
       "\n",
       "                       max  \n",
       "no_of_trainings        6.0  \n",
       "age                   60.0  \n",
       "previous_year_rating   5.0  \n",
       "length_of_service     16.0  \n",
       "is_kpi_met             1.0  \n",
       "has_won_awards         1.0  \n",
       "avg_training_score    98.0  \n",
       "is_promoted            1.0  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Is there anyone who is promoted whose education is null??\n",
    "mask = (df.education.isnull() == True) & (df.is_promoted == 1)\n",
    "print('Is there anyone who is promoted whose education is null?? : ', df[mask].shape[0], 'folks!')\n",
    "df[mask].describe().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:12.003539Z",
     "start_time": "2018-09-14T04:02:11.984590Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[\"Master's & above\", \"Bachelor's\", nan, 'Below Secondary']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(df.education.unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:29:22.730389Z",
     "start_time": "2018-09-14T04:29:22.410209Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x216 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f = plt.figure(figsize=(8,3))\n",
    "ax = f.add_subplot(1, 2, 1)\n",
    "sns.countplot(x='education', data=df)    \n",
    "\n",
    "with sns.axes_style(\"darkgrid\"):    \n",
    "    f.add_subplot(1, 2, 2)\n",
    "    sns.countplot(x='education', data=fillna_pipe.fit_transform(df))    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:35.837394Z",
     "start_time": "2018-09-14T04:02:35.656876Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25a14724c18>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# sns.countplot(x='previous_year_rating', data=df)\n",
    "sns.countplot(x='previous_year_rating', data=fillna_pipe.fit_transform(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:39.089521Z",
     "start_time": "2018-09-14T04:02:39.063622Z"
    }
   },
   "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "      <th>no_of_trainings</th>\n",
       "      <th>age</th>\n",
       "      <th>previous_year_rating</th>\n",
       "      <th>length_of_service</th>\n",
       "      <th>is_kpi_met</th>\n",
       "      <th>has_won_awards</th>\n",
       "      <th>avg_training_score</th>\n",
       "      <th>is_promoted</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>29934</th>\n",
       "      <td>Technology</td>\n",
       "      <td>region_23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>77</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71177</th>\n",
       "      <td>Procurement</td>\n",
       "      <td>region_5</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>70</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74759</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_4</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>44</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45709</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_31</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>f</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26599</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_16</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>2</td>\n",
       "      <td>27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>47</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    department     region   education gender  \\\n",
       "employee_id                                                    \n",
       "29934               Technology  region_23         NaN      m   \n",
       "71177              Procurement   region_5  Bachelor's      m   \n",
       "74759        Sales & Marketing   region_4  Bachelor's      m   \n",
       "45709        Sales & Marketing  region_31  Bachelor's      f   \n",
       "26599        Sales & Marketing  region_16  Bachelor's      m   \n",
       "\n",
       "            recruitment_channel  no_of_trainings  age  previous_year_rating  \\\n",
       "employee_id                                                                   \n",
       "29934                  sourcing                1   30                   NaN   \n",
       "71177                     other                1   27                   NaN   \n",
       "74759                  sourcing                1   26                   NaN   \n",
       "45709                     other                1   29                   NaN   \n",
       "26599                     other                2   27                   NaN   \n",
       "\n",
       "             length_of_service  is_kpi_met  has_won_awards  \\\n",
       "employee_id                                                  \n",
       "29934                        1           0               0   \n",
       "71177                        1           0               0   \n",
       "74759                        1           0               0   \n",
       "45709                        1           0               0   \n",
       "26599                        1           1               0   \n",
       "\n",
       "             avg_training_score  is_promoted  \n",
       "employee_id                                   \n",
       "29934                        77            0  \n",
       "71177                        70            0  \n",
       "74759                        44            0  \n",
       "45709                        49            0  \n",
       "26599                        47            0  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.previous_year_rating.isnull() == True].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:40.951035Z",
     "start_time": "2018-09-14T04:02:40.939067Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Is there anyone who is promoted whose previous_year_rating is null?? :  339 folks!\n"
     ]
    }
   ],
   "source": [
    "# Is there anyone who is promoted whose previous_year_rating is null??\n",
    "mask = (df.previous_year_rating.isnull() == True) & (df.is_promoted == 1)\n",
    "print('Is there anyone who is promoted whose previous_year_rating is null?? : ', df[mask].shape[0], 'folks!')\n",
    "# df[mask].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:43.850777Z",
     "start_time": "2018-09-14T04:02:43.840830Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Is there anyone whose length_of_service>1  with previous_year_rating as null?? :  0 folks!\n"
     ]
    }
   ],
   "source": [
    "# Is there anyone whose length_of_service>1  with previous_year_rating as null??\n",
    "mask = (df.previous_year_rating.isnull() == True) & (df.length_of_service > 1)\n",
    "print('Is there anyone whose length_of_service>1  with previous_year_rating as null?? : ', df[mask].shape[0], 'folks!')\n",
    "# df[mask].head()\n",
    "# Observation\n",
    "# If only length_of_service=1, the previous_year_rating = NaN\n",
    "# Possibility of FeatureEngineering?? length_of_service*previous_year_rating, setting NaN=0 \n",
    "# [Could be wrong bcoz only last rating is available. What if last ratig is bad but exp is more?]\n",
    "# How does this look for test dataset?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:02:45.959817Z",
     "start_time": "2018-09-14T04:02:45.950841Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Is there anyone whose length_of_service>1  with previous_year_rating as null in TEST DATASET?? :  0 folks!\n"
     ]
    }
   ],
   "source": [
    "# How does this look for test dataset?\n",
    "# Is there anyone whose length_of_service>1  with previous_year_rating as null??\n",
    "mask = (test.previous_year_rating.isnull() == True) & (test.length_of_service > 1)\n",
    "print('Is there anyone whose length_of_service>1  with previous_year_rating as null in TEST DATASET?? : ', test[mask].shape[0], 'folks!')\n",
    "# None, so we can set the value as Zero"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:04:23.756301Z",
     "start_time": "2018-09-14T04:04:23.579774Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x25a148f5278>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x='gender', data=df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:04:28.713995Z",
     "start_time": "2018-09-14T04:04:28.381666Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Possible Regions :  ['region_7' 'region_22' 'region_19' 'region_23' 'region_26' 'region_2'\n",
      " 'region_20' 'region_34' 'region_1' 'region_4' 'region_29' 'region_31'\n",
      " 'region_15' 'region_14' 'region_11' 'region_5' 'region_28' 'region_17'\n",
      " 'region_13' 'region_16' 'region_25' 'region_10' 'region_27' 'region_30'\n",
      " 'region_12' 'region_21' 'region_8' 'region_32' 'region_6' 'region_33'\n",
      " 'region_24' 'region_3' 'region_9' 'region_18']\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
       "        17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33]),\n",
       " <a list of 34 Text xticklabel objects>)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('Possible Regions : ', df.region.unique())\n",
    "ax = sns.countplot(x='region', data=df)\n",
    "# ax.set_xticklabels(ax.get_xticklabels(), rotation=90)\n",
    "plt.xticks(rotation=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:10:27.715990Z",
     "start_time": "2018-09-14T04:10:27.639194Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 54808 entries, 65438 to 51526\n",
      "Data columns (total 14 columns):\n",
      "department              54808 non-null object\n",
      "region                  54808 non-null object\n",
      "education               52399 non-null object\n",
      "gender                  54808 non-null object\n",
      "recruitment_channel     54808 non-null object\n",
      "no_of_trainings         54808 non-null int64\n",
      "age                     54808 non-null int64\n",
      "previous_year_rating    50684 non-null float64\n",
      "length_of_service       54808 non-null int64\n",
      "is_kpi_met              54808 non-null int64\n",
      "has_won_awards          54808 non-null int64\n",
      "avg_training_score      54808 non-null int64\n",
      "is_promoted             54808 non-null int64\n",
      "iregion                 54808 non-null int8\n",
      "dtypes: float64(1), int64(7), int8(1), object(5)\n",
      "memory usage: 8.4+ MB\n"
     ]
    }
   ],
   "source": [
    "df['iregion'] = df['region'].apply(lambda x: x.split('_')[1])\n",
    "df['iregion'] = df['iregion'].astype(int)\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:14:46.028872Z",
     "start_time": "2018-09-14T04:14:46.010918Z"
    }
   },
   "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "      <th>no_of_trainings</th>\n",
       "      <th>age</th>\n",
       "      <th>previous_year_rating</th>\n",
       "      <th>length_of_service</th>\n",
       "      <th>is_kpi_met</th>\n",
       "      <th>has_won_awards</th>\n",
       "      <th>avg_training_score</th>\n",
       "      <th>is_promoted</th>\n",
       "      <th>iregion</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>65438</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_7</td>\n",
       "      <td>Master's &amp; above</td>\n",
       "      <td>f</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>35</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>49</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65141</th>\n",
       "      <td>Operations</td>\n",
       "      <td>region_22</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>60</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7513</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_19</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>sourcing</td>\n",
       "      <td>1</td>\n",
       "      <td>34</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2542</th>\n",
       "      <td>Sales &amp; Marketing</td>\n",
       "      <td>region_23</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>0</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48945</th>\n",
       "      <td>Technology</td>\n",
       "      <td>region_26</td>\n",
       "      <td>Bachelor's</td>\n",
       "      <td>m</td>\n",
       "      <td>other</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>73</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    department     region         education gender  \\\n",
       "employee_id                                                          \n",
       "65438        Sales & Marketing   region_7  Master's & above      f   \n",
       "65141               Operations  region_22        Bachelor's      m   \n",
       "7513         Sales & Marketing  region_19        Bachelor's      m   \n",
       "2542         Sales & Marketing  region_23        Bachelor's      m   \n",
       "48945               Technology  region_26        Bachelor's      m   \n",
       "\n",
       "            recruitment_channel  no_of_trainings  age  previous_year_rating  \\\n",
       "employee_id                                                                   \n",
       "65438                  sourcing                1   35                   5.0   \n",
       "65141                     other                1   30                   5.0   \n",
       "7513                   sourcing                1   34                   3.0   \n",
       "2542                      other                2   39                   1.0   \n",
       "48945                     other                1   45                   3.0   \n",
       "\n",
       "             length_of_service  is_kpi_met  has_won_awards  \\\n",
       "employee_id                                                  \n",
       "65438                        8           1               0   \n",
       "65141                        4           0               0   \n",
       "7513                         7           0               0   \n",
       "2542                        10           0               0   \n",
       "48945                        2           0               0   \n",
       "\n",
       "             avg_training_score  is_promoted  iregion  \n",
       "employee_id                                            \n",
       "65438                        49            0        7  \n",
       "65141                        60            0       22  \n",
       "7513                         50            0       19  \n",
       "2542                         50            0       23  \n",
       "48945                        73            0       26  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:18:41.191862Z",
     "start_time": "2018-09-14T04:18:41.152966Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 54808 entries, 65438 to 51526\n",
      "Data columns (total 14 columns):\n",
      "department              54808 non-null object\n",
      "region                  54808 non-null object\n",
      "education               52399 non-null object\n",
      "gender                  54808 non-null object\n",
      "recruitment_channel     54808 non-null object\n",
      "no_of_trainings         54808 non-null int64\n",
      "age                     54808 non-null int64\n",
      "previous_year_rating    50684 non-null float64\n",
      "length_of_service       54808 non-null int64\n",
      "is_kpi_met              54808 non-null int64\n",
      "has_won_awards          54808 non-null int64\n",
      "avg_training_score      54808 non-null int8\n",
      "is_promoted             54808 non-null int8\n",
      "iregion                 54808 non-null int8\n",
      "dtypes: float64(1), int64(5), int8(3), object(5)\n",
      "memory usage: 7.7+ MB\n"
     ]
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T04:34:04.426117Z",
     "start_time": "2018-09-14T04:34:04.328432Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 54808 entries, 65438 to 51526\n",
      "Data columns (total 14 columns):\n",
      "department              54808 non-null object\n",
      "region                  54808 non-null object\n",
      "education               52399 non-null object\n",
      "gender                  54808 non-null object\n",
      "recruitment_channel     54808 non-null object\n",
      "no_of_trainings         54808 non-null int8\n",
      "age                     54808 non-null int8\n",
      "previous_year_rating    50684 non-null float64\n",
      "length_of_service       54808 non-null int8\n",
      "is_kpi_met              54808 non-null int8\n",
      "has_won_awards          54808 non-null int8\n",
      "avg_training_score      54808 non-null int8\n",
      "is_promoted             54808 non-null int8\n",
      "iregion                 54808 non-null int8\n",
      "dtypes: float64(1), int8(8), object(5)\n",
      "memory usage: 5.8+ MB\n",
      "None\n",
      "\n",
      "After processing:\n",
      "\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 54808 entries, 65438 to 51526\n",
      "Data columns (total 14 columns):\n",
      "department              54808 non-null category\n",
      "region                  54808 non-null category\n",
      "education               52399 non-null category\n",
      "gender                  54808 non-null category\n",
      "recruitment_channel     54808 non-null category\n",
      "no_of_trainings         54808 non-null int8\n",
      "age                     54808 non-null int8\n",
      "previous_year_rating    50684 non-null float16\n",
      "length_of_service       54808 non-null int8\n",
      "is_kpi_met              54808 non-null int8\n",
      "has_won_awards          54808 non-null int8\n",
      "avg_training_score      54808 non-null int8\n",
      "is_promoted             54808 non-null int8\n",
      "iregion                 54808 non-null int8\n",
      "dtypes: category(5), float16(1), int8(8)\n",
      "memory usage: 3.7 MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "print(df.info())\n",
    "\n",
    "df['iregion'] = df['iregion'].astype(np.int8)\n",
    "df['is_promoted'] = df['is_promoted'].astype(np.int8)\n",
    "df['avg_training_score'] = df['avg_training_score'].astype(np.int8)\n",
    "df['has_won_awards'] = df['has_won_awards'].astype(np.int8)\n",
    "df['is_kpi_met'] = df['is_kpi_met'].astype(np.int8)\n",
    "df['length_of_service'] = df['length_of_service'].astype(np.int8)\n",
    "df['age'] = df['age'].astype(np.int8)\n",
    "df['no_of_trainings'] = df['no_of_trainings'].astype(np.int8)\n",
    "\n",
    "df['previous_year_rating'] = df['previous_year_rating'].astype(np.float16)\n",
    "\n",
    "df['department'] = df['department'].astype('category')\n",
    "df['region'] = df['region'].astype('category')\n",
    "df['education'] = df['education'].astype('category')\n",
    "df['gender'] = df['gender'].astype('category')\n",
    "df['recruitment_channel'] = df['recruitment_channel'].astype('category')\n",
    "\n",
    "print('\\nAfter processing:\\n',)\n",
    "print(df.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T05:30:50.707911Z",
     "start_time": "2018-09-14T05:30:50.701924Z"
    }
   },
   "outputs": [],
   "source": [
    "class RatingsImputer(BaseEstimator, TransformerMixin):\n",
    "    def fit (self, X, y=None, **fit_params):\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X):\n",
    "        assert isinstance(X, pd.DataFrame)\n",
    "        #X.loc[X.previous_year_rating.isnull(), 'previous_year_rating'] = 0\n",
    "        X = X.fillna({'previous_year_rating':0})\n",
    "        return X\n",
    "\n",
    "class EducationCleanser(BaseEstimator, TransformerMixin):\n",
    "    def fit (self, X, y=None, **fit_params):\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X):\n",
    "        assert isinstance(X, pd.DataFrame)\n",
    "        X = X.fillna({'education':\"Bachelor's\"})\n",
    "        return X\n",
    "\n",
    "# fillna_pipe.fit_transform(df).isnull().sum() # For Testing purpose only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T07:45:06.296019Z",
     "start_time": "2018-09-14T07:45:06.260084Z"
    }
   },
   "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>department</th>\n",
       "      <th>region</th>\n",
       "      <th>education</th>\n",
       "      <th>gender</th>\n",
       "      <th>recruitment_channel</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>employee_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>65438</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65141</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7513</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2542</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48945</th>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             department  region  education  gender  recruitment_channel\n",
       "employee_id                                                            \n",
       "65438                 1       4          1       0                    1\n",
       "65141                 0       1          0       1                    0\n",
       "7513                  1       0          0       1                    1\n",
       "2542                  1       2          0       1                    0\n",
       "48945                 2       3          0       1                    0"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class ColumnSelector(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, columns=None):\n",
    "        self.columns = columns\n",
    "    \n",
    "    def fit (self, X, y=None, **fit_params):\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X):\n",
    "        assert isinstance(X, pd.DataFrame)\n",
    "        try:\n",
    "            return X[self.columns]\n",
    "        except KeyError:\n",
    "            cols_error = list(set(self.columns) - set(X.columns))\n",
    "            raise KeyError(\"The DataFrame does not include the columns: %s\" % cols_error)\n",
    "\n",
    "# class ModifiedLabelEncoder(LabelEncoder):\n",
    "#     def fit_transform(self, y, *args, **kwargs):\n",
    "#         print(type(y))\n",
    "#         print(y.head(5))\n",
    "#         return super().fit_transform(y).reshape(-1, 1)\n",
    "\n",
    "#     def transform(self, y, *args, **kwargs):\n",
    "#         return super().transform(y).reshape(-1, 1)\n",
    "\n",
    "class MultiColumnLabelEncoder:\n",
    "    def __init__(self,columns = None):\n",
    "        self.columns = columns # array of column names to encode\n",
    "\n",
    "    def fit(self,X,y=None):\n",
    "        return self # not relevant here\n",
    "\n",
    "    def transform(self,X):\n",
    "        '''\n",
    "        Transforms columns of X specified in self.columns using LabelEncoder(). \n",
    "        If no columns specified, transforms all columns in X.\n",
    "        '''\n",
    "        output = X.copy()\n",
    "        if self.columns is not None:\n",
    "            for col in self.columns:\n",
    "                output[col] = LabelEncoder().fit_transform(output[col])\n",
    "        else:\n",
    "            for colname,col in output.iteritems():\n",
    "                output[colname] = LabelEncoder().fit_transform(col)\n",
    "        return output\n",
    "\n",
    "    def fit_transform(self,X,y=None):\n",
    "        return self.fit(X,y).transform(X)\n",
    "\n",
    "categorical_cols = ['department', 'region', 'education', 'gender', 'recruitment_channel'] # list of categorical column names    \n",
    "numeric_cols = [ col for col in df.columns.difference(categorical_cols) if col not in ['is_promoted', 'iregion']]\n",
    "\n",
    "fillna_pipe = make_pipeline(RatingsImputer(), EducationCleanser())\n",
    "# categorical_cols_pipe = make_pipeline(ColumnSelector(columns=categorical_cols), ModifiedLabelEncoder(), OneHotEncoder(sparse=False))\n",
    "categorical_cols_pipe = make_pipeline(ColumnSelector(columns=categorical_cols), MultiColumnLabelEncoder())\n",
    "numeric_cols_pipe = make_pipeline(ColumnSelector(columns=numeric_cols),MinMaxScaler())\n",
    "\n",
    "# categorical_cols_pipe.fit_transform(df.head(5))\n",
    "# numeric_cols_pipe.fit_transform(df.head(5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T08:17:50.918668Z",
     "start_time": "2018-09-14T08:17:50.684659Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(54808, 12)\n",
      "[[ 0.375       0.16666667  0.          1.          0.19444444  0.\n",
      "   1.          7.         31.          2.          0.          2.        ]\n",
      " [ 0.25        0.35        0.          0.          0.08333333  0.\n",
      "   1.          4.         14.          0.          1.          0.        ]\n",
      " [ 0.35        0.18333333  0.          0.          0.16666667  0.\n",
      "   0.6         7.         10.          0.          1.          2.        ]]\n"
     ]
    }
   ],
   "source": [
    "# [ col for col in df.columns.difference(categorical_cols) if col not in ['is_promoted', 'iregion']]\n",
    "fu = make_union(numeric_cols_pipe, categorical_cols_pipe)\n",
    "pipe = make_pipeline(fillna_pipe, fu)\n",
    "train_vec = pipe.fit_transform(df)\n",
    "print(train_vec.shape)\n",
    "print(train_vec[:3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T08:30:28.810785Z",
     "start_time": "2018-09-14T08:30:28.785823Z"
    }
   },
   "outputs": [],
   "source": [
    "# Split train data-set\n",
    "from sklearn.model_selection import train_test_split\n",
    "x_train, x_test, y_train, y_test = train_test_split(train_vec, df.is_promoted, train_size = 0.75, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T08:14:09.092398Z",
     "start_time": "2018-09-14T08:14:08.979700Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.1       ,  0.63333333,  0.        ,  1.        ,  0.        ,\n",
       "         0.        ,  0.        ,  8.        , 18.        ,  0.        ,\n",
       "         1.        ,  2.        ],\n",
       "       [ 0.275     ,  0.2       ,  0.        ,  0.        ,  0.11111111,\n",
       "         0.        ,  0.6       ,  2.        , 28.        ,  0.        ,\n",
       "         0.        ,  0.        ],\n",
       "       [ 0.275     ,  0.13333333,  0.        ,  0.        ,  0.08333333,\n",
       "         0.        ,  0.2       ,  7.        ,  4.        ,  0.        ,\n",
       "         1.        ,  0.        ],\n",
       "       [ 0.275     ,  0.43333333,  0.        ,  0.        ,  0.22222222,\n",
       "         0.22222222,  0.4       ,  5.        , 11.        ,  0.        ,\n",
       "         0.        ,  0.        ],\n",
       "       [ 0.25      ,  0.36666667,  0.        ,  0.        ,  0.16666667,\n",
       "         0.        ,  0.8       ,  1.        , 21.        ,  0.        ,\n",
       "         1.        ,  2.        ]])"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_vec = pipe.transform(test)\n",
    "test_vec[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SGDClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T09:16:53.197691Z",
     "start_time": "2018-09-14T09:16:53.092969Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.27378641, 0.25193108, 0.03135314])"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import SGDClassifier\n",
    "from sklearn.model_selection import cross_val_score, cross_validate\n",
    "\n",
    "# Ref.: https://stackoverflow.com/questions/37413302/f1-score-of-all-classes-from-scikits-cross-val-score\n",
    "# Ref.: https://stackoverflow.com/questions/35876508/evaluate-multiple-scores-on-sklearn-cross-val-score\n",
    "svm_clf = SGDClassifier()\n",
    "scoring = {'acc': 'accuracy',\n",
    "           'prec_macro': 'precision_macro',\n",
    "           'rec_micro': 'recall_macro'}\n",
    "# scores = cross_val_score(svm_clf, x_train, y_train, cv=3, scoring='precision')\n",
    "# scores = cross_val_score(svm_clf, x_train, y_train, cv=3, scoring='recall')\n",
    "scores = cross_val_score(svm_clf, x_train, y_train, cv=3, scoring='f1')\n",
    "# scores = cross_validate(svm_clf, x_train, y_train, cv=3, scoring=scoring)\n",
    "scores # Check against the recall score, for this is something the  Public LeaderBoard is scoring against."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T08:56:50.581640Z",
     "start_time": "2018-09-14T08:56:50.547731Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy =  0.9162166107137644\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "Not Promoted       0.92      1.00      0.96     12571\n",
      "    Promoted       0.39      0.03      0.05      1131\n",
      "\n",
      " avg / total       0.88      0.92      0.88     13702\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[12524,    47],\n",
       "       [ 1101,    30]], dtype=int64)"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "\n",
    "svm_clf.fit(x_train, y_train)\n",
    "preds = svm_clf.predict(x_test)\n",
    "\n",
    "print('Accuracy = ', np.mean(preds == y_test))\n",
    "\n",
    "creport = classification_report(y_test, preds, target_names=['Not Promoted', 'Promoted'])\n",
    "print(creport)\n",
    "\n",
    "confusion_matrix(y_test, preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T09:01:44.651818Z",
     "start_time": "2018-09-14T09:01:44.645835Z"
    }
   },
   "outputs": [],
   "source": [
    "def conv2df(preds):\n",
    "    df = pd.DataFrame(data={\n",
    "        'employee_id': test.index.values,\n",
    "        'is_promoted': preds\n",
    "    })\n",
    "    df['is_promoted'] = df['is_promoted'].astype(np.int8)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2018-09-14T09:01:48.184355Z",
     "start_time": "2018-09-14T09:01:46.740220Z"
    }
   },
   "outputs": [],
   "source": [
    "conv2df(svm_clf.predict(test_vec)).to_csv('data/output/SGDClassifier_default.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Possible areas of improvement\n",
    "* Is there a better  way to fill missing values for the 'education' column???"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}