Arvato_Project_Workbook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Capstone Project: Create a Customer Segmentation Report for Arvato Financial Services\n",
    "\n",
    "In this project, I will analyze demographics data for customers of a mail-order sales company in Germany, comparing it against demographics information for the general population. I'll use unsupervised learning techniques to perform customer segmentation, identifying the parts of the population that best describe the core customer base of the company. Then, I'll apply what I've learned on a third dataset with demographics information for targets of a marketing campaign for the company, and use a model to predict which individuals are most likely to convert into becoming customers for the company. The data that I will use has been provided by Bertelsmann Arvato Analytics, and represents a real-life data science task.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Table of Contents\n",
    "\n",
    "[Part 0: Get-to-Know-the-Data](#Get-to-Know-the-Data)<br>\n",
    "[Part 1: Customer Segmentation Report](#part1)<br>\n",
    "[Part 2: Supervised Learning Model](#part2)<br>\n",
    "[Part 3: Kaggle Competition](#part3)<br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Import libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "import os\n",
    "import random\n",
    "import progressbar\n",
    "\n",
    "from sklearn.preprocessing import Imputer\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "import xgboost as xgb\n",
    "from sklearn.ensemble import AdaBoostRegressor \n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from xgboost.sklearn import XGBRegressor\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import learning_curve\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "from sklearn import metrics\n",
    "\n",
    "# Pretty display for notebooks\n",
    "%matplotlib inline\n",
    "%config InlineBachend.figure_format = 'retina'\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <a class=\"anchor\" id=\"Get-to-Know-the-Data\">Part 0: Get to Know the Data</a>\n",
    "\n",
    "There are four data files associated with this project:\n",
    "\n",
    "- `Udacity_AZDIAS_052018.csv`: Demographics data for the general population of Germany; 891 211 persons (rows) x 366 features (columns).\n",
    "- `Udacity_CUSTOMERS_052018.csv`: Demographics data for customers of a mail-order company; 191 652 persons (rows) x 369 features (columns).\n",
    "- `Udacity_MAILOUT_052018_TRAIN.csv`: Demographics data for individuals who were targets of a marketing campaign; 42 982 persons (rows) x 367 (columns).\n",
    "- `Udacity_MAILOUT_052018_TEST.csv`: Demographics data for individuals who were targets of a marketing campaign; 42 833 persons (rows) x 366 (columns).\n",
    "\n",
    "Each row of the demographics files represents a single person, but also includes information outside of individuals, including information about their household, building, and neighborhood. I will use the information from the first two files to figure out how customers (\"CUSTOMERS\") are similar to or differ from the general population at large (\"AZDIAS\"), then use my analysis to make predictions on the other two files (\"MAILOUT\"), predicting which recipients are most likely to become a customer for the mail-order company.\n",
    "\n",
    "The \"CUSTOMERS\" file contains three extra columns ('CUSTOMER_GROUP', 'ONLINE_PURCHASE', and 'PRODUCT_GROUP'), which provide broad information about the customers depicted in the file. The original \"MAILOUT\" file included one additional column, \"RESPONSE\", which indicated whether or not each recipient became a customer of the company. For the \"TRAIN\" subset, this column has been retained, but in the \"TEST\" subset it has been removed; it is against that withheld column that will be assessed in the Kaggle competition.\n",
    "\n",
    "Otherwise, all of the remaining columns are the same between the three data files. For more information about the columns depicted in the files, I will refer to two Excel spreadsheets provided in the workspace. [One of them](./DIAS Information Levels - Attributes 2017.xlsx) is a top-level list of attributes and descriptions, organized by informational category. [The other](./DIAS Attributes - Values 2017.xlsx) is a detailed mapping of data values for each feature in alphabetical order.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (18,19) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\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>LNR</th>\n",
       "      <th>AGER_TYP</th>\n",
       "      <th>AKT_DAT_KL</th>\n",
       "      <th>ALTER_HH</th>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <th>ANZ_HAUSHALTE_AKTIV</th>\n",
       "      <th>...</th>\n",
       "      <th>VHN</th>\n",
       "      <th>VK_DHT4A</th>\n",
       "      <th>VK_DISTANZ</th>\n",
       "      <th>VK_ZG11</th>\n",
       "      <th>W_KEIT_KIND_HH</th>\n",
       "      <th>WOHNDAUER_2008</th>\n",
       "      <th>WOHNLAGE</th>\n",
       "      <th>ZABEOTYP</th>\n",
       "      <th>ANREDE_KZ</th>\n",
       "      <th>ALTERSKATEGORIE_GROB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>910215</td>\n",
       "      <td>-1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>910220</td>\n",
       "      <td>-1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>910225</td>\n",
       "      <td>-1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>910226</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>910241</td>\n",
       "      <td>-1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 366 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      LNR  AGER_TYP  AKT_DAT_KL  ALTER_HH  ALTER_KIND1  ALTER_KIND2  \\\n",
       "0  910215        -1         NaN       NaN          NaN          NaN   \n",
       "1  910220        -1         9.0       0.0          NaN          NaN   \n",
       "2  910225        -1         9.0      17.0          NaN          NaN   \n",
       "3  910226         2         1.0      13.0          NaN          NaN   \n",
       "4  910241        -1         1.0      20.0          NaN          NaN   \n",
       "\n",
       "   ALTER_KIND3  ALTER_KIND4  ALTERSKATEGORIE_FEIN  ANZ_HAUSHALTE_AKTIV  ...  \\\n",
       "0          NaN          NaN                   NaN                  NaN  ...   \n",
       "1          NaN          NaN                  21.0                 11.0  ...   \n",
       "2          NaN          NaN                  17.0                 10.0  ...   \n",
       "3          NaN          NaN                  13.0                  1.0  ...   \n",
       "4          NaN          NaN                  14.0                  3.0  ...   \n",
       "\n",
       "   VHN  VK_DHT4A  VK_DISTANZ  VK_ZG11  W_KEIT_KIND_HH  WOHNDAUER_2008  \\\n",
       "0  NaN       NaN         NaN      NaN             NaN             NaN   \n",
       "1  4.0       8.0        11.0     10.0             3.0             9.0   \n",
       "2  2.0       9.0         9.0      6.0             3.0             9.0   \n",
       "3  0.0       7.0        10.0     11.0             NaN             9.0   \n",
       "4  2.0       3.0         5.0      4.0             2.0             9.0   \n",
       "\n",
       "   WOHNLAGE ZABEOTYP ANREDE_KZ ALTERSKATEGORIE_GROB  \n",
       "0       NaN        3         1                    2  \n",
       "1       4.0        5         2                    1  \n",
       "2       2.0        5         2                    3  \n",
       "3       7.0        3         2                    4  \n",
       "4       3.0        4         1                    3  \n",
       "\n",
       "[5 rows x 366 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load data\n",
    "azdias = pd.read_csv('Arvato_data/Udacity_AZDIAS_052018.csv', sep=';')\n",
    "\n",
    "azdias.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(891221, 366)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Data shape\n",
    "azdias.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891221 entries, 0 to 891220\n",
      "Columns: 366 entries, LNR to ALTERSKATEGORIE_GROB\n",
      "dtypes: float64(267), int64(93), object(6)\n",
      "memory usage: 2.4+ GB\n"
     ]
    }
   ],
   "source": [
    "# Data info\n",
    "azdias.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": 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>CAMEO_DEU_2015</th>\n",
       "      <th>CAMEO_DEUG_2015</th>\n",
       "      <th>CAMEO_INTL_2015</th>\n",
       "      <th>D19_LETZTER_KAUF_BRANCHE</th>\n",
       "      <th>EINGEFUEGT_AM</th>\n",
       "      <th>OST_WEST_KZ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8A</td>\n",
       "      <td>8</td>\n",
       "      <td>51</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1992-02-10 00:00:00</td>\n",
       "      <td>W</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4C</td>\n",
       "      <td>4</td>\n",
       "      <td>24</td>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "      <td>W</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2A</td>\n",
       "      <td>2</td>\n",
       "      <td>12</td>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1997-04-21 00:00:00</td>\n",
       "      <td>W</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6B</td>\n",
       "      <td>6</td>\n",
       "      <td>43</td>\n",
       "      <td>D19_SCHUHE</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "      <td>W</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  CAMEO_DEU_2015 CAMEO_DEUG_2015 CAMEO_INTL_2015 D19_LETZTER_KAUF_BRANCHE  \\\n",
       "0            NaN             NaN             NaN                      NaN   \n",
       "1             8A               8              51                      NaN   \n",
       "2             4C               4              24            D19_UNBEKANNT   \n",
       "3             2A               2              12            D19_UNBEKANNT   \n",
       "4             6B               6              43               D19_SCHUHE   \n",
       "\n",
       "         EINGEFUEGT_AM OST_WEST_KZ  \n",
       "0                  NaN         NaN  \n",
       "1  1992-02-10 00:00:00           W  \n",
       "2  1992-02-12 00:00:00           W  \n",
       "3  1997-04-21 00:00:00           W  \n",
       "4  1992-02-12 00:00:00           W  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check object features\n",
    "df_object = azdias.select_dtypes(include='object')\n",
    "\n",
    "df_object.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <a class=\"anchor\" id=\"part1\">Part 1: Customer Segmentation Report</a>\n",
    "\n",
    "The main bulk of the analysis will come in this part of the project. Here, I will use unsupervised learning techniques to describe the relationship between the demographics of the company's existing customers and the general population of Germany. By the end of this part, I should be able to describe parts of the general population that are more likely to be part of the mail-order company's main customer base, and which parts of the general population are less so."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1: Preprocessing\n",
    "\n",
    "####  Step 1.1:  Assess Missing Data\n",
    "\n",
    "\n",
    "I will use the feature summary file from the previous project and 'DIAS Attributes - Values 2017.xlsx' file to build a data frame of columns properties in the demographics data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>information_level</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>person</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1,0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>person</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,0,9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ANREDE_KZ</td>\n",
       "      <td>person</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1,0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CJT_GESAMTTYP</td>\n",
       "      <td>person</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>FINANZ_MINIMALIST</td>\n",
       "      <td>person</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              attribute information_level         type missing_or_unknown\n",
       "0              AGER_TYP            person  categorical             [-1,0]\n",
       "1  ALTERSKATEGORIE_GROB            person      ordinal           [-1,0,9]\n",
       "2             ANREDE_KZ            person  categorical             [-1,0]\n",
       "3         CJT_GESAMTTYP            person  categorical                [0]\n",
       "4     FINANZ_MINIMALIST            person      ordinal               [-1]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load in the feature summary file.\n",
    "feat_info = pd.read_csv('Arvato_data/AZDIAS_Feature_Summary.csv', sep=';')\n",
    "feat_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(85, 4)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# feat_info shape\n",
    "feat_info.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ordinal        49\n",
       "categorical    21\n",
       "mixed           7\n",
       "numeric         7\n",
       "interval        1\n",
       "Name: type, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feat_info.type.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": 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>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>BIP_FLAG</td>\n",
       "      <td>business-flag indicating companies in the buil...</td>\n",
       "      <td>-1</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>CAMEO_DEUINTL_2015</td>\n",
       "      <td>CAMEO classification 2015 - international typo...</td>\n",
       "      <td>-1</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>D19_BANKEN_ANZ_12</td>\n",
       "      <td>transaction activity BANKS in the last 12 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>D19_BANKEN_ANZ_24</td>\n",
       "      <td>transaction activity BANKS in the last 24 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>D19_BANKEN_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Attribute                                        Description  \\\n",
       "48             BIP_FLAG  business-flag indicating companies in the buil...   \n",
       "105  CAMEO_DEUINTL_2015  CAMEO classification 2015 - international typo...   \n",
       "138   D19_BANKEN_ANZ_12   transaction activity BANKS in the last 12 months   \n",
       "145   D19_BANKEN_ANZ_24   transaction activity BANKS in the last 24 months   \n",
       "152    D19_BANKEN_DATUM  actuality of the last transaction for the segm...   \n",
       "\n",
       "    Value                                     Meaning  \n",
       "48     -1                                     unknown  \n",
       "105    -1                                     unknown  \n",
       "138     0                       no transactions known  \n",
       "145     0                       no transactions known  \n",
       "152     1  highest activity within the last 12 months  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Laod DIAS Attributes\n",
    "dias_attributes = pd.read_excel('Arvato_data/DIAS Attributes - Values 2017.xlsx',\n",
    "                                skiprows=[0])\n",
    "del dias_attributes['Unnamed: 0']\n",
    "\n",
    "# Drop NaN attribute\n",
    "dias_attributes = dias_attributes[~dias_attributes['Attribute'].isnull()]\n",
    "\n",
    "# Drop attribute described in feat_info\n",
    "dias_attributes = dias_attributes[\n",
    "    ~dias_attributes['Attribute'].isin(feat_info.attribute)\n",
    "]\n",
    "\n",
    "dias_attributes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(233, 4)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dias_attributes shape\n",
    "dias_attributes.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "unknown                                             172\n",
       "no transaction known                                 34\n",
       "highest activity within the last 12 months           12\n",
       "no transactions known                                10\n",
       "no Online-transactions within the last 12 months      3\n",
       "Universal                                             1\n",
       "very low car quote                                    1\n",
       "Name: Meaning, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# dias_attributes Meanning counts\n",
    "dias_attributes.Meaning.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": 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>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>765</th>\n",
       "      <td>KBA05_ALTER1</td>\n",
       "      <td>share of car owners less than 31 years old</td>\n",
       "      <td>-1, 9</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>771</th>\n",
       "      <td>KBA05_ALTER2</td>\n",
       "      <td>share of car owners inbetween 31 and 45 years ...</td>\n",
       "      <td>-1, 9</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>777</th>\n",
       "      <td>KBA05_ALTER3</td>\n",
       "      <td>share of car owners inbetween 45 and 60 years ...</td>\n",
       "      <td>-1, 9</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>783</th>\n",
       "      <td>KBA05_ALTER4</td>\n",
       "      <td>share of cars owners elder than 61 years</td>\n",
       "      <td>-1, 9</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>790</th>\n",
       "      <td>KBA05_ANHANG</td>\n",
       "      <td>share of trailers in the microcell</td>\n",
       "      <td>-1, 9</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Attribute                                        Description  Value  \\\n",
       "765  KBA05_ALTER1         share of car owners less than 31 years old  -1, 9   \n",
       "771  KBA05_ALTER2  share of car owners inbetween 31 and 45 years ...  -1, 9   \n",
       "777  KBA05_ALTER3  share of car owners inbetween 45 and 60 years ...  -1, 9   \n",
       "783  KBA05_ALTER4           share of cars owners elder than 61 years  -1, 9   \n",
       "790  KBA05_ANHANG                 share of trailers in the microcell  -1, 9   \n",
       "\n",
       "     Meaning  \n",
       "765  unknown  \n",
       "771  unknown  \n",
       "777  unknown  \n",
       "783  unknown  \n",
       "790  unknown  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Investigate attribute with unknown Meaning\n",
    "unknown_Meaning = dias_attributes[dias_attributes['Meaning'] == 'unknown']\n",
    "\n",
    "# Keep only attribute of azdias\n",
    "unknown_Meaning = unknown_Meaning[\n",
    "    unknown_Meaning['Attribute'].isin(azdias.columns)\n",
    "]\n",
    "\n",
    "# unknown_Meaning shape\n",
    "unknown_Meaning.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Get first part of attribute\n",
    "unknown_Meaning['first_att'] = unknown_Meaning.Attribute.apply(lambda x: x.split('_')[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KBA13    106\n",
       "KBA05     58\n",
       "Name: first_att, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unknown_Meaning.first_att.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 2.0    228625\n",
       " 1.0    167046\n",
       " 3.0    166129\n",
       "NaN     133324\n",
       " 0.0    102789\n",
       " 4.0     78522\n",
       " 9.0     14786\n",
       "Name: KBA05_ALTER1, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Value of KBA05_ALTER1  \n",
    "azdias.KBA05_ALTER1.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 3.0    361449\n",
       " 2.0    167076\n",
       " 4.0    158150\n",
       "NaN     105800\n",
       " 1.0     50939\n",
       " 5.0     47807\n",
       "Name: KBA13_VORB_1, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Values of KBA13_VORB_1  \n",
    "azdias.KBA13_VORB_1.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**KBA13 and KBA05 are ordinals variabeles**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "unknown_Meaning['type'] = 'ordinal'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": 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>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "      <th>first_att</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>D19_BANKEN_ANZ_12</td>\n",
       "      <td>transaction activity BANKS in the last 12 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>D19_BANKEN_ANZ_24</td>\n",
       "      <td>transaction activity BANKS in the last 24 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305</th>\n",
       "      <td>D19_GESAMT_ANZ_12</td>\n",
       "      <td>transaction activity TOTAL POOL in the last 12...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>D19_GESAMT_ANZ_24</td>\n",
       "      <td>transaction activity TOTAL POOL in the last 24...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>D19_TELKO_ANZ_12</td>\n",
       "      <td>transaction activity TELCO in the last 12 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>485</th>\n",
       "      <td>D19_TELKO_ANZ_24</td>\n",
       "      <td>transaction activity TELCO in the last 24 months</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>546</th>\n",
       "      <td>D19_VERSAND_ANZ_12</td>\n",
       "      <td>transaction activity MAIL-ORDER in the last 12...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>553</th>\n",
       "      <td>D19_VERSAND_ANZ_24</td>\n",
       "      <td>transaction activity MAIL-ORDER in the last 24...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>617</th>\n",
       "      <td>D19_VERSI_ANZ_12</td>\n",
       "      <td>transaction activity INSURANCE in the last 12 ...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>624</th>\n",
       "      <td>D19_VERSI_ANZ_24</td>\n",
       "      <td>transaction activity INSURANCE in the last 24 ...</td>\n",
       "      <td>0</td>\n",
       "      <td>no transactions known</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Attribute                                        Description  \\\n",
       "138   D19_BANKEN_ANZ_12   transaction activity BANKS in the last 12 months   \n",
       "145   D19_BANKEN_ANZ_24   transaction activity BANKS in the last 24 months   \n",
       "305   D19_GESAMT_ANZ_12  transaction activity TOTAL POOL in the last 12...   \n",
       "312   D19_GESAMT_ANZ_24  transaction activity TOTAL POOL in the last 24...   \n",
       "478    D19_TELKO_ANZ_12   transaction activity TELCO in the last 12 months   \n",
       "485    D19_TELKO_ANZ_24   transaction activity TELCO in the last 24 months   \n",
       "546  D19_VERSAND_ANZ_12  transaction activity MAIL-ORDER in the last 12...   \n",
       "553  D19_VERSAND_ANZ_24  transaction activity MAIL-ORDER in the last 24...   \n",
       "617    D19_VERSI_ANZ_12  transaction activity INSURANCE in the last 12 ...   \n",
       "624    D19_VERSI_ANZ_24  transaction activity INSURANCE in the last 24 ...   \n",
       "\n",
       "    Value                Meaning first_att  \n",
       "138     0  no transactions known       D19  \n",
       "145     0  no transactions known       D19  \n",
       "305     0  no transactions known       D19  \n",
       "312     0  no transactions known       D19  \n",
       "478     0  no transactions known       D19  \n",
       "485     0  no transactions known       D19  \n",
       "546     0  no transactions known       D19  \n",
       "553     0  no transactions known       D19  \n",
       "617     0  no transactions known       D19  \n",
       "624     0  no transactions known       D19  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Investigate attribute with no transaction known\n",
    "no_transaction = dias_attributes[dias_attributes['Meaning'].isin(['no transaction known',\n",
    "                                                                  'no transactions known'])]\n",
    "\n",
    "# Keep only attribute of azdias\n",
    "no_transaction = no_transaction[no_transaction['Attribute'].isin(azdias.columns)]\n",
    "\n",
    "# Get first part of attribute\n",
    "no_transaction['first_att'] = no_transaction.Attribute.apply(lambda x: x.split('_')[0])\n",
    "\n",
    "# unknown_Meaning shape\n",
    "no_transaction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    831734\n",
       "1     29771\n",
       "2     18067\n",
       "3      5708\n",
       "4      4082\n",
       "5      1483\n",
       "6       376\n",
       "Name: D19_BANKEN_ANZ_12, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Values of D19_BANKEN_ANZ_12  \n",
    "azdias.D19_BANKEN_ANZ_12.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**D19 are ordinals variables**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "no_transaction['type'] = 'ordinal'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": 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>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "      <th>first_att</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>D19_BANKEN_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>D19_BANKEN_OFFLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>D19_BANKEN_ONLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>D19_BANKEN_ONLINE_QUOTE_12</td>\n",
       "      <td>amount of online transactions within all trans...</td>\n",
       "      <td>0</td>\n",
       "      <td>no Online-transactions within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319</th>\n",
       "      <td>D19_GESAMT_DATUM</td>\n",
       "      <td>actuality of the last transaction with the com...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>D19_GESAMT_OFFLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction with the com...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>339</th>\n",
       "      <td>D19_GESAMT_ONLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction with the com...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>349</th>\n",
       "      <td>D19_GESAMT_ONLINE_QUOTE_12</td>\n",
       "      <td>amount of online transactions within all trans...</td>\n",
       "      <td>0</td>\n",
       "      <td>no Online-transactions within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>384</th>\n",
       "      <td>D19_KONSUMTYP</td>\n",
       "      <td>consumption type</td>\n",
       "      <td>1</td>\n",
       "      <td>Universal</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>492</th>\n",
       "      <td>D19_TELKO_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>510</th>\n",
       "      <td>D19_TELKO_OFFLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>520</th>\n",
       "      <td>D19_TELKO_ONLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>560</th>\n",
       "      <td>D19_VERSAND_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>570</th>\n",
       "      <td>D19_VERSAND_OFFLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>580</th>\n",
       "      <td>D19_VERSAND_ONLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>1</td>\n",
       "      <td>highest activity within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>590</th>\n",
       "      <td>D19_VERSAND_ONLINE_QUOTE_12</td>\n",
       "      <td>amount of online transactions within all trans...</td>\n",
       "      <td>0</td>\n",
       "      <td>no Online-transactions within the last 12 months</td>\n",
       "      <td>D19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>816</th>\n",
       "      <td>KBA05_AUTOQUOT</td>\n",
       "      <td>share of cars per household</td>\n",
       "      <td>1</td>\n",
       "      <td>very low car quote</td>\n",
       "      <td>KBA05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Attribute  \\\n",
       "152             D19_BANKEN_DATUM   \n",
       "186     D19_BANKEN_OFFLINE_DATUM   \n",
       "196      D19_BANKEN_ONLINE_DATUM   \n",
       "206   D19_BANKEN_ONLINE_QUOTE_12   \n",
       "319             D19_GESAMT_DATUM   \n",
       "329     D19_GESAMT_OFFLINE_DATUM   \n",
       "339      D19_GESAMT_ONLINE_DATUM   \n",
       "349   D19_GESAMT_ONLINE_QUOTE_12   \n",
       "384                D19_KONSUMTYP   \n",
       "492              D19_TELKO_DATUM   \n",
       "510      D19_TELKO_OFFLINE_DATUM   \n",
       "520       D19_TELKO_ONLINE_DATUM   \n",
       "560            D19_VERSAND_DATUM   \n",
       "570    D19_VERSAND_OFFLINE_DATUM   \n",
       "580     D19_VERSAND_ONLINE_DATUM   \n",
       "590  D19_VERSAND_ONLINE_QUOTE_12   \n",
       "816               KBA05_AUTOQUOT   \n",
       "\n",
       "                                           Description Value  \\\n",
       "152  actuality of the last transaction for the segm...     1   \n",
       "186  actuality of the last transaction for the segm...     1   \n",
       "196  actuality of the last transaction for the segm...     1   \n",
       "206  amount of online transactions within all trans...     0   \n",
       "319  actuality of the last transaction with the com...     1   \n",
       "329  actuality of the last transaction with the com...     1   \n",
       "339  actuality of the last transaction with the com...     1   \n",
       "349  amount of online transactions within all trans...     0   \n",
       "384                                  consumption type      1   \n",
       "492  actuality of the last transaction for the segm...     1   \n",
       "510  actuality of the last transaction for the segm...     1   \n",
       "520  actuality of the last transaction for the segm...     1   \n",
       "560  actuality of the last transaction for the segm...     1   \n",
       "570  actuality of the last transaction for the segm...     1   \n",
       "580  actuality of the last transaction for the segm...     1   \n",
       "590  amount of online transactions within all trans...     0   \n",
       "816                        share of cars per household     1   \n",
       "\n",
       "                                              Meaning first_att  \n",
       "152        highest activity within the last 12 months       D19  \n",
       "186        highest activity within the last 12 months       D19  \n",
       "196        highest activity within the last 12 months       D19  \n",
       "206  no Online-transactions within the last 12 months       D19  \n",
       "319        highest activity within the last 12 months       D19  \n",
       "329        highest activity within the last 12 months       D19  \n",
       "339        highest activity within the last 12 months       D19  \n",
       "349  no Online-transactions within the last 12 months       D19  \n",
       "384                                         Universal       D19  \n",
       "492        highest activity within the last 12 months       D19  \n",
       "510        highest activity within the last 12 months       D19  \n",
       "520        highest activity within the last 12 months       D19  \n",
       "560        highest activity within the last 12 months       D19  \n",
       "570        highest activity within the last 12 months       D19  \n",
       "580        highest activity within the last 12 months       D19  \n",
       "590  no Online-transactions within the last 12 months       D19  \n",
       "816                                very low car quote     KBA05  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Investigate the rest of attributes \n",
    "sub_dias = dias_attributes[~dias_attributes['Meaning'].isin(['no transaction known',\n",
    "                                                             'no transactions known',\n",
    "                                                            'unknown'])]\n",
    "\n",
    "# Keep only attribute of azdias\n",
    "sub_dias = sub_dias[sub_dias['Attribute'].isin(azdias.columns)]\n",
    "\n",
    "# Get first part of attribute\n",
    "sub_dias['first_att'] = sub_dias.Attribute.apply(lambda x: x.split('_')[0])\n",
    "\n",
    "# unknown_Meaning shape\n",
    "sub_dias"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NaN     257113\n",
       " 9.0    254296\n",
       " 1.0    117912\n",
       " 4.0     78262\n",
       " 6.0     56562\n",
       " 3.0     53330\n",
       " 2.0     49324\n",
       " 5.0     24422\n",
       "Name: D19_KONSUMTYP, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Values of D19_KONSUMTYP\n",
    "azdias.D19_KONSUMTYP.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "**All attributes in sub_dias are ordinal except D19_KONSUMTYP which is categorical**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sub_dias['type'] = sub_dias.Meaning.apply(\n",
    "    lambda x: 'categorical' if x=='Universal' else 'ordinal')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>765</th>\n",
       "      <td>KBA05_ALTER1</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,  9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>771</th>\n",
       "      <td>KBA05_ALTER2</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,  9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>777</th>\n",
       "      <td>KBA05_ALTER3</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,  9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>783</th>\n",
       "      <td>KBA05_ALTER4</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,  9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>790</th>\n",
       "      <td>KBA05_ANHANG</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1,  9]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        attribute     type missing_or_unknown\n",
       "765  KBA05_ALTER1  ordinal           [-1,  9]\n",
       "771  KBA05_ALTER2  ordinal           [-1,  9]\n",
       "777  KBA05_ALTER3  ordinal           [-1,  9]\n",
       "783  KBA05_ALTER4  ordinal           [-1,  9]\n",
       "790  KBA05_ANHANG  ordinal           [-1,  9]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Make a dataframe of features proprieties\n",
    "def_cl = lambda x: [str(k) for k in str(x).split(',')]\n",
    "unknown_Meaning['missing_or_unknown'] = unknown_Meaning.Value.apply(def_cl)\n",
    "\n",
    "# Drop colums\n",
    "unknown_Meaning.drop(['Description', 'Value', 'Meaning', 'first_att'],\n",
    "                    axis=1, inplace=True)\n",
    "\n",
    "# Change columns name\n",
    "unknown_Meaning.columns = ['attribute', 'type', 'missing_or_unknown']\n",
    "\n",
    "unknown_Meaning.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>D19_BANKEN_ANZ_12</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>D19_BANKEN_ANZ_24</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305</th>\n",
       "      <td>D19_GESAMT_ANZ_12</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>D19_GESAMT_ANZ_24</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>478</th>\n",
       "      <td>D19_TELKO_ANZ_12</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             attribute     type missing_or_unknown\n",
       "138  D19_BANKEN_ANZ_12  ordinal                [0]\n",
       "145  D19_BANKEN_ANZ_24  ordinal                [0]\n",
       "305  D19_GESAMT_ANZ_12  ordinal                [0]\n",
       "312  D19_GESAMT_ANZ_24  ordinal                [0]\n",
       "478   D19_TELKO_ANZ_12  ordinal                [0]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# no_transaction\n",
    "no_transaction['missing_or_unknown'] = no_transaction.Value.apply(lambda\n",
    "                                                                 x: [str(x)])\n",
    "\n",
    "# Drop colums\n",
    "no_transaction.drop(['Description', 'Value', 'Meaning', 'first_att'],\n",
    "                    axis=1, inplace=True)\n",
    "\n",
    "# Change columns name\n",
    "no_transaction.columns = ['attribute', 'type', 'missing_or_unknown']\n",
    "\n",
    "no_transaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>D19_BANKEN_DATUM</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>D19_BANKEN_OFFLINE_DATUM</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>D19_BANKEN_ONLINE_DATUM</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>D19_BANKEN_ONLINE_QUOTE_12</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>319</th>\n",
       "      <td>D19_GESAMT_DATUM</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      attribute     type missing_or_unknown\n",
       "152            D19_BANKEN_DATUM  ordinal                 []\n",
       "186    D19_BANKEN_OFFLINE_DATUM  ordinal                 []\n",
       "196     D19_BANKEN_ONLINE_DATUM  ordinal                 []\n",
       "206  D19_BANKEN_ONLINE_QUOTE_12  ordinal                 []\n",
       "319            D19_GESAMT_DATUM  ordinal                 []"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "sub_dias['missing_or_unknown'] = sub_dias.Value.apply(lambda x: [])\n",
    "\n",
    "# Drop colums\n",
    "sub_dias.drop(['Description', 'Value', 'Meaning', 'first_att'],\n",
    "              axis=1, inplace=True)\n",
    "\n",
    "# Change columns name\n",
    "sub_dias.columns = ['attribute', 'type', 'missing_or_unknown']\n",
    "\n",
    "sub_dias.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1, 0, 9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ANREDE_KZ</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CJT_GESAMTTYP</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>FINANZ_MINIMALIST</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              attribute         type missing_or_unknown\n",
       "0              AGER_TYP  categorical            [-1, 0]\n",
       "1  ALTERSKATEGORIE_GROB      ordinal         [-1, 0, 9]\n",
       "2             ANREDE_KZ  categorical            [-1, 0]\n",
       "3         CJT_GESAMTTYP  categorical                [0]\n",
       "4     FINANZ_MINIMALIST      ordinal               [-1]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# feat_info\n",
    "feat_info['missing_or_unknown'] = feat_info['missing_or_unknown'].apply(lambda \n",
    "                                                                        x: x[1:-1].split(','))\n",
    "\n",
    "# Drop information_level \n",
    "feat_info.drop(['information_level'], axis=1, inplace=True)\n",
    "\n",
    "feat_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Concat feat info\n",
    "new_feat_info = pd.concat([feat_info, sub_dias, no_transaction,\n",
    "                           unknown_Meaning], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(276, 3)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of new_feat_info\n",
    "new_feat_info.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Columns without type\n",
    "col = azdias.columns[~azdias.columns.isin(new_feat_info.attribute)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": 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>D19_LETZTER_KAUF_BRANCHE</th>\n",
       "      <th>EINGEFUEGT_AM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1992-02-10 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1997-04-21 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>D19_SCHUHE</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  D19_LETZTER_KAUF_BRANCHE        EINGEFUEGT_AM\n",
       "0                      NaN                  NaN\n",
       "1                      NaN  1992-02-10 00:00:00\n",
       "2            D19_UNBEKANNT  1992-02-12 00:00:00\n",
       "3            D19_UNBEKANNT  1997-04-21 00:00:00\n",
       "4               D19_SCHUHE  1992-02-12 00:00:00"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check object features\n",
    "sub_object = azdias[col].select_dtypes(include='object')\n",
    "sub_object.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NaN                       257113\n",
       "D19_UNBEKANNT             195338\n",
       "D19_VERSICHERUNGEN         57734\n",
       "D19_SONSTIGE               44722\n",
       "D19_VOLLSORTIMENT          34812\n",
       "D19_SCHUHE                 32578\n",
       "D19_BUCH_CD                28754\n",
       "D19_VERSAND_REST           26034\n",
       "D19_DROGERIEARTIKEL        24072\n",
       "D19_BANKEN_DIREKT          23273\n",
       "D19_BEKLEIDUNG_REST        21796\n",
       "D19_HAUS_DEKO              20858\n",
       "D19_TELKO_MOBILE           14447\n",
       "D19_ENERGIE                12084\n",
       "D19_TELKO_REST             11472\n",
       "D19_BANKEN_GROSS           10550\n",
       "D19_BEKLEIDUNG_GEH         10272\n",
       "D19_KINDERARTIKEL           7301\n",
       "D19_FREIZEIT                7257\n",
       "D19_TECHNIK                 7002\n",
       "D19_LEBENSMITTEL            6458\n",
       "D19_BANKEN_REST             5247\n",
       "D19_RATGEBER                4931\n",
       "D19_NAHRUNGSERGAENZUNG      4061\n",
       "D19_DIGIT_SERV              3577\n",
       "D19_REISEN                  3122\n",
       "D19_TIERARTIKEL             2578\n",
       "D19_SAMMELARTIKEL           2443\n",
       "D19_HANDWERK                2227\n",
       "D19_WEIN_FEINKOST           2164\n",
       "D19_GARTEN                  1646\n",
       "D19_BANKEN_LOKAL            1442\n",
       "D19_BIO_OEKO                1232\n",
       "D19_BILDUNG                  980\n",
       "D19_LOTTO                    839\n",
       "D19_KOSMETIK                 805\n",
       "Name: D19_LETZTER_KAUF_BRANCHE, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sub_object.D19_LETZTER_KAUF_BRANCHE.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": 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>LNR</th>\n",
       "      <th>AKT_DAT_KL</th>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <th>ANZ_KINDER</th>\n",
       "      <th>ANZ_STATISTISCHE_HAUSHALTE</th>\n",
       "      <th>CJT_KATALOGNUTZER</th>\n",
       "      <th>...</th>\n",
       "      <th>STRUKTURTYP</th>\n",
       "      <th>UMFELD_ALT</th>\n",
       "      <th>UMFELD_JUNG</th>\n",
       "      <th>UNGLEICHENN_FLAG</th>\n",
       "      <th>VERDICHTUNGSRAUM</th>\n",
       "      <th>VHA</th>\n",
       "      <th>VHN</th>\n",
       "      <th>VK_DHT4A</th>\n",
       "      <th>VK_DISTANZ</th>\n",
       "      <th>VK_ZG11</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8.912210e+05</td>\n",
       "      <td>817722.000000</td>\n",
       "      <td>81058.000000</td>\n",
       "      <td>29499.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>1205.000000</td>\n",
       "      <td>628274.000000</td>\n",
       "      <td>817722.000000</td>\n",
       "      <td>798073.000000</td>\n",
       "      <td>886367.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>793947.000000</td>\n",
       "      <td>793435.000000</td>\n",
       "      <td>793435.000000</td>\n",
       "      <td>817722.000000</td>\n",
       "      <td>793947.00000</td>\n",
       "      <td>817722.00000</td>\n",
       "      <td>770025.000000</td>\n",
       "      <td>815304.000000</td>\n",
       "      <td>815304.000000</td>\n",
       "      <td>815304.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>6.372630e+05</td>\n",
       "      <td>4.421928</td>\n",
       "      <td>11.745392</td>\n",
       "      <td>13.402658</td>\n",
       "      <td>14.476013</td>\n",
       "      <td>15.089627</td>\n",
       "      <td>13.700717</td>\n",
       "      <td>0.154018</td>\n",
       "      <td>7.599356</td>\n",
       "      <td>3.335264</td>\n",
       "      <td>...</td>\n",
       "      <td>2.539212</td>\n",
       "      <td>3.223125</td>\n",
       "      <td>4.017526</td>\n",
       "      <td>0.090067</td>\n",
       "      <td>4.58576</td>\n",
       "      <td>0.43882</td>\n",
       "      <td>2.417322</td>\n",
       "      <td>6.001214</td>\n",
       "      <td>7.532130</td>\n",
       "      <td>5.945972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.572735e+05</td>\n",
       "      <td>3.638805</td>\n",
       "      <td>4.097660</td>\n",
       "      <td>3.243300</td>\n",
       "      <td>2.712427</td>\n",
       "      <td>2.452932</td>\n",
       "      <td>5.079849</td>\n",
       "      <td>0.502389</td>\n",
       "      <td>14.332201</td>\n",
       "      <td>1.493633</td>\n",
       "      <td>...</td>\n",
       "      <td>0.754926</td>\n",
       "      <td>1.255785</td>\n",
       "      <td>1.118399</td>\n",
       "      <td>0.286278</td>\n",
       "      <td>8.47152</td>\n",
       "      <td>1.14329</td>\n",
       "      <td>1.166572</td>\n",
       "      <td>2.856091</td>\n",
       "      <td>3.247789</td>\n",
       "      <td>2.771464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.916530e+05</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>4.144580e+05</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6.372630e+05</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.600680e+05</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>17.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>8.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.082873e+06</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>449.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>45.00000</td>\n",
       "      <td>5.00000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>13.000000</td>\n",
       "      <td>11.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 88 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                LNR     AKT_DAT_KL   ALTER_KIND1   ALTER_KIND2  ALTER_KIND3  \\\n",
       "count  8.912210e+05  817722.000000  81058.000000  29499.000000  6170.000000   \n",
       "mean   6.372630e+05       4.421928     11.745392     13.402658    14.476013   \n",
       "std    2.572735e+05       3.638805      4.097660      3.243300     2.712427   \n",
       "min    1.916530e+05       1.000000      2.000000      2.000000     4.000000   \n",
       "25%    4.144580e+05       1.000000      8.000000     11.000000    13.000000   \n",
       "50%    6.372630e+05       3.000000     12.000000     14.000000    15.000000   \n",
       "75%    8.600680e+05       9.000000     15.000000     16.000000    17.000000   \n",
       "max    1.082873e+06       9.000000     18.000000     18.000000    18.000000   \n",
       "\n",
       "       ALTER_KIND4  ALTERSKATEGORIE_FEIN     ANZ_KINDER  \\\n",
       "count  1205.000000         628274.000000  817722.000000   \n",
       "mean     15.089627             13.700717       0.154018   \n",
       "std       2.452932              5.079849       0.502389   \n",
       "min       7.000000              0.000000       0.000000   \n",
       "25%      14.000000             11.000000       0.000000   \n",
       "50%      15.000000             14.000000       0.000000   \n",
       "75%      17.000000             17.000000       0.000000   \n",
       "max      18.000000             25.000000      11.000000   \n",
       "\n",
       "       ANZ_STATISTISCHE_HAUSHALTE  CJT_KATALOGNUTZER  ...    STRUKTURTYP  \\\n",
       "count               798073.000000      886367.000000  ...  793947.000000   \n",
       "mean                     7.599356           3.335264  ...       2.539212   \n",
       "std                     14.332201           1.493633  ...       0.754926   \n",
       "min                      0.000000           1.000000  ...       1.000000   \n",
       "25%                      1.000000           2.000000  ...       2.000000   \n",
       "50%                      3.000000           4.000000  ...       3.000000   \n",
       "75%                      9.000000           5.000000  ...       3.000000   \n",
       "max                    449.000000           5.000000  ...       3.000000   \n",
       "\n",
       "          UMFELD_ALT    UMFELD_JUNG  UNGLEICHENN_FLAG  VERDICHTUNGSRAUM  \\\n",
       "count  793435.000000  793435.000000     817722.000000      793947.00000   \n",
       "mean        3.223125       4.017526          0.090067           4.58576   \n",
       "std         1.255785       1.118399          0.286278           8.47152   \n",
       "min         1.000000       1.000000          0.000000           0.00000   \n",
       "25%         2.000000       3.000000          0.000000           0.00000   \n",
       "50%         3.000000       4.000000          0.000000           1.00000   \n",
       "75%         4.000000       5.000000          0.000000           5.00000   \n",
       "max         5.000000       5.000000          1.000000          45.00000   \n",
       "\n",
       "                VHA            VHN       VK_DHT4A     VK_DISTANZ  \\\n",
       "count  817722.00000  770025.000000  815304.000000  815304.000000   \n",
       "mean        0.43882       2.417322       6.001214       7.532130   \n",
       "std         1.14329       1.166572       2.856091       3.247789   \n",
       "min         0.00000       0.000000       1.000000       1.000000   \n",
       "25%         0.00000       2.000000       3.000000       5.000000   \n",
       "50%         0.00000       2.000000       6.000000       8.000000   \n",
       "75%         0.00000       3.000000       9.000000      10.000000   \n",
       "max         5.00000       4.000000      11.000000      13.000000   \n",
       "\n",
       "             VK_ZG11  \n",
       "count  815304.000000  \n",
       "mean        5.945972  \n",
       "std         2.771464  \n",
       "min         1.000000  \n",
       "25%         4.000000  \n",
       "50%         6.000000  \n",
       "75%         8.000000  \n",
       "max        11.000000  \n",
       "\n",
       "[8 rows x 88 columns]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "describe = azdias[col].describe()\n",
    "describe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": 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>Information level</th>\n",
       "      <th>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Additional notes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>NaN</td>\n",
       "      <td>D19_VERSI_OFFLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>NaN</td>\n",
       "      <td>D19_VERSI_ONLINE_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>NaN</td>\n",
       "      <td>D19_VERSI_DATUM</td>\n",
       "      <td>actuality of the last transaction for the segm...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Information level                Attribute  \\\n",
       "67               NaN  D19_VERSI_OFFLINE_DATUM   \n",
       "68               NaN   D19_VERSI_ONLINE_DATUM   \n",
       "69               NaN          D19_VERSI_DATUM   \n",
       "\n",
       "                                          Description Additional notes  \n",
       "67  actuality of the last transaction for the segm...              NaN  \n",
       "68  actuality of the last transaction for the segm...              NaN  \n",
       "69  actuality of the last transaction for the segm...              NaN  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Data information\n",
    "dias_info = pd.read_excel('Arvato_data/DIAS Information Levels - Attributes 2017.xlsx',\n",
    "                          skiprows=[0])\n",
    "del dias_info['Unnamed: 0']\n",
    "dias_info = dias_info[dias_info.Attribute.isin(col)]\n",
    "\n",
    "dias_info"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# List of columns to drop\n",
    "\n",
    "col_to_drop = list(sub_object.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['D19_LETZTER_KAUF_BRANCHE', 'EINGEFUEGT_AM']"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "col_to_drop"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discussion\n",
    "\n",
    "**D19_LETZTER_KAUF_BRANCHE  contains some variables names, and EINGEFUEGT_AM includes timestamps. Therefore, I will drop them.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LNR</td>\n",
       "      <td>numeric</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AKT_DAT_KL</td>\n",
       "      <td>numeric</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ALTER_KIND1</td>\n",
       "      <td>numeric</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ALTER_KIND2</td>\n",
       "      <td>numeric</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ALTER_KIND3</td>\n",
       "      <td>numeric</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     attribute     type missing_or_unknown\n",
       "0          LNR  numeric                 []\n",
       "1   AKT_DAT_KL  numeric                 []\n",
       "2  ALTER_KIND1  numeric                 []\n",
       "3  ALTER_KIND2  numeric                 []\n",
       "4  ALTER_KIND3  numeric                 []"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "last_att_info = pd.DataFrame({'attribute': list(describe.columns)})\n",
    "last_att_info['type'] = 'numeric'\n",
    "last_att_info['missing_or_unknown'] = last_att_info['type'].apply(lambda\n",
    "                                                                 x: [])\n",
    "\n",
    "last_att_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(364, 3)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Concat last_att_info and new_feat_info\n",
    "new_feat_info = pd.concat([new_feat_info, last_att_info], ignore_index=True)\n",
    "\n",
    "new_feat_info.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.1.1: Convert Missing Value Codes to NaNs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/pandas/core/ops.py:1649: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  result = method(y)\n"
     ]
    }
   ],
   "source": [
    "# Identify missing or unknown data values and convert them to NaNs.\n",
    "for attrib, missing_values in zip(new_feat_info['attribute'],\n",
    "                                  new_feat_info['missing_or_unknown']):\n",
    "    if len(missing_values) != 0:\n",
    "        for value in missing_values:\n",
    "            if value.isnumeric() or value.lstrip('-').isnumeric():\n",
    "                value = int(value)\n",
    "            azdias.loc[azdias[attrib] == value, attrib] = np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(891221, 364)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Drop col_to_drop\n",
    "azdias.drop(col_to_drop, axis=1, inplace=True)\n",
    "\n",
    "# check shape\n",
    "azdias.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.1.2: Assess Missing Data in Each Column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Missing Data in Each Column\n",
    "count_nan = azdias.isnull().sum()/len(azdias)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiUAAAGHCAYAAABvUSKTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10aaef9cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "sns.distplot(count_nan.values, kde=False)\n",
    "plt.ylabel('count of columns')\n",
    "plt.xlabel('% of NaN')\n",
    "plt.title('Distribution of missing value per columns')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discusion\n",
    "\n",
    "**Most attributes contain less than 30% of missing values. Therefore I will drop all features that include more than 30% of missing values.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get columns that have more than 30% of nan\n",
    "more_30 = count_nan[count_nan > 0.3]\n",
    "missing_col = list(more_30.index)\n",
    "len(missing_col)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(891221, 343)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Remove the outlier missing_col from the dataset.\n",
    "azdias.drop(missing_col, axis=1, inplace=True)\n",
    "\n",
    "azdias.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.1.2: Assess Missing Data in Each row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Missing Data in Each Column\n",
    "row_nan = azdias.isnull().sum(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtoAAAHUCAYAAAAEDI/eAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4ELbzMzMzKwA/wNLCO5CgqCxvQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ad66d048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 5))\n",
    "sns.distplot(row_nan.values/343, kde=False)\n",
    "plt.ylabel('count of rows')\n",
    "plt.xlabel('% of NaN')\n",
    "plt.title('Distribution of missing value per row')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fig.savefig('row_nan.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Discussion\n",
    "\n",
    "**Rows that contain many missing values are harmful to machine learning algorithms. Therefore I will drop rows that contain more than 50% of missing values.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(791987, 343)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Drop row that contains more than 50% of nan values\n",
    "azdias = azdias.loc[row_nan[row_nan <= 0.5*343].index]\n",
    "\n",
    "azdias.reset_index(drop=True, inplace=True)\n",
    "\n",
    "azdias.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891221 entries, 0 to 891220\n",
      "Columns: 343 entries, LNR to ALTERSKATEGORIE_GROB\n",
      "dtypes: float64(288), int64(51), object(4)\n",
      "memory usage: 2.3+ GB\n"
     ]
    }
   ],
   "source": [
    "azdias.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1.2: Select and Re-Encode Features\n",
    "\n",
    "Machine learning techniques work only on data that is encoded numerically, while almost all of the values in the dataset are encoded using numbers, not all of them represent numeric values.  I will use the features description to re-encode features.\n",
    "- I will keep numeric, ordinal and interval data without changes.\n",
    "- Special handling is necessary for the remaining two variable types: categorical, and  mixed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1, 0, 9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ANREDE_KZ</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CJT_GESAMTTYP</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>FINANZ_MINIMALIST</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              attribute         type missing_or_unknown\n",
       "0              AGER_TYP  categorical            [-1, 0]\n",
       "1  ALTERSKATEGORIE_GROB      ordinal         [-1, 0, 9]\n",
       "2             ANREDE_KZ  categorical            [-1, 0]\n",
       "3         CJT_GESAMTTYP  categorical                [0]\n",
       "4     FINANZ_MINIMALIST      ordinal               [-1]"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_feat_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "343"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# keep only azdias attribute\n",
    "final_feat_info = new_feat_info[new_feat_info.attribute.isin(azdias.columns)]\n",
    "\n",
    "len(final_feat_info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ordinal        229\n",
       "numeric         89\n",
       "categorical     19\n",
       "mixed            6\n",
       "Name: type, dtype: int64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# How many features are there of each data type?\n",
    "final_feat_info.type.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "count_type = final_feat_info.type.value_counts(dropna=False)\n",
    "\n",
    "df_count = pd.DataFrame(count_type.values, index=count_type.index, columns=['count columns'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": 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>count columns</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ordinal</th>\n",
       "      <td>229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>numeric</th>\n",
       "      <td>89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>categorical</th>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mixed</th>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             count columns\n",
       "ordinal                229\n",
       "numeric                 89\n",
       "categorical             19\n",
       "mixed                    6"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_count"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 1.2.1: Re-Encode Categorical Features\n",
    "\n",
    "For categorical data, depending on the number of categories,  I will perform one of the following:\n",
    "-  For binary (two-level) categoricals that take numeric values, I will keep them without any change.\n",
    "-  For binary variables that take on non-numeric values, I will re-encod the values as numbers.\n",
    "- For multi-level categoricals variables (three or more values), I will re-encod the values using multiple dummies variables. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": 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>attribute</th>\n",
       "      <th>type</th>\n",
       "      <th>missing_or_unknown</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1, 0, 9]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ANREDE_KZ</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[-1, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CJT_GESAMTTYP</td>\n",
       "      <td>categorical</td>\n",
       "      <td>[0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>FINANZ_MINIMALIST</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>FINANZ_SPARER</td>\n",
       "      <td>ordinal</td>\n",
       "      <td>[-1]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              attribute         type missing_or_unknown\n",
       "1  ALTERSKATEGORIE_GROB      ordinal         [-1, 0, 9]\n",
       "2             ANREDE_KZ  categorical            [-1, 0]\n",
       "3         CJT_GESAMTTYP  categorical                [0]\n",
       "4     FINANZ_MINIMALIST      ordinal               [-1]\n",
       "5         FINANZ_SPARER      ordinal               [-1]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_feat_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Get list of Categorical Features\n",
    "Cat_feat = list(final_feat_info.attribute[\n",
    "        final_feat_info.type == 'categorical'])\n",
    "\n",
    "binary_num_attribute = []\n",
    "binary_str_attribute = []\n",
    "multi_level_attribute = []\n",
    "\n",
    "for att in Cat_feat:\n",
    "    dtype = azdias[att].dtype\n",
    "    count = len(azdias[att].value_counts())\n",
    "\n",
    "    if count > 2:\n",
    "        multi_level_attribute.append(att)\n",
    "    else:\n",
    "        if dtype == 'object':\n",
    "            binary_str_attribute.append(att)\n",
    "        else:\n",
    "            binary_num_attribute.append(att)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ANREDE_KZ', 'GREEN_AVANTGARDE', 'SOHO_KZ', 'VERS_TYP']"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# binary_num_attribute\n",
    "binary_num_attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['OST_WEST_KZ']"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# binary_str_attribute\n",
    "binary_str_attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['CJT_GESAMTTYP',\n",
       " 'FINANZTYP',\n",
       " 'GFK_URLAUBERTYP',\n",
       " 'LP_FAMILIE_FEIN',\n",
       " 'LP_FAMILIE_GROB',\n",
       " 'LP_STATUS_FEIN',\n",
       " 'LP_STATUS_GROB',\n",
       " 'NATIONALITAET_KZ',\n",
       " 'SHOPPER_TYP',\n",
       " 'ZABEOTYP',\n",
       " 'GEBAEUDETYP',\n",
       " 'CAMEO_DEUG_2015',\n",
       " 'CAMEO_DEU_2015',\n",
       " 'D19_KONSUMTYP']"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# multi_level_attribute\n",
    "multi_level_attribute"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Re-encode OST_WEST_KZ as dummy variable\n",
    "azdias['OST_WEST_KZ'] = azdias['OST_WEST_KZ'].map({'W': 1, 'O': 2})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Convert cat to cat to dummies variables\n",
    "\n",
    "\n",
    "def encode_cat(df, list_cat):\n",
    "    list_df_dum = []\n",
    "    for col_dum in list_cat:\n",
    "        list_df_dum.append(pd.get_dummies(df[col_dum], prefix=col_dum))\n",
    "\n",
    "    assert len(list_df_dum) == len(list_cat)\n",
    "    \n",
    "    # Drop cat col from df\n",
    "    df.drop(list_cat, axis=1, inplace=True)\n",
    "    list_df_dum.append(df)\n",
    "    df = pd.concat(list_df_dum, axis=1)\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "new_azdias = encode_cat(azdias, multi_level_attribute)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Step 1.2.2: Engineer Mixed-Type Features\n",
    "\n",
    "There are  six mixed features. I will use the feature summary given in the previous project to encode them as ordinal and dummies variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['LP_LEBENSPHASE_FEIN',\n",
       " 'LP_LEBENSPHASE_GROB',\n",
       " 'PRAEGENDE_JUGENDJAHRE',\n",
       " 'WOHNLAGE',\n",
       " 'CAMEO_INTL_2015',\n",
       " 'PLZ8_BAUMAX']"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mixed_att = list(final_feat_info.attribute[final_feat_info.type == 'mixed'])\n",
    "mixed_att"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**\"PRAEGENDE_JUGENDJAHRE\" Combines information on three dimensions: generation by decade, movement (mainstream vs. avantgarde), and nation (east vs. west). While there aren't enough levels to disentangle east from west, I will create two new variables to capture the other two dimensions: an interval-type variable for decade, and a binary variable for movement.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Map decade\n",
    "dic_dec = {1: 40, 2: 40, 3: 50, 4: 50, 5: 60, 6: 60, 7: 60,\n",
    "           8: 70, 9: 70, 10: 80, 11: 80, 12: 80, 13: 80, 14: 90,\n",
    "           15: 90, np.nan: np.nan, -1: np.nan, 0: np.nan}\n",
    "\n",
    "# Map mouvement\n",
    "dic_mov = {1: 0, 2: 1, 3: 0, 4: 1, 5: 0, 6: 1, 7: 1, 8: 0,\n",
    "           9: 1, 10: 0, 11: 1, 12: 0, 13: 1, 14: 0, 15: 1,\n",
    "           np.nan: np.nan, -1: np.nan, 0: np.nan}\n",
    "\n",
    "new_azdias[\"PRAEGENDE_JUGENDJAHRE_dec\"] = new_azdias[\"PRAEGENDE_JUGENDJAHRE\"].map(dic_dec)\n",
    "\n",
    "# Mainstream : 0, Avantgarde : 1\n",
    "new_azdias[\"PRAEGENDE_JUGENDJAHRE_mov\"] = new_azdias[\"PRAEGENDE_JUGENDJAHRE\"].map(dic_mov)\n",
    "\n",
    "# Drop PRAEGENDE_JUGENDJAHRE\n",
    "new_azdias.drop('PRAEGENDE_JUGENDJAHRE', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**\"CAMEO_INTL_2015\" combines information on two axes: wealth and life stage. I will break up the two-digit codes by their 'tens'-place and 'ones'-place digits into two new ordinal variables.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def encode_CAMEO_INTL_2015(x):\n",
    "\n",
    "    if np.isnan(x):\n",
    "        return [np.nan, np.nan]\n",
    "    else:\n",
    "        x = str(x).split('.')[0]\n",
    "        return [int(x[0]), int(x[1])]\n",
    "\n",
    "\n",
    "new_azdias[\"CAMEO_INTL_2015\"] = new_azdias[\"CAMEO_INTL_2015\"].astype('float')\n",
    "new_azdias[\"CAMEO_INTL_2015\"] = new_azdias[\"CAMEO_INTL_2015\"].apply(encode_CAMEO_INTL_2015)\n",
    "\n",
    "# CAMEO_INTL_2015_wealth\n",
    "new_azdias[\"CAMEO_INTL_2015_wealth\"] = new_azdias[\"CAMEO_INTL_2015\"].apply(lambda x: x[0])\n",
    "\n",
    "# CAMEO_INTL_2015_life_stage\n",
    "new_azdias[\"CAMEO_INTL_2015_life_stage\"] = new_azdias[\"CAMEO_INTL_2015\"].apply(lambda x: x[1])\n",
    "\n",
    "# drop CAMEO_INTL_2015\n",
    "new_azdias.drop('CAMEO_INTL_2015', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**\"LP_LEBENSPHASE_FEIN\" contains information about life stage and fine scale. I will create two new ordinals features, one for life_stage and other for fine_scale**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "life_stage = {1: 'younger_age', 2: 'middle_age', 3: 'younger_age',\n",
    "              4: 'middle_age', 5: 'advanced_age', 6: 'retirement_age',\n",
    "              7: 'advanced_age', 8: 'retirement_age', 9: 'middle_age',\n",
    "              10: 'middle_age', 11: 'advanced_age', 12: 'retirement_age',\n",
    "              13: 'advanced_age', 14: 'younger_age', 15: 'advanced_age',\n",
    "              16: 'advanced_age', 17: 'middle_age', 18: 'younger_age',\n",
    "              19: 'advanced_age', 20: 'advanced_age', 21: 'middle_age',\n",
    "              22: 'middle_age', 23: 'middle_age', 24: 'middle_age',\n",
    "              25: 'middle_age', 26: 'middle_age', 27: 'middle_age',\n",
    "              28: 'middle_age', 29: 'younger_age', 30: 'younger_age',\n",
    "              31: 'advanced_age', 32: 'advanced_age', 33: 'younger_age',\n",
    "              34: 'younger_age', 35: 'younger_age', 36: 'advanced_age',\n",
    "              37: 'advanced_age', 38: 'retirement_age', 39: 'middle_age',\n",
    "              40: 'retirement_age'}\n",
    "\n",
    "fine_scale = {1: 'low', 2: 'low', 3: 'average', 4: 'average', 5: 'low', 6: 'low',\n",
    "              7: 'average', 8: 'average', 9: 'average', 10: 'wealthy', 11: 'average',\n",
    "              12: 'average', 13: 'top', 14: 'average', 15: 'low', 16: 'average',\n",
    "              17: 'average', 18: 'wealthy', 19: 'wealthy', 20: 'top', 21: 'low',\n",
    "              22: 'average', 23: 'wealthy', 24: 'low', 25: 'average', 26: 'average',\n",
    "              27: 'average', 28: 'top', 29: 'low', 30: 'average', 31: 'low',\n",
    "              32: 'average', 33: 'average', 34: 'average', 35: 'top', 36: 'average',\n",
    "              37: 'average', 38: 'average', 39: 'top', 40: 'top'}\n",
    "\n",
    "new_azdias['LP_life_stage'] = new_azdias['LP_LEBENSPHASE_FEIN'].map(life_stage)\n",
    "new_azdias['LP_fine_scale'] = new_azdias['LP_LEBENSPHASE_FEIN'].map(fine_scale)\n",
    "\n",
    "# Drop LP_LEBENSPHASE_FEIN\n",
    "new_azdias.drop('LP_LEBENSPHASE_FEIN', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# R-encode LP_life_stage and LP_fine_scale as ordinal\n",
    "dic_life = {'younger_age': 1, 'middle_age': 2, 'advanced_age': 3,\n",
    "            'retirement_age': 4}\n",
    "dic_scale = {'low': 1, 'average': 2, 'wealthy': 3, 'top': 4}\n",
    "\n",
    "new_azdias['LP_life_stage'] = new_azdias['LP_life_stage'].map(dic_life)\n",
    "new_azdias['LP_fine_scale'] = new_azdias['LP_fine_scale'].map(dic_scale)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " \n",
    "**LP_LEBENSPHASE_GROB contains the same information as LP_LEBENSPHASE_FEIN. Therefore I will drop it.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "new_azdias.drop('LP_LEBENSPHASE_GROB', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I will Re-encode WOHNLAGE  as dummies variables**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# replace 0 by nan\n",
    "new_azdias['WOHNLAGE'] = new_azdias['WOHNLAGE'].replace(0, np.nan)\n",
    "\n",
    "# Get dummumies variable\n",
    "WOHNLAGE = pd.get_dummies(new_azdias['WOHNLAGE'], prefix='WOHNLAGE')\n",
    "\n",
    "# Drop WOHNLAGE\n",
    "new_azdias.drop('WOHNLAGE', axis=1, inplace=True)\n",
    "\n",
    "# concat WOHNLAGE with new_azdias\n",
    "new_azdias = pd.concat([new_azdias, WOHNLAGE], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**I will Re-encode PLZ8_BAUMAX  as dummies variables**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Get dummumies variable\n",
    "PLZ8_BAUMAX = pd.get_dummies(new_azdias['PLZ8_BAUMAX'], prefix='PLZ8_BAUMAX')\n",
    "\n",
    "# Drop PLZ8_BAUMAX\n",
    "new_azdias.drop('PLZ8_BAUMAX', axis=1, inplace=True)\n",
    "\n",
    "# concat WOHNLAGE with new_azdias\n",
    "new_azdias = pd.concat([new_azdias, PLZ8_BAUMAX], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 1.3: Create a Cleaning Function\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def clean_data(df, drop_col=True, drop_row=True):\n",
    "    \"\"\"\n",
    "    Perform feature trimming, re-encoding, and engineering for demographics\n",
    "    data\n",
    "\n",
    "    INPUT: Demographics DataFrame\n",
    "    OUTPUT: Trimmed and cleaned demographics DataFrame\n",
    "    \"\"\"\n",
    "\n",
    "    # Drop 'D19_LETZTER_KAUF_BRANCHE' and 'EINGEFUEGT_AM'\n",
    "    df.drop(col_to_drop, axis=1, inplace=True)\n",
    "\n",
    "    # convert missing value codes into NaNs\n",
    "    for attrib, missing_values in zip(new_feat_info['attribute'],\n",
    "                                      new_feat_info['missing_or_unknown']):\n",
    "        if len(missing_values) != 0:\n",
    "            for value in missing_values:\n",
    "                if value.isnumeric() or value.lstrip('-').isnumeric():\n",
    "                    value = int(value)\n",
    "                df.loc[df[attrib] == value, attrib] = np.nan\n",
    "\n",
    "    # remove selected columns\n",
    "    if drop_col:\n",
    "        df.drop(missing_col, axis=1, inplace=True)\n",
    "    \n",
    "    # assessing missing value per row\n",
    "    row_nan = df.isnull().sum(axis=1)\n",
    "    a = df.shape[1]\n",
    "    \n",
    "    if drop_row:\n",
    "        df = df.loc[row_nan[row_nan<=0.5*a].index]\n",
    "        df.reset_index(drop=True, inplace=True)\n",
    "\n",
    "    # Re-encode OST_WEST_KZ as dummy variable\n",
    "    df['OST_WEST_KZ'] =  df['OST_WEST_KZ'].map({'W': 1, 'O': 2})\n",
    "    \n",
    "    # Convert cat to cat to dummies variables\n",
    "    df = encode_cat(df, multi_level_attribute)\n",
    "    \n",
    "    # Mix\n",
    "    df[\"PRAEGENDE_JUGENDJAHRE_dec\"] = df[\"PRAEGENDE_JUGENDJAHRE\"].map(dic_dec)\n",
    "\n",
    "    # Mainstream : 0, Avantgarde : 1\n",
    "    df[\"PRAEGENDE_JUGENDJAHRE_mov\"] = df[\"PRAEGENDE_JUGENDJAHRE\"].map(dic_mov)\n",
    "\n",
    "    # Drop PRAEGENDE_JUGENDJAHRE\n",
    "    df.drop('PRAEGENDE_JUGENDJAHRE', axis=1, inplace=True)\n",
    "    \n",
    "    # re\n",
    "    df[\"CAMEO_INTL_2015\"] = df[\"CAMEO_INTL_2015\"].astype('float')\n",
    "    df[\"CAMEO_INTL_2015\"] = df[\"CAMEO_INTL_2015\"].apply(encode_CAMEO_INTL_2015)\n",
    "\n",
    "    # CAMEO_INTL_2015_wealth\n",
    "    df[\"CAMEO_INTL_2015_wealth\"] = df[\"CAMEO_INTL_2015\"].apply(lambda x: x[0])\n",
    "\n",
    "    # CAMEO_INTL_2015_life_stage\n",
    "    df[\"CAMEO_INTL_2015_life_stage\"] = df[\"CAMEO_INTL_2015\"].apply(lambda x: x[1])\n",
    "\n",
    "    # drop CAMEO_INTL_2015\n",
    "    df.drop('CAMEO_INTL_2015', axis=1, inplace=True)\n",
    "    \n",
    "    #\n",
    "    df['LP_life_stage'] = df['LP_LEBENSPHASE_FEIN'].map(life_stage)\n",
    "    df['LP_fine_scale'] = df['LP_LEBENSPHASE_FEIN'].map(fine_scale)\n",
    "\n",
    "    # Drop LP_LEBENSPHASE_FEIN \n",
    "    df.drop('LP_LEBENSPHASE_FEIN', axis=1, inplace=True)\n",
    "    \n",
    "    df['LP_life_stage'] = df['LP_life_stage'].map(dic_life)\n",
    "    df['LP_fine_scale'] = df['LP_fine_scale'].map(dic_scale)\n",
    "    \n",
    "    # Drop LP_LEBENSPHASE_GROB\n",
    "    df.drop('LP_LEBENSPHASE_GROB', axis=1, inplace=True)\n",
    "    \n",
    "    # replace 0 by nan\n",
    "    df['WOHNLAGE'] = df['WOHNLAGE'].replace(0, np.nan)\n",
    "\n",
    "    # Get dummumies variable\n",
    "    WOHNLAGE = pd.get_dummies(df['WOHNLAGE'], prefix='WOHNLAGE')\n",
    "\n",
    "    # Drop WOHNLAGE\n",
    "    df.drop('WOHNLAGE', axis=1, inplace=True)\n",
    "\n",
    "    # concat WOHNLAGE with new_azdias\n",
    "    df = pd.concat([df, WOHNLAGE], axis=1)\n",
    "    \n",
    "    # Get dummumies variable\n",
    "    PLZ8_BAUMAX = pd.get_dummies(df['PLZ8_BAUMAX'], prefix='PLZ8_BAUMAX')\n",
    "\n",
    "    # Drop PLZ8_BAUMAX\n",
    "    df.drop('PLZ8_BAUMAX', axis=1, inplace=True)\n",
    "\n",
    "    # concat WOHNLAGE with new_azdias\n",
    "    df = pd.concat([df, PLZ8_BAUMAX], axis=1)\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Step 2: Feature Transformation\n",
    "\n",
    "### Step 2.1: Apply Feature Scaling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**LNR is not useful for machine learning because it contains a unique value for each row. Therefore, I will drop it.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# drop LNR from data to process\n",
    "LNR_azdias = new_azdias['LNR']\n",
    "new_azdias.drop('LNR', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/sklearn/utils/deprecation.py:58: DeprecationWarning: Class Imputer is deprecated; Imputer was deprecated in version 0.20 and will be removed in 0.22. Import impute.SimpleImputer from sklearn instead.\n",
      "  warnings.warn(msg, category=DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "# Imputer\n",
    "Imp = Imputer(missing_values=np.nan, strategy='median')\n",
    "new_azdias = pd.DataFrame(Imp.fit_transform(new_azdias.values),\n",
    "                          columns=new_azdias.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhsAAAFkCAYAAACJu/k0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ab1615c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check features range\n",
    "#describe_azdias = new_azdias.describe()\n",
    "\n",
    "plt.figure()\n",
    "plt.plot(describe_azdias.loc['min'].values, label='min')\n",
    "plt.plot(describe_azdias.loc['max'].values, label='max')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "** Somme features have a big range. I will scale the data to avoid  that some feature imbalance  weight**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Apply feature scaling to the general population demographics data.\n",
    "scaler = RobustScaler()\n",
    "scaled_azdias = pd.DataFrame(scaler.fit_transform(new_azdias.values),\n",
    "                          columns=new_azdias.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 2.2: Perform Dimensionality Reduction\n",
    "\n",
    "I will use sklearn's [PCA](http://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html) class to apply principal component analysis on the data, thus finding the vectors of maximal variance in the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Apply PCA to the azdias data.\n",
    "pca = PCA()\n",
    "azdias_pca = pca.fit_transform(scaled_azdias)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def scree_plot(pca):\n",
    "    '''\n",
    "    Creates a scree plot associated with the principal components\n",
    "    \n",
    "    INPUT: pca - the result of instantian of PCA in scikit learn\n",
    "            \n",
    "    OUTPUT:\n",
    "            None\n",
    "    '''\n",
    "    num_components = len(pca.explained_variance_ratio_)\n",
    "    ind = np.arange(num_components)\n",
    "    vals = pca.explained_variance_ratio_\n",
    " \n",
    "    plt.figure(figsize=(16, 10))\n",
    "    ax = plt.subplot(111)\n",
    "    cumvals = np.cumsum(vals)\n",
    "    ax.bar(ind, vals)\n",
    "    ax.plot(ind, cumvals)\n",
    "    for i in range(num_components):\n",
    "        ax.annotate(r\"%s%%\" % ((str(vals[i]*100)[:4])), (ind[i]+0.2, vals[i]),\n",
    "                    va=\"bottom\", ha=\"center\", fontsize=12)\n",
    " \n",
    "    ax.xaxis.set_tick_params(width=0)\n",
    "    ax.yaxis.set_tick_params(width=2, length=12)\n",
    " \n",
    "    ax.set_xlabel(\"Principal Component\")\n",
    "    ax.set_ylabel(\"Variance Explained (%)\")\n",
    "    plt.title('Explained Variance Per Principal Component')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABSkAAANYCAYAAAAooWhNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf5JTzPmC8/MTAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10aa66fa20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Investigate the variance accounted for by each principal component.\n",
    "scree_plot(pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Re-apply PCA to the data whith 200 components.\n",
    "pca = PCA(n_components=200)\n",
    "azdias_pca = pca.fit_transform(scaled_azdias)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discussion\n",
    "\n",
    "**Looking at the plot above, we may notice that after 200 components the cumulative explained variance becomes superior to 0.95. So I will choose 200 features to re-do the PCA.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 3: Clustering\n",
    "\n",
    "### Step 3.1: Apply Clustering to General Population\n",
    "\n",
    "**I will use sklearn's [KMeans](http://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html#sklearn.cluster.KMeans) class to perform k-means clustering on the PCA-transformed data.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Compute the average within-cluster distances.\n",
    "\n",
    "\n",
    "def get_kmeans_score(data, center):\n",
    "    '''\n",
    "    returns the kmeans score regarding SSE for points to centers\n",
    "    INPUT:\n",
    "        data - the dataset you want to fit kmeans to\n",
    "        center - the number of centers you want (the k value)\n",
    "    OUTPUT:\n",
    "        score - the SSE score for the kmeans model fit to the data\n",
    "    '''\n",
    "    # instantiate kmeans\n",
    "    kmeans = KMeans(n_clusters=center, random_state=42, n_jobs=-1)\n",
    "\n",
    "    # Then fit the model to your data using the fit method\n",
    "    model = kmeans.fit(data)\n",
    "    \n",
    "    # Obtain a score related to the model fit\n",
    "    score = np.abs(model.score(data))\n",
    "    \n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Apply the elbow method\n",
    "score_list = []\n",
    "centers = np.arange(2, 15, 1)\n",
    "\n",
    "\n",
    "for center in centers: \n",
    "    # Append score to \n",
    "    score_list.append(get_kmeans_score(azdias_pca, center))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6AAAAL3CAYAAACUOShKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCaiZmZmZmZlV4v8Bq3kxZpdwKZEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ab973908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Investigate the change in within-cluster distance across number of clusters.\n",
    "\n",
    "plt.figure(figsize=(10, 8)) \n",
    "plt.plot(centers, score_list, linestyle='--', marker='o', color='b');\n",
    "matplotlib.rc('xtick', labelsize=30)  \n",
    "matplotlib.rc('ytick', labelsize=30)\n",
    "plt.xlabel('K', fontsize=20)\n",
    "plt.ylabel('SSE', fontsize=20)\n",
    "plt.title('SSE vs. K', fontsize=20)\n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Discussion\n",
    "\n",
    "**Looking at the plot above, we can notice that the score keeps decreasing and there is no visible elbow, but after 6 clusters the decrease is weaker. I will choose six clusters to do KMeans clustering.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Re-fit the k-means model with 6 class\n",
    "kmeans = KMeans(n_clusters=6, random_state=42, n_jobs=-1)\n",
    "labels_azdias = kmeans.fit_predict(azdias_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "collapsed": 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>% of data</th>\n",
       "      <th>cluster</th>\n",
       "      <th>data</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>38.772875</td>\n",
       "      <td>4</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>21.748814</td>\n",
       "      <td>1</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>20.541818</td>\n",
       "      <td>2</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12.935175</td>\n",
       "      <td>0</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.342245</td>\n",
       "      <td>5</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.659072</td>\n",
       "      <td>3</td>\n",
       "      <td>general population</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   % of data  cluster                data\n",
       "0  38.772875        4  general population\n",
       "1  21.748814        1  general population\n",
       "2  20.541818        2  general population\n",
       "3  12.935175        0  general population\n",
       "4   4.342245        5  general population\n",
       "5   1.659072        3  general population"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "class_count = pd.Series(labels_azdias).value_counts()\n",
    "\n",
    "cls_azdias = pd.DataFrame({'cluster': class_count.index,\n",
    "                           '% of data': 100*class_count.values/len(labels_azdias)})\n",
    "cls_azdias['data'] = 'general population'\n",
    "\n",
    "# print cls_azdias\n",
    "cls_azdias"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Step 3.2: Apply All Steps to the Customer Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (18,19) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\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>LNR</th>\n",
       "      <th>AGER_TYP</th>\n",
       "      <th>AKT_DAT_KL</th>\n",
       "      <th>ALTER_HH</th>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <th>ANZ_HAUSHALTE_AKTIV</th>\n",
       "      <th>...</th>\n",
       "      <th>VK_ZG11</th>\n",
       "      <th>W_KEIT_KIND_HH</th>\n",
       "      <th>WOHNDAUER_2008</th>\n",
       "      <th>WOHNLAGE</th>\n",
       "      <th>ZABEOTYP</th>\n",
       "      <th>PRODUCT_GROUP</th>\n",
       "      <th>CUSTOMER_GROUP</th>\n",
       "      <th>ONLINE_PURCHASE</th>\n",
       "      <th>ANREDE_KZ</th>\n",
       "      <th>ALTERSKATEGORIE_GROB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9626</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3</td>\n",
       "      <td>COSMETIC_AND_FOOD</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9628</td>\n",
       "      <td>-1</td>\n",
       "      <td>9.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>FOOD</td>\n",
       "      <td>SINGLE_BUYER</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>143872</td>\n",
       "      <td>-1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>11.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>COSMETIC_AND_FOOD</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>143873</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>COSMETIC</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>143874</td>\n",
       "      <td>-1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>FOOD</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 369 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      LNR  AGER_TYP  AKT_DAT_KL  ALTER_HH  ALTER_KIND1  ALTER_KIND2  \\\n",
       "0    9626         2         1.0      10.0          NaN          NaN   \n",
       "1    9628        -1         9.0      11.0          NaN          NaN   \n",
       "2  143872        -1         1.0       6.0          NaN          NaN   \n",
       "3  143873         1         1.0       8.0          NaN          NaN   \n",
       "4  143874        -1         1.0      20.0          NaN          NaN   \n",
       "\n",
       "   ALTER_KIND3  ALTER_KIND4  ALTERSKATEGORIE_FEIN  ANZ_HAUSHALTE_AKTIV  ...  \\\n",
       "0          NaN          NaN                  10.0                  1.0  ...   \n",
       "1          NaN          NaN                   NaN                  NaN  ...   \n",
       "2          NaN          NaN                   0.0                  1.0  ...   \n",
       "3          NaN          NaN                   8.0                  0.0  ...   \n",
       "4          NaN          NaN                  14.0                  7.0  ...   \n",
       "\n",
       "   VK_ZG11  W_KEIT_KIND_HH  WOHNDAUER_2008  WOHNLAGE  ZABEOTYP  \\\n",
       "0      2.0             6.0             9.0       7.0         3   \n",
       "1      3.0             0.0             9.0       NaN         3   \n",
       "2     11.0             6.0             9.0       2.0         3   \n",
       "3      2.0             NaN             9.0       7.0         1   \n",
       "4      4.0             2.0             9.0       3.0         1   \n",
       "\n",
       "       PRODUCT_GROUP  CUSTOMER_GROUP ONLINE_PURCHASE ANREDE_KZ  \\\n",
       "0  COSMETIC_AND_FOOD     MULTI_BUYER               0         1   \n",
       "1               FOOD    SINGLE_BUYER               0         1   \n",
       "2  COSMETIC_AND_FOOD     MULTI_BUYER               0         2   \n",
       "3           COSMETIC     MULTI_BUYER               0         1   \n",
       "4               FOOD     MULTI_BUYER               0         1   \n",
       "\n",
       "  ALTERSKATEGORIE_GROB  \n",
       "0                    4  \n",
       "1                    4  \n",
       "2                    4  \n",
       "3                    4  \n",
       "4                    3  \n",
       "\n",
       "[5 rows x 369 columns]"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load in the customer demographics data.\n",
    "customers = pd.read_csv('Arvato_data/Udacity_CUSTOMERS_052018.csv', sep=';')\n",
    "customers.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/pandas/core/ops.py:1649: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  result = method(y)\n"
     ]
    }
   ],
   "source": [
    "# clean customer demographics data.\n",
    "new_customers = clean_data(customers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Remove additional columns\n",
    "sub_customers = new_customers[['LNR', 'CUSTOMER_GROUP', 'ONLINE_PURCHASE',\n",
    "                               'PRODUCT_GROUP']]\n",
    "\n",
    "new_customers.drop(['LNR', 'CUSTOMER_GROUP', 'ONLINE_PURCHASE',\n",
    "                    'PRODUCT_GROUP'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "new_azdias shape (891221, 484)\n",
      "new_customers shape (140899, 483)\n"
     ]
    }
   ],
   "source": [
    "print('new_azdias shape', new_azdias.shape)\n",
    "print('new_customers shape', new_customers.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['GEBAEUDETYP_5.0']"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#  Check for missing column in `new_customers`\n",
    "mask = new_azdias.columns.isin(new_customers.columns)\n",
    "\n",
    "miss_att = list(new_azdias.columns[~mask])\n",
    "miss_att"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Add the missing column with default value of 0\n",
    "new_customers['GEBAEUDETYP_5.0'] = 0\n",
    "new_customers['GEBAEUDETYP_5.0'] = new_customers['GEBAEUDETYP_5.0'].astype('uint8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# replace missing value\n",
    "new_customers = pd.DataFrame(Imp.transform(new_customers.values),\n",
    "                         columns=new_customers.columns)\n",
    "\n",
    "# scale customers variables\n",
    "new_customers = pd.DataFrame(scaler.transform(new_customers.values),\n",
    "                         columns=new_customers.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# cluster predictions for the  customer demographics data.\n",
    "customers_pca = pca.transform(new_customers)\n",
    "labels_customers = kmeans.predict(customers_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": 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>% of data</th>\n",
       "      <th>cluster</th>\n",
       "      <th>data</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>99.429378</td>\n",
       "      <td>1</td>\n",
       "      <td>customers data</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.570622</td>\n",
       "      <td>3</td>\n",
       "      <td>customers data</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   % of data  cluster            data\n",
       "0  99.429378        1  customers data\n",
       "1   0.570622        3  customers data"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cus_class = pd.Series(labels_customers).value_counts()\n",
    "\n",
    "cls_customers = pd.DataFrame({'cluster': cus_class.index,\n",
    "                              '% of data': 100*cus_class.values/len(labels_customers)})\n",
    "cls_customers['data'] = 'customers data'\n",
    "\n",
    "# print cls_azdias\n",
    "cls_customers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Step 3.3: Compare Customer Data to Demographics Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABp8AAAQuCAYAAAAqdyVgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ac5c2668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 12))\n",
    "sns.barplot(x=\"cluster\", y=\"% of data\", hue=\"data\",\n",
    "            data=pd.concat([cls_customers, cls_azdias]))\n",
    "plt.xlabel('Cluster', fontsize=24)\n",
    "plt.ylabel('% of data', fontsize=24)\n",
    "plt.title('Compare Customer Data to Demographics Data', fontsize=24)\n",
    "plt.legend(fontsize=24)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "collapsed": 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>CJT_GESAMTTYP_1.0</th>\n",
       "      <th>CJT_GESAMTTYP_2.0</th>\n",
       "      <th>CJT_GESAMTTYP_3.0</th>\n",
       "      <th>CJT_GESAMTTYP_4.0</th>\n",
       "      <th>CJT_GESAMTTYP_5.0</th>\n",
       "      <th>CJT_GESAMTTYP_6.0</th>\n",
       "      <th>FINANZTYP_1.0</th>\n",
       "      <th>FINANZTYP_2.0</th>\n",
       "      <th>FINANZTYP_3.0</th>\n",
       "      <th>FINANZTYP_4.0</th>\n",
       "      <th>...</th>\n",
       "      <th>WOHNLAGE_4.0</th>\n",
       "      <th>WOHNLAGE_5.0</th>\n",
       "      <th>WOHNLAGE_7.0</th>\n",
       "      <th>WOHNLAGE_8.0</th>\n",
       "      <th>PLZ8_BAUMAX_1.0</th>\n",
       "      <th>PLZ8_BAUMAX_2.0</th>\n",
       "      <th>PLZ8_BAUMAX_3.0</th>\n",
       "      <th>PLZ8_BAUMAX_4.0</th>\n",
       "      <th>PLZ8_BAUMAX_5.0</th>\n",
       "      <th>LNR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>645165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>612595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>612606</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</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",
       "      <td>665447</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>507916</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 485 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   CJT_GESAMTTYP_1.0  CJT_GESAMTTYP_2.0  CJT_GESAMTTYP_3.0  CJT_GESAMTTYP_4.0  \\\n",
       "0                0.0                0.0               -0.0                0.0   \n",
       "1                0.0                0.0                1.0                0.0   \n",
       "2                0.0                0.0               -0.0                1.0   \n",
       "3                0.0                0.0                0.0                0.0   \n",
       "4                0.0                0.0                0.0                0.0   \n",
       "\n",
       "   CJT_GESAMTTYP_5.0  CJT_GESAMTTYP_6.0  FINANZTYP_1.0  FINANZTYP_2.0  \\\n",
       "0                0.0                0.0            0.0            0.0   \n",
       "1                0.0               -0.0            0.0            0.0   \n",
       "2                0.0               -0.0            1.0            0.0   \n",
       "3                0.0                0.0            1.0            0.0   \n",
       "4                0.0                0.0            0.0            0.0   \n",
       "\n",
       "   FINANZTYP_3.0  FINANZTYP_4.0  ...  WOHNLAGE_4.0  WOHNLAGE_5.0  \\\n",
       "0           -0.0           -0.0  ...           1.0           0.0   \n",
       "1            0.0            0.0  ...           1.0           0.0   \n",
       "2            0.0            0.0  ...           1.0           0.0   \n",
       "3            0.0            0.0  ...           1.0           0.0   \n",
       "4           -0.0           -0.0  ...           0.0           0.0   \n",
       "\n",
       "   WOHNLAGE_7.0  WOHNLAGE_8.0  PLZ8_BAUMAX_1.0  PLZ8_BAUMAX_2.0  \\\n",
       "0          -0.0          -0.0              0.0              0.0   \n",
       "1          -0.0          -0.0             -0.0              0.0   \n",
       "2           0.0          -0.0              0.0              0.0   \n",
       "3          -0.0          -0.0              0.0              0.0   \n",
       "4           0.0          -0.0              0.0              0.0   \n",
       "\n",
       "   PLZ8_BAUMAX_3.0  PLZ8_BAUMAX_4.0  PLZ8_BAUMAX_5.0     LNR  \n",
       "0              0.0              0.0             -0.0  645165  \n",
       "1              0.0              0.0              0.0  612595  \n",
       "2              0.0              0.0              0.0  612606  \n",
       "3              0.0              0.0              1.0  665447  \n",
       "4              0.0              0.0              0.0  507916  \n",
       "\n",
       "[5 rows x 485 columns]"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select peaple in the target class\n",
    "target_class = [1]\n",
    "target_pca = pca.inverse_transform(azdias_pca[np.isin(labels_azdias, target_class)])\n",
    "target_azdias = scaler.inverse_transform(target_pca).round()\n",
    "target_azdias = pd.DataFrame(target_azdias, columns=new_azdias.columns)\n",
    "\n",
    "# Get LNR variable\n",
    "target_LNR = LNR_azdias[np.isin(labels_azdias, target_class)]\n",
    "target_azdias['LNR'] = target_LNR.values \n",
    " \n",
    "target_azdias.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {
    "collapsed": 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>CJT_GESAMTTYP_1.0</th>\n",
       "      <th>CJT_GESAMTTYP_2.0</th>\n",
       "      <th>CJT_GESAMTTYP_3.0</th>\n",
       "      <th>CJT_GESAMTTYP_4.0</th>\n",
       "      <th>CJT_GESAMTTYP_5.0</th>\n",
       "      <th>CJT_GESAMTTYP_6.0</th>\n",
       "      <th>FINANZTYP_1.0</th>\n",
       "      <th>FINANZTYP_2.0</th>\n",
       "      <th>FINANZTYP_3.0</th>\n",
       "      <th>FINANZTYP_4.0</th>\n",
       "      <th>...</th>\n",
       "      <th>WOHNLAGE_4.0</th>\n",
       "      <th>WOHNLAGE_5.0</th>\n",
       "      <th>WOHNLAGE_7.0</th>\n",
       "      <th>WOHNLAGE_8.0</th>\n",
       "      <th>PLZ8_BAUMAX_1.0</th>\n",
       "      <th>PLZ8_BAUMAX_2.0</th>\n",
       "      <th>PLZ8_BAUMAX_3.0</th>\n",
       "      <th>PLZ8_BAUMAX_4.0</th>\n",
       "      <th>PLZ8_BAUMAX_5.0</th>\n",
       "      <th>LNR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.153986</td>\n",
       "      <td>0.438981</td>\n",
       "      <td>0.195428</td>\n",
       "      <td>0.005338</td>\n",
       "      <td>0.032109</td>\n",
       "      <td>0.178488</td>\n",
       "      <td>0.101494</td>\n",
       "      <td>0.022236</td>\n",
       "      <td>0.030009</td>\n",
       "      <td>0.832592</td>\n",
       "      <td>...</td>\n",
       "      <td>0.088633</td>\n",
       "      <td>0.008109</td>\n",
       "      <td>-0.030853</td>\n",
       "      <td>0.006746</td>\n",
       "      <td>-0.968481</td>\n",
       "      <td>0.004420</td>\n",
       "      <td>0.030635</td>\n",
       "      <td>0.001991</td>\n",
       "      <td>0.008158</td>\n",
       "      <td>910215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.016553</td>\n",
       "      <td>0.140481</td>\n",
       "      <td>-0.043271</td>\n",
       "      <td>0.205858</td>\n",
       "      <td>0.352741</td>\n",
       "      <td>0.347740</td>\n",
       "      <td>1.071068</td>\n",
       "      <td>0.024275</td>\n",
       "      <td>-0.011223</td>\n",
       "      <td>-0.073766</td>\n",
       "      <td>...</td>\n",
       "      <td>0.332211</td>\n",
       "      <td>0.111176</td>\n",
       "      <td>0.169287</td>\n",
       "      <td>0.053518</td>\n",
       "      <td>-0.408341</td>\n",
       "      <td>0.178000</td>\n",
       "      <td>0.025323</td>\n",
       "      <td>0.081406</td>\n",
       "      <td>0.196422</td>\n",
       "      <td>910220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.068808</td>\n",
       "      <td>0.167344</td>\n",
       "      <td>0.690430</td>\n",
       "      <td>0.005184</td>\n",
       "      <td>-0.037214</td>\n",
       "      <td>0.107954</td>\n",
       "      <td>0.655271</td>\n",
       "      <td>0.145848</td>\n",
       "      <td>0.015969</td>\n",
       "      <td>0.120313</td>\n",
       "      <td>...</td>\n",
       "      <td>0.201650</td>\n",
       "      <td>0.054984</td>\n",
       "      <td>0.182772</td>\n",
       "      <td>-0.000965</td>\n",
       "      <td>-0.131393</td>\n",
       "      <td>0.190337</td>\n",
       "      <td>-0.072414</td>\n",
       "      <td>-0.040423</td>\n",
       "      <td>0.092618</td>\n",
       "      <td>910225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.049052</td>\n",
       "      <td>0.619099</td>\n",
       "      <td>0.085748</td>\n",
       "      <td>0.046782</td>\n",
       "      <td>0.157899</td>\n",
       "      <td>0.130038</td>\n",
       "      <td>-0.129433</td>\n",
       "      <td>0.189162</td>\n",
       "      <td>0.047865</td>\n",
       "      <td>-0.109202</td>\n",
       "      <td>...</td>\n",
       "      <td>0.089831</td>\n",
       "      <td>0.080626</td>\n",
       "      <td>0.681299</td>\n",
       "      <td>0.101693</td>\n",
       "      <td>-0.092381</td>\n",
       "      <td>-0.005103</td>\n",
       "      <td>0.116265</td>\n",
       "      <td>-0.051851</td>\n",
       "      <td>0.015575</td>\n",
       "      <td>910226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.253009</td>\n",
       "      <td>-0.429921</td>\n",
       "      <td>0.569092</td>\n",
       "      <td>0.260723</td>\n",
       "      <td>0.193215</td>\n",
       "      <td>0.144168</td>\n",
       "      <td>0.187647</td>\n",
       "      <td>0.189173</td>\n",
       "      <td>0.078393</td>\n",
       "      <td>-0.027416</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.128274</td>\n",
       "      <td>0.098169</td>\n",
       "      <td>-0.016136</td>\n",
       "      <td>-0.006731</td>\n",
       "      <td>-0.744264</td>\n",
       "      <td>0.428140</td>\n",
       "      <td>0.113225</td>\n",
       "      <td>0.038011</td>\n",
       "      <td>-0.006381</td>\n",
       "      <td>910241</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 485 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   CJT_GESAMTTYP_1.0  CJT_GESAMTTYP_2.0  CJT_GESAMTTYP_3.0  CJT_GESAMTTYP_4.0  \\\n",
       "0           0.153986           0.438981           0.195428           0.005338   \n",
       "1          -0.016553           0.140481          -0.043271           0.205858   \n",
       "2           0.068808           0.167344           0.690430           0.005184   \n",
       "3          -0.049052           0.619099           0.085748           0.046782   \n",
       "4           0.253009          -0.429921           0.569092           0.260723   \n",
       "\n",
       "   CJT_GESAMTTYP_5.0  CJT_GESAMTTYP_6.0  FINANZTYP_1.0  FINANZTYP_2.0  \\\n",
       "0           0.032109           0.178488       0.101494       0.022236   \n",
       "1           0.352741           0.347740       1.071068       0.024275   \n",
       "2          -0.037214           0.107954       0.655271       0.145848   \n",
       "3           0.157899           0.130038      -0.129433       0.189162   \n",
       "4           0.193215           0.144168       0.187647       0.189173   \n",
       "\n",
       "   FINANZTYP_3.0  FINANZTYP_4.0  ...  WOHNLAGE_4.0  WOHNLAGE_5.0  \\\n",
       "0       0.030009       0.832592  ...      0.088633      0.008109   \n",
       "1      -0.011223      -0.073766  ...      0.332211      0.111176   \n",
       "2       0.015969       0.120313  ...      0.201650      0.054984   \n",
       "3       0.047865      -0.109202  ...      0.089831      0.080626   \n",
       "4       0.078393      -0.027416  ...     -0.128274      0.098169   \n",
       "\n",
       "   WOHNLAGE_7.0  WOHNLAGE_8.0  PLZ8_BAUMAX_1.0  PLZ8_BAUMAX_2.0  \\\n",
       "0     -0.030853      0.006746        -0.968481         0.004420   \n",
       "1      0.169287      0.053518        -0.408341         0.178000   \n",
       "2      0.182772     -0.000965        -0.131393         0.190337   \n",
       "3      0.681299      0.101693        -0.092381        -0.005103   \n",
       "4     -0.016136     -0.006731        -0.744264         0.428140   \n",
       "\n",
       "   PLZ8_BAUMAX_3.0  PLZ8_BAUMAX_4.0  PLZ8_BAUMAX_5.0     LNR  \n",
       "0         0.030635         0.001991         0.008158  910215  \n",
       "1         0.025323         0.081406         0.196422  910220  \n",
       "2        -0.072414        -0.040423         0.092618  910225  \n",
       "3         0.116265        -0.051851         0.015575  910226  \n",
       "4         0.113225         0.038011        -0.006381  910241  \n",
       "\n",
       "[5 rows x 485 columns]"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Select peaple in the no target class\n",
    "no_target_class = [0, 2, 3, 4, 5]\n",
    "no_target_pca = pca.inverse_transform(azdias_pca[np.isin(labels_azdias, no_target_class)])\n",
    "no_target_azdias = scaler.inverse_transform(no_target_pca).round()\n",
    "no_target_pca = pd.DataFrame(no_target_pca, columns=new_azdias.columns)\n",
    "\n",
    "# Get LNR variable\n",
    "no_target_LNR = LNR_azdias[np.isin(labels_azdias, no_target_class)]\n",
    "no_target_pca['LNR'] = no_target_LNR.values \n",
    " \n",
    "no_target_pca.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discussion\n",
    "\n",
    "**99 % percent of the customers' data fit in the second class which contains 30% of azdias data. So the company can take this population as future customers in Germany.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  <a class=\"anchor\" id=\"part2\">Part 2: Supervised Learning Model</a>\n",
    "\n",
    "Now that I've found which parts of the population are more likely to be customers of the mail-order company, it's time to build a prediction model. Each of the rows in the \"MAILOUT\" data files represents an individual that was targeted for a mailout campaign. Ideally, we should be able to use the demographic information from each individual to decide whether or not it will be worth it to include that person in the campaign.\n",
    "\n",
    "The \"MAILOUT\" data has been split into two approximately equal parts, each with almost 43 000 data rows. In this part, you can verify your model with the \"TRAIN\" partition, which includes a column, \"RESPONSE\", that states whether or not a person became a customer of the company following the campaign. In the next part, I'll need to create predictions on the \"TEST\" partition, where the \"RESPONSE\" column has been withheld."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 1: Exploring the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (18,19) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\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>LNR</th>\n",
       "      <th>AGER_TYP</th>\n",
       "      <th>AKT_DAT_KL</th>\n",
       "      <th>ALTER_HH</th>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <th>ANZ_HAUSHALTE_AKTIV</th>\n",
       "      <th>...</th>\n",
       "      <th>VK_DHT4A</th>\n",
       "      <th>VK_DISTANZ</th>\n",
       "      <th>VK_ZG11</th>\n",
       "      <th>W_KEIT_KIND_HH</th>\n",
       "      <th>WOHNDAUER_2008</th>\n",
       "      <th>WOHNLAGE</th>\n",
       "      <th>ZABEOTYP</th>\n",
       "      <th>RESPONSE</th>\n",
       "      <th>ANREDE_KZ</th>\n",
       "      <th>ALTERSKATEGORIE_GROB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1763</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1771</td>\n",
       "      <td>1</td>\n",
       "      <td>4.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1776</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1460</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1783</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>53.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 367 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    LNR  AGER_TYP  AKT_DAT_KL  ALTER_HH  ALTER_KIND1  ALTER_KIND2  \\\n",
       "0  1763         2         1.0       8.0          NaN          NaN   \n",
       "1  1771         1         4.0      13.0          NaN          NaN   \n",
       "2  1776         1         1.0       9.0          NaN          NaN   \n",
       "3  1460         2         1.0       6.0          NaN          NaN   \n",
       "4  1783         2         1.0       9.0          NaN          NaN   \n",
       "\n",
       "   ALTER_KIND3  ALTER_KIND4  ALTERSKATEGORIE_FEIN  ANZ_HAUSHALTE_AKTIV  ...  \\\n",
       "0          NaN          NaN                   8.0                 15.0  ...   \n",
       "1          NaN          NaN                  13.0                  1.0  ...   \n",
       "2          NaN          NaN                   7.0                  0.0  ...   \n",
       "3          NaN          NaN                   6.0                  4.0  ...   \n",
       "4          NaN          NaN                   9.0                 53.0  ...   \n",
       "\n",
       "   VK_DHT4A  VK_DISTANZ  VK_ZG11  W_KEIT_KIND_HH  WOHNDAUER_2008  WOHNLAGE  \\\n",
       "0       5.0         2.0      1.0             6.0             9.0       3.0   \n",
       "1       1.0         2.0      1.0             4.0             9.0       7.0   \n",
       "2       6.0         4.0      2.0             NaN             9.0       2.0   \n",
       "3       8.0        11.0     11.0             6.0             9.0       1.0   \n",
       "4       2.0         2.0      1.0             6.0             9.0       3.0   \n",
       "\n",
       "   ZABEOTYP RESPONSE ANREDE_KZ ALTERSKATEGORIE_GROB  \n",
       "0         3        0         2                    4  \n",
       "1         1        0         2                    3  \n",
       "2         3        0         1                    4  \n",
       "3         3        0         2                    4  \n",
       "4         3        0         1                    3  \n",
       "\n",
       "[5 rows x 367 columns]"
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mailout_train = pd.read_csv('Arvato_data/Udacity_MAILOUT_052018_TRAIN.csv',\n",
    "                            sep=';')\n",
    "\n",
    "mailout_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42962, 367)"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Data shape\n",
    "mailout_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Response values\n",
    "cout = mailout_train.RESPONSE.value_counts(dropna=False)\n",
    "df_sp = pd.DataFrame({'RESPONSE': cout.index, '% of data': cout.values})\n",
    "df_sp['% of data'] = 100*df_sp['% of data']/42962"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuIAAAI+CAYAAAD5H29rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ac3f7b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 6))\n",
    "sns.barplot(x=\"RESPONSE\", y=\"% of data\", data=df_sp)\n",
    "matplotlib.rc('xtick', labelsize=25)  \n",
    "matplotlib.rc('ytick', labelsize=25)\n",
    "plt.xlabel('RESPONSE', fontsize=15)\n",
    "plt.ylabel('% of data', fontsize=15)\n",
    "plt.title('Distribution of customers responses', fontsize=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Out of all the 42 962 individuals in the mailout campaign, only 532 of the individuals became customers. The dataset is highly imbalanced.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 2: Preparing the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/pandas/core/ops.py:1649: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  result = method(y)\n"
     ]
    }
   ],
   "source": [
    "# Cleaning data\n",
    "clean_mailout = clean_data(mailout_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    34663\n",
       "1      437\n",
       "Name: RESPONSE, dtype: int64"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Response values\n",
    "clean_mailout.RESPONSE.value_counts(dropna=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Cheking missing values\n",
    "sum_null = clean_mailout.isnull().sum(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAGHCAYAAABWNpPIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10aac5c5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "sns.distplot(sum_null.values/485, kde=False)\n",
    "plt.ylabel('count')\n",
    "plt.xlabel('% of nan')\n",
    "plt.title('Distribution of missing value per row')\n",
    "matplotlib.rc('xtick', labelsize=10)  \n",
    "matplotlib.rc('ytick', labelsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Get LNR variable\n",
    "mailout_LNR = clean_mailout['LNR'] \n",
    "\n",
    "# Drop LNR \n",
    "clean_mailout.drop('LNR', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Imputer\n",
    "new_mailout = pd.DataFrame(Imp.fit_transform(clean_mailout.values),\n",
    "                          columns=clean_mailout.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# feature scaling\n",
    "new_mailout = pd.DataFrame(scaler.fit_transform(new_mailout.values),\n",
    "                          columns=new_mailout.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# split data to features and labels\n",
    "response = new_mailout['RESPONSE']\n",
    "\n",
    "# drop RESPONSE variable\n",
    "features = new_mailout.drop('RESPONSE', axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 3: Modeling the data \n",
    "\n",
    "#### Evaluating Model Performance\n",
    "\n",
    "Since the data is highly imbalanced, using accuracy as a metric to evaluate model performance will give us a misleading result. Instead, I will use ROC-AUC to assess models performances.\n",
    "\n",
    "\n",
    "#### Supervised Learning Models\n",
    "Ensemble methods have become some of the most popular methods used to compete in competitions on Kaggle and used in industry across applications because of their ability to optimize both variance and bias. Therefore in this project, I will test the following three ensemble algorithms:\n",
    "\n",
    " - [RandomForestRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html)\n",
    " - [XGBRegressor](https://xgboost.readthedocs.io/en/latest/python/python_api.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Initialize the three models\n",
    "model_A = XGBRegressor(objective='binary:logistic', random_state=42)\n",
    "model_B = RandomForestRegressor(random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# randomize the data\n",
    "\n",
    "\n",
    "def randomize(X, Y):\n",
    "    permutation = np.random.permutation(Y.shape[0])\n",
    "    X2 = X[permutation, :]\n",
    "    Y2 = Y[permutation]\n",
    "    \n",
    "    return X2, Y2\n",
    "\n",
    "\n",
    "X_train, y_train = randomize(np.asarray(features), np.asarray(response))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Learning Curves\n",
    "\n",
    "\n",
    "def draw_learning_curves(X, y, estimator, title=\"Learning Curves\"):\n",
    "    \n",
    "    X2, y2 = randomize(X, y)\n",
    "    \n",
    "    train_sizes = np.linspace(.1, 1.0, 10)\n",
    "    train_sizes, train_scores, test_scores = learning_curve(estimator,\n",
    "                                                            X2,\n",
    "                                                            y2,\n",
    "                                                            cv=5,\n",
    "                                                            scoring='roc_auc',\n",
    "                                                            n_jobs=-1,\n",
    "                                                            train_sizes=train_sizes,\n",
    "                                                            shuffle=True,\n",
    "                                                            verbose=10\n",
    "                                                           )\n",
    "\n",
    "    train_scores_mean = np.mean(train_scores, axis=1)\n",
    "    train_scores_std = np.std(train_scores, axis=1)\n",
    "    test_scores_mean = np.mean(test_scores, axis=1)\n",
    "    test_scores_std = np.std(test_scores, axis=1)\n",
    "\n",
    "    fig = plt.figure()\n",
    "    plt.grid()\n",
    "    plt.title(title)\n",
    "    plt.xlabel(\"Training examples\")\n",
    "    plt.ylabel(\"Auc score\")\n",
    "\n",
    "    plt.plot(train_scores_mean, 'o-', color=\"g\",\n",
    "             label=\"Training score\")\n",
    "    plt.plot(test_scores_mean, 'o-', color=\"y\",\n",
    "             label=\"Cross-validation score\")\n",
    "\n",
    "    plt.legend(loc=\"best\")\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[learning_curve] Training set sizes: [ 2808  5616  8424 11232 14040 16848 19656 22464 25272 28080]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 16 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of  50 | elapsed:   19.4s remaining:  5.1min\n",
      "[Parallel(n_jobs=-1)]: Done   9 out of  50 | elapsed:   28.5s remaining:  2.2min\n",
      "[Parallel(n_jobs=-1)]: Done  15 out of  50 | elapsed:   41.5s remaining:  1.6min\n",
      "[Parallel(n_jobs=-1)]: Done  21 out of  50 | elapsed:   50.2s remaining:  1.2min\n",
      "[Parallel(n_jobs=-1)]: Done  27 out of  50 | elapsed:   58.8s remaining:   50.1s\n",
      "[Parallel(n_jobs=-1)]: Done  33 out of  50 | elapsed:  1.1min remaining:   34.1s\n",
      "[Parallel(n_jobs=-1)]: Done  39 out of  50 | elapsed:  1.3min remaining:   21.6s\n",
      "[Parallel(n_jobs=-1)]: Done  45 out of  50 | elapsed:  1.4min remaining:    9.6s\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed:  1.7min finished\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAGHCAYAAACNjTnqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzs3Xl8VOX1+PHPSchCQliDLLIpKmAVV9C41LZarAup9leL\nS6ui1vpVtMUWtRUrrVDFrri1tlKtbcW1ImhbqNqqVNzAHdCK7AoCCsREss35/fHcSW4mE0iGSebO\nw3m/XvPKzL3PvXlOZpJ78mxXVBVjjDHGmEzIyXQFjDHGGLP7skTEGGOMMRljiYgxxhhjMsYSEWOM\nMcZkjCUixhhjjMkYS0SMMcYYkzGWiBhjjDEmYywRMcYYY0zGWCJijDHGmIyxRMSYFInI+SISE5FB\nma6LSU5EDheR/4rIpyJSLyIjM10nY0xTloiYjBKR84KL+aGZrksKNHhkjIgcLCJ/EZHVIrJdRDaL\nyL+CJGm3/v0WkU7Aw0AP4HvAt4BVLZS9Ifgcfj7JvjODfZcmbBcROVdE5ovIRhGpEZENIjJPRL4t\nIvkJ5WMJj09F5G0RuVZEOieUvTuh7HYReUdEfiIiBbv4ozEmUsTuNWMySUTOA/4IjFLVxZmuT1uI\niAB5qlqToe9/EfBbYD3wZ+B/QAlwPHAKMFlVb8pE3aJARIYBS4ELVfXunZQtAN4CaoGRqloXbO8W\nnGOVqpaFyhcCs4ExwH+BucAGoCdwHHAy8CdV/XbomBgwH7g32NQFOBY4B3hIVceFyt4NjAMuBATo\nBnw1+H5/VdVvpfAjMSaSOmW6AsZEhYgUqur21pZXl8VnKgk5EpeE/Bc4WVWrQrtvCVqYDkjT9ypK\nOH+26BN83bqzgqpaLSL/h0sUfgjcEOyaDpQCJyYc8hvgy8AVqnpbwr5fi8jQYH+id1X1vtDr3wdJ\n0NdEJD8hqa1T1Vmh178VkeeBs0TkSlXduLO40ikbPwfZWOfd0W7ddGuyh4jkB83S/wuaqVeLyPQk\nzd/jReSpoIl8e9D0fUmS860UkTkiMkZEXhaRz4CLg30xEblFRL4qIm8G53lLRE5MOEezMSKh8x4t\nIi+KyGcislxEmv0HKyIjReQZEakSkTVBE/34Vo47uR6IAeck+0OrqotV9d7g+xyXrNtBRAYH288N\nbbtHRCpEZG8R+buIbAP+IiK3BtsLk8QxS0Q+CFqI4ttOEpFng+6HbSLyuIjsn3Bcn6ALYk3wM/5A\nRGa3InZE5Esi8lxw/k+C44aH9t8N/AfXdfZwEOfTOzqnqj4J3Af8UET2EZEy4NvAb1T1zdC5B+Ba\nKv6RJAmJn2u5qv5uZ3EENgT1rGtF2QW4FpK9wxvF+V7wOf1MRNaLyO9EpHuSclNEZJ2IVAa/KyOC\nz+0fQ+XiXaafF5E7RGQDsCa0v7+I/DH4PvHfj/GJlRWRy4N9lSLycfC7dmZofxcR+Y2IrAjOs0Fc\nV9fBCec5Q0ReCX5XNorIn0Wkf0KZpJ/dVvxMTYZZi4iJvOACNxc4CrgTWAYcCEwE9gW+Fip+Ca6J\n/THcH/axwB0iIqr621A5BYbjLjx3Ar8H3gntPzY47x1ABXAF7oI2SFU/CZ0jsW9Tgzo9BMwE7gEu\nAO4WkVdUdWkQU3/g30A9MA2oAi7CtbDssL9U3HiCLwHPquq6HZVNqFdry3UC5gHPAd8P6rYKuBTX\n5fNIQl1OBf4YtBARJF33AP8ErgKKgP8DnhORQ1R1dXD434ARwC3B+ffAtSIMAuJlmhGRE4C/A8tx\nCVln3PuzQEQODc7/O2AtcC0wA3gZd8HfmSuBr+A+D71wF98pCWVOwv0T99dWnC9RoYj0Cp4XA8cA\n5+K6W2KtOH6v4OsnCdt/H5znj7h49wIuBw4WkaNVtT4odxMwCff7MR84CPdetzTu5A7gI+AnQX0R\nkT2AF3Gf3VuATbifyUwRKVHVW4Jy3w7q8iCuBakQGAkcAdwfnP9O3O/ZrbgusF7Bz2QE8FpwnvOD\nuF4ErsG1dH0POCr4PG0LztXSZ9dEnarawx4ZewDn4f6gHbqDMt/E9d2XJWy/ODj2yNC2giTH/wP4\nX8K2FcGxJyQpHwM+A4aEth0YbL80Sd0HJTnvUaFtpcH5bg5tuwWXKB0Y2tYd90e9yTmT1C9el1+1\n8md8XHDOzydsHxyc59zQtruDslOTnGcN8GDCtjOC8kcHr4uBj4HfJpTrjbt4/i543S343lem8Jl5\nFfgQ6JbwM6kD7k6IOwZ8rY3n/3ZwXD1wapL9vwz2HZiwPQ93IY0/eib5XNUHX8OPR3BjjcJl7wa2\nhc61N+7CWg+8llD2mOA84xK2fznYfmbweg9covtwQrkfB+X+mPDZjuFalSSh/F24JK97wvb7gve+\nIHj9KPDGTn7WnwC37GB/J9wYqNeA/ND2k4P6Xd+az649ov2wrhmTDb6O+2/pXRHpFX/gWhQE+GK8\noKpWx5+LSNeg3LPA3iJSknDeFeqa45P5l6quDJ33TdyFYe8WyoctUdXnQ8duwrW2hI89EViooSZ/\nVd1C6/7L7hp8rWhF2VQl61Z4CDhZRIpC28YB61T1v8HrL+OSjPsT3ivF/Ucbf68+w10Uv5DYfbAj\nItIX91/83araMPYj+Dn+C3eB2lWbgq9VuDE4ieI//08Ttp8MbAw9ViY59jHghOBRDvwM15owK0nZ\nLqFzvQf8HNc1c1pCua8DW4CnEn7mrwZ1jP/MTwBycWOLwm5N8r3BvWd/0OAqH/I1XAtlbsL3m49L\npuMz4LYAA0Tk8BbOHy9zhIj0a2H/4bgE6g4NjZ9R1b/jWkZPSXJMa7vETERYImKywb7A52j6R34j\n7uKuuD9UAIgbm/GkiHyK+yO3Edf1Ae4CGbZiB99zTZJtn+Cmgu5Msm6FxGMH4y4uiZJtSxRvik5M\nrNKlTlXXJtn+AK6bpRxARIpxF9EHQ2X2xSWH/6bpe/URLknZAyC4qFwdHL9B3FiZSSLShx0bHHx9\nN8m+pUCpJEyFbQsR6YJrrVoG5OMGqyaKJ4BdErYvoDHJmN/Ct1irqk8Hj8dVdTIwGThdRBIvqp/h\nZkCdAJwPLMH9/D5LKLcvLgH4iOY/82Iafz/iY2+afMbUdTUmdvXErQy/EJHewfe6mOa/j3+k6e/j\ndFwi9JKIvCsit4nIUQnnvwo3qHqNuDFV14vIXqH9g4NzJnu/l9H4eYhr6bNrIszGiJhskAO8iRsT\nIkn2rwEQkb2BJ3EXpInB9hrcf03fo3ninfgHPay+he3Jvn86j22N9wi6dVpZvqXxIbktbK9OtlFV\nXxSRlcA3cH385bh+/3AikhN8v2+SfExGw4BMVZ0hInNw/+GfCPwUN1D0i6r6eovRtK+f4cYgjAXO\nAr4vIner6sJQmWW49/IA3OcSAFXdDDwNDeNkWuup4HyfB54Iba9X1X/HX4jI/OB730nTVpEc3M/6\nbJJ/xnZldk3i70j8d+gvwJ9aOOYNAFVdJm4K9am4cTdfAy4VkZ+o6k+CMg+JyLPA6bipyT8ArhaR\n01V1Xgr1TfrZNdFmiYjJBstxazv8eyflxuL+ix2roUGcInJ8e1YuRauAfZJs33dnB6rqZ+JmgHxR\nRPbUnQ9Y/QR3gUrsAhnSmoomeBC4Img5GAesVNWXQvuXB99ro6rucJYKgKquAH5N45TX13FjIc5t\n4ZD4gmTDkuwbDmxS1R0lmC0KuhAuBWao6msi8i4uxt8FgyLjg0n/gUs2zyF5l0pbxf8OJ7awNKGq\n60Xk18CPRWR06Oe+HNdy8ny4azKJ+M9un9BzRKQnrWvpA5fUVAC5rXx/P8N16T0kboG5R4FrReTG\neFeLqm7Adaf8TkRKcV1K1+IGna7CfZ6G4carhA2jhQXqTHaxrhmTDR7E9TV/O3GHiBSGxizEWyJy\nQvu74Zq1o2YeUCahJceDC8LZrTz+J7g4/xx0kTQhIodJ47TcVQSDVROKXUrbV4Z9ADfD4nxcK8YD\nCfvn4bqOfhRceBLrVRp87SzNVwhdgbvItbhyqKrGBy6eJyLxsRqIyAG4/6ifaOnYHRG3Cu2dwAe4\nwZuomxZ9OY0ztOJ1WIPrhjhJRC5r4ZRt+dtajnsfXmtF2VtxrRTXhLY9iEtmfpxYWERyg98BcC0v\n9bgZTGGXt7aiQTL2CPD/RORzSb5faeh5z4Rj63CtlQLkiUhO+D0MymzCvQfxz8AruC6mS0QkL3Tu\nk3Azax5vbd1NdFmLiIkCAS4M/rgk+g1u1dBv4BZ0+iJuAGEu7g/RGbgL0GJcv3wt8LiI3IkbQ3ER\nrtm6b3sH0UY347ovnhSRW4FKXF1X4f473WGCoKoLg4vg7cAyEQmvrPoF3MXt2qDsNhF5CNeSAe4/\n6FNxM1naRFVfFZHluHE3+TTtlkFVK8QtDHYvsFhE7sf9Fz0I10W2ADfVdj/c4MoHcWMf6nBN93uw\n81aGSbjpuy+IyEzcuJUJuJafn7Q1psB3gYNxM2wqQ/HMDbqPrheRB0LjD76Ha1G6Rdy6GHNxF8xS\n4Ghc69zSJN9nPxE5J3heBJThWn/+RyvWvFDVj8WtkfJ/IjJMVd9R1WeDz/s14tbfiP8e7IcbyHoF\n8DdV/UhEZgBXishjuOnVB+HG6Wyk+Weupa7Ea3CfsRdF5A+4968ncBhuWnk8GZkvIutxv68bgP2B\ny4DHVbUySJDWisjDuJawT3HjiA7HTaNGVetE5Gpc4vesiMzC/S5fAbyP+/tgsl2mp+3YY/d+0DgF\ntqVH/6BcLq7/+A3cbIZNwEu4i22X0PlOwTXtVuIuuN/H/feeOM32feCxFupUj2ueT9z+PjAzSd13\nel7c4M2nEraNxDU3V+EGuP4Q999pPdC7lT+/g3GJ2hpgO2765FO4+6pIqFwvXNJQEfzsbsclcvU0\nn767dSff84bguGU7KPN5XLLwcfBevItbV+WQYH9P3KDQt3EtKB8Dz9PKqba4mSDP4i5en+Ca/Icl\nlIlPW97hOYE9gzrMbmH/wGD/ownbBZdE/At3Ia/GXXDn45LK/ITyiZ/tGlzieQdQmlC2xfcBt0ZI\nDaHptsH2C3G/E/GB2q8RjHlJqPMUYF1Q7klcl9ZG4PYkn+2k0+pxycYtuMGs24PzzQcuCJW5CPe5\n/yj4jL8L3Ejw+4qb7nwT7p+ILcHPeDFwcZLv93Vc60hVUNc/Af1a+zOzR7Qfdq8ZYyJERH6DW8ei\ni9ovp+kAQcvEJ8C1qnpjputjdj+RGCMiIseKWxZ7nbhlhctbccwXRGSRuGWB3xV387TEMmeIyFJx\nSx6/3kLTvzEZIQnLpQdrMXwTeM6SENMeEj9zgYm4bpn/dGxtjHGiMkakGNeMOBO37PMOicgQ3CCl\nO3CD+04A7hKRD1T1X0GZo3Ar/V2NG8B2DjA7GP2+pB1iMKatForIf3BjCfriloIvofGGa8ak2zhx\nS6b/Hdc1cyxwJvBPbTpF2ZgOE7muGXG3yj5NVefsoMx04CRVDc84mIVb8vnk4PX9QJGqlofKLARe\nVdVL2y0AY1pJRKbi+r4H4P4jXQT8RHc+TdmYlIjIIbiFxg7GrRC7AXgYuE7tLrUmQ6LSItJWR+IG\nWYXNw61HEFeGuydEYpmvtmO9jGk1bVxV05gOoaqv4maZGRMZkRgjkoK+NF+1cQPQNbQ2QUtlojaN\n0xhjjNltZWuLSLsIBgueSOOUNGOMMca0TiFufZ156m550CrZmoisx90PIqwPsE0blzhuqcz6HZz3\nRFp391NjjDHGJHcObrJIq2RrIrIQtxpg2Jhge7jM8bhFd+K+nFAm0UrAre8YrA3Y75l+PH5fZlYR\nro/VUxurpba+lrpYHbX1tdTGgufh7bFaaupq+NFVP2LrF7Y2nuCfuFtNBQqfKuTYi4+lqraKqtoq\nPq35lKraKiprKqmsraS2vnaH9emc15nivGKK84spyiuiS34X97yTe16UV9Swv6FMXheK8oPtwb7O\neZ3JkZ33Ct4842Ye3PggOlCbxSNrhHG9xzHpu5Pa9kONkIkTJ/LrX/965wWzgE+xgMUTZT7FAn7F\ns3TpUr75zW9Cwl2bdyYSiUhwr4x9aFxSeG8ROQj4WFXXiMiNuBU242uF/A64LJg980dcwvF14OTQ\naWcA/xGRK3HTd8/CLUHc7H4lIa47phToDznv5XDGV8/g0EMPTUeY7e7nP/k5W/ttbfwpFgL9g+cK\nfbv2Zf5VLd2dHGrqa6iormBb9TYqaoKvO3sdfF1Xvc69/tS9rqmv2WFdS/JLKCkooWtBV0ryg68J\nr+e9Mg89Q5PGo/2UF+a+kDXvTTLdunXL6vqH+RQLWDxR5lMs4F88gTYNbYhEIoK7t8C/cVMYlcbZ\nLn/Cra3QF7fMMgCqulJETsHNkrkCWAtcqKpPhsosFJGzcffEmIa7l8NXW7uGSM57OYx4bwRT75i6\nq7F1mLEnjOX2928nNjS4SeinjftyludQ/uUdrxOXn5tPr6Je9Crqtct1qa6rpqKmos2JzftV77Ot\nehvbtm9ja2xr07tdhOJBYEPNBn705I/Yp9c+DO0xlKE9h9K/pH+rWluiYP36HfUSZhefYgGLJ8p8\nigX8iycVkUhEVPUZdjCDR1XHJ9n2LK6FY0fnfQR3p8g26fdsP84oP4Opd0ylpKSkrYdnzLTrpvH0\nmKdZqktdMlIBqEtCOjqpKuhUQEGnAkqLSndeuAV73bsXK3VlYzJSEdqpEKuO8Zc3/8LabWvR4H5d\nhZ0K2bvH3i4x6TGUfXruw9Ce7vmQ7kPIy81L/DYZs27dukxXIW18igUsnijzKRbwL55URCIRiZrH\n//p4VjaVlZSUsHD+QiZPncycuXNYn7OevnP7Un5CedYlVZCkhadf476c5Tl85/TvMGPiDLbXbWfl\nlpW89/F7LP94Ocs/Wc57H7/H39/7Oys+WUFtzI19yZEcBncb3JCYJCYqxfnFHRrfYYftMI/OKj7F\nAhZPlPkUC/gXTyosEfFMSUkJM6bPYAYzuO+++zj77LMzXaWUNWvhOZCkLTyFnQoZXjqc4aXDm52j\nPlbP2m1rG5KTeKLy4roX+eubf+XTmsb+nj7FfZokJvHunn167kOvzr0Qaemu6Kk588wz03q+TDrr\nrLMyXYW0sniiy6dYwL94UhG5Jd4zSUQOBRYtWrQoK1tEfFRRUeFaeJ6cQ21OLXmxPNfCM3nXW3hU\nlY1VG5u0oiz/ZHnD648qP2oo27Wga2Ni0iOUrPQcyoCuA1o9LqWiooJrb7iWuU/OpTa3lrz6PMae\nMJZp103LuhYrY4wJW7x4cbyF5zBVXdza4ywRCbFEJNpUNe2tEjtSUV3RJDEJJyprtq0hpq7LqCC3\ngL167NXiuJSCTm6x34qKCsrGlLF0n6CFR3AtPO/nMOJ/I1g4f6ElI8aYrJVqImJdMx4bP348d999\nd6arkTYXXHBBh8ZTUlDCwX0P5uC+BzfbV11XzcotK5slKvPfn8/vFv2uYfqyIAzqNoihPYeyYe4G\nluyzBN0nSP5nA6dBbGiMpbqUyVMnM2P6jA6LL518+6xZPNHlUyzgXzypsETEY2PG+HVvqyjFU9Cp\ngGGlwxhWOqzZvpjGWLdtXbOungWvLUDPCbVADg0dMzTGo3MezdpEJErvTTpYPNHlUyzgXzypsK6Z\nEOuaMe1FVRk4eiDrTt3BVL1ZMOSSIZQNLKNsQBllA8s4qM9BkZpybIwxLbGuGWMiTETIq89zy/Ul\nG+ai0DuvN6ePOJ2FaxfyyNJHqKmvobBTIYf3P9wlJgPKOHLAkfQr6ZfkBMYYk50sETGmgzRbFyUk\nZ3kOZ518Fr868VeAG4Py6vpXWbhmIS+se4H737qfnz//cwAGdxtM2cAyjtzzSMoGlnFw34PJz83v\n0FiMMSZdsmMtbJOSBQsWZLoKaZXt8Uy7bhoj/jeCnPdyXMvIKtysmfjtBCY3rnxb0KmAIwccycSy\niTzw9QdYPXE1ayeu5aEzHuLr+3+d1VtXc/WTV3PEXUfQ9cauHP3Ho/nB/B/wyJJHWLet41dqzPb3\nJpHFE10+xQL+xZMKGyMS4tsYkfLycubMmZPpaqSND/GE10VZv3w9fYf2TXldlJr6Gl5b/xoL1yxk\n4dqFvLD2BVZtXQXAwK4DG8aaHDngSA7pe0jDNOL24MN7E2bxRJdPsYBf8dg6ImngWyJSVVVFUVFR\npquRNr7FU1lZSXFxepeV/6DiA15Y+wIvrH2BhWsX8soHr7C9bjv5ufkc1u8wjhxwZMNA2AFdB6Tt\n+/r23lg80eVTLOBXPJaIpIFviYgxNfU1vL7+9YbEZOHahazcshKAPUv2bJyhM6CMQ/odQmGnwpS+\nT0cvNmeMiR6bNWOMaSY/N59Re45i1J6juPyIywFY/+l6l5gEXTqTn57MZ3WfkZ+bzyF9D2nozikb\nWMbArgNbTDBsuXpjTDpYi0iItYiY3VFtfS1vbHijYZzJwrULef+T9wHoX9K/MTEZUMZh/Q+jsFOh\nLVdvjGkm1RYRmzXjsUmTJmW6Cmll8bSPvNw8Dut/GBNGT+AvX/sLy69Yzvrvr+exMx/j3JHnsqlq\nE9f/53qOufsYut7YldF/GM2xFx3Lkn2WENsnSELmAxIsV7+PW64+m0XlvUkXn+LxKRbwL55UWNeM\nxwYNGpTpKqSVxdNx+nTpQ/mwcsqHlQNQF6vjzQ1vNowzuX/x/U2Xq+/W+DQ2NMbsObOzdrl6iPZ7\nkwqf4vEpFvAvnlRY10yIdc0Ys3OtXa6+94W9GdF7BMN7DWd4aeNjULdB5ObkdlyFjTEdwgarGmM6\nRGuXq58wegLLNi3jpQ9e4s9v/JnP6j4DoLBTIfv12s8lJqEkZb9e+1Gcn97pzMaY6LNExBjTZq1Z\nrv7Hx/24YVtMY6zZuoZlm5Y1PjYv49lVz7L+0/UN5QZ1G8Tw0uGMKB3RpBWlT3Efmx5sjKesaybE\nt66ZZcuWMXz48ExXI20snuhoNmtmE1DqkpAR77Vt1syW7Vt4Z9M7TRKUZZuW8d7H71EXqwOgW0G3\nJolJ/DG0x9B2uTvx0qVLGTFiRNrPmynZ/FlL5FMs4Fc8tqBZGviWiPi0dDBYPFGTzuXqk6mtr2X5\nJ8ubtqJsWsbSTUvZVr0NgE45nRjaY2jSJKV7Yfc2xxNfF2X9++vpu3dfb9ZFyfbPWphPsYBf8Vgi\nkga+JSKrV6/2akS2xRNdq1atYvDgwR3yvVSVDZUbmiUoyzYta7jXDkCf4j7NkpMRpSMY2G0gOdJ0\n5YJmLTxbgW7+rIvi02fNp1jAr3gsEUkD3xIRY3Y3lTWVvLv53WbdPO9seofq+moAOnfqzLDSYU0G\nyz72u8d4YOsDbl2UBDnv5TCh/4Ssno5sTEewWTPGmN1ecX4xh/Q7hEP6HdJke32sntVbVzcbLPv0\niqf5qPIj+DtwbvJzxobGmDN3DjOwRMSY9mCJiDHGe7k5uezVYy/26rEXJ+17UpN9m6s2M/yx4WyS\nTckPFlhdtZov3PMFPtf7c4zoPYL9e+/PiNIR9O3S12bzGLOLbIl3j02fPj3TVUgriye6sjmWXkW9\n6EIXty5K3ILQc4USKaF3cW+eWfUMV867kuPvPZ7+v+pPz5t7ctTMo7hozkX88vlf8o///YOVW1YS\n0+ZdPJmUze9PIp9iAf/iSYW1iHisqqoq01VIK4snurI9lmbrotQ27stZnsN5Y89jxhmua6a2vpb3\nP3mfJRuXsHTTUpZsXMKr61/lvjfva1i0rSiviBGlI1zrSen+Da0oe/fYm045Hf9nN9vfnzCfYgH/\n4kmFDVYNscGqxuyeWrybcBvWRYlpjNVbV7sEZePSJonK1uqtAOTn5rNfr/0YUdrYvbN/7/3Zr9d+\nFHQqaLf4VNW6kEy7s8GqxhiTopKSEhbOX+jWRZk7h9qcWvJieW5dlDtaty5KjuQwpPsQhnQfwsn7\nntywXVVZ/+n6JonJ0k1L+e0rv3UDZYNjh/YY2qwFZXjpcLrkd0kppvC6KLW5teTV53mzLorxi7WI\nhFiLiDEGOq4FYXPVZpZuWtqsBWXNtjUNZQZ1G9Sk9STe5dOzc88Wz9tiC48n66KYaLIWEdPMpk2b\nKC0tzXQ10sbiiS6fYgHYvHlzh8TTq6gXxww6hmMGHdNke0V1RcMqsvEEZe67c5nx4oyGgbB9ivs0\nTVCCVpQ+xX249oZrXRISXxelEih2U5GX6lImT52cteui+PZZ8y2eVFiLSIhvLSI+LR0MFk+U+RQL\nRDee7XXbeXfzu81aUN7d/C61MTfCtnthdz6b+RnVZ1c33h35PuDs4LnCkLlDWLFoRSZC2GVRfW9S\n5VM8Wd8iIiKXAT8A+gKvA5er6ss7KX8ZMARYBfxMVf8c2n8ecDdNb1a+XVWL2iWACJoyZUqmq5BW\nFk90+RQLRDeewk6FjOwzkpF9RjbZXherY/nHy1m6aSlvf/Q20/4yrfGvHsAXQs8FttRv4ZV1r3BA\nnwMo7FTYEVVPm6i+N6nyLZ5URKJFRETGAX8CLgZeAiYCZwD7qWqzVYZE5P+AG4GLgFeAI4A/AGep\n6hNBmfOA3wD70fgrqaq6cQf18KpFxBize9rr0L1YWb6yaTISp8C9wHmQK7ns12s/RvYZyUF9DnJf\n+x7EniV72iwb02bZ3iIyEbhTVe8FEJFLgFOAC4Cbk5T/ZlD+4eD1ShEZBVwNPBEqt8PEwxhjfNRs\nXZSQnOU5fOf/fYfzLjyPNza8wRsb3uD1Da/zz/f+2TDNuEdhjybJycg+IzlgjwPonNe5o0Mxu4GM\nJyIikgccBvwsvk1VVUSeBMpaOKwA2J6wbTswWkRyVbU+2NZFRFbiVpBdDPxIVZeks/7GGBM1066b\nxtNjnmapJl8XZfod0ykpKeGIAUc0HKOqrN66uklyMm/5PG596VYUJUdy2LfnvhzU9yBG7uFaTkb2\nGcnArgOt9cTskigs8V4K5AIbErZvwI0XSWYecFHQlYKIHA5cCOQF5wN4B9eiUg6cg4v1eRHpn9ba\nR9jMmTMzXYW0sniiy6dYIPvjia+LMqH/BIbMHUL3u7szZO4QJvSf0OLUXRFhcPfBjB02lms/fy0P\nnvEgyyYy/SEBAAAgAElEQVQs49MffcpLF73E70/9PWOGjmHDpxv4xcJfMHbWWAb/ZjA9b+7Jcfcc\nx+V/v5w/LPoDL617icqaynaLLdvfm0S+xZOKKCQiqbgB+AewUERqgUeBe4J9MQBVfUFV/6Kqb6jq\nc8DXgI3AdzJQ34xYvLjVXXRZweKJLp9iAT/iKSkpYcb0GaxYtIKzTj6LFYtWMGP6jDavH1KUV8So\nPUdx4aEXcstJt/Cf8//Dx1d9zOrvrebxsx5n0lGT6NelH0+ueJJLnriEI+46gpIbSxh22zDOeOgM\nbnjmBua8M4eVW1aSjjGJixYt2uVzRIkPn7VdpqoZfeBaMWqB8oTt9wCP7uTYXKA/ruHxEmDLTso/\nCPx1B/sPBbRPnz46duzYJo8jjzxSH330UQ2bN2+ejh07VhNdeumletdddzXZtmjRIh07dqxu3Lix\nyfYf//jHetNNNzXZtmrVKh07dqwuXbq0yfZbbrlFf/CDHzTZVllZqWPHjtXnnnuuyfb77rtPzz//\n/GZ1+8Y3vmFxWBwWh8XRLnFU1VTpy+te1pmLZ+pxlxynA78yUHvc1EOZgjIFLflJifY8uKeePv10\nvfOVO3XhmoVaUV2x0zi2bduml0+6XIccMkR77dtLO3ftrJdPuly3bdvWLnGEZfP70Z5x3HfffQ3X\nxvg18/Of/7zihkMfqm3IA6Iya+YF4EVV/W7wWoDVwC2q+vNWnuM/wBpV/VYL+3OAt4EnVPUHLZSx\nWTPGGJNGqsoHFR/w+obXG8aevLHhDd7Z9A71Wo8gDO05tNng2CHdh5AjObZKbBbJ9lkzvwLuEZFF\nNE7fLSLobhGRG4H+qnpe8HpfYDTwItATuBL4HHBu/IQich3wAvAe0B24ChgE3NUhERljjEFE2LPr\nnuzZdc8m9+DZXredJRuXuORk/eu88dEb3PLiLWz+bDMAJfklHNjnQLb9cxtL9lmC7hP6p1n8WCXW\nOJFIRFT1QREpBX4K9AFeA07Uxqm3fYGBoUNyge/j1gipBf4NHKWqq0NlegC/D479BFgElKnqsvaM\nxRhjzM4Vdirk0H6Hcmi/xtZnVeXDTz9skpw89OpD6DnJW+5jQ2Pc98h9lH+nnCHdhzCw20Dyc/M7\nKgSTJpHomokK37pmfFo6GCyeKPMpFrB4okJVGTh6IOtOXde4MbxcPcAs4ExAQBD6l/RncPfBDOk+\nhMHdmn4d1G1Q5NZCydb3Jpls75ox7WDChAmZrkJaWTzR5VMsYPFEhYiQV5/X9EYdo0MFFAYXDebJ\nK55k5ZaVrNqyyn3d6r4uWL2AtdvWNtwoENzNAltKVAZ3H0yX/C4dGSKXXXZZh36/KLIWkRDfWkSM\nMSbbXXHVFdy+voVVYt/LYUL/CTscI1JbX8u6inVJE5VVW1exeutq6mJ1DeV7de61w0Sle2H3XY6p\noqKCa2+4lrlPzqU2t5a8+jzGnjCWaddNy+qBt6m2iFgiEmKJiDHGREuLs2aCVWJ3ddZMfayeDz/9\nsMVEZdWWVVTXVzeU71bQrcVEZUj3IfTs3HOHK836PAvIumaMMcZ4J75K7OSpk5kzdw61ObXkxfIo\nP6GcqXdM3eWLdm5OLgO6DmBA1wEcM+iYZvtjGuOjyo+SJipPrXiKlVtWUlVb1VC+OK+4ofVkSLch\nzZKWaTdMc0nIPqEWnt18FpC1iIT41iIye/ZsTjvttExXI20snujyKRaweKLs0Ucf5fTTT890NRqo\nKps/29xii8rKLSvZVr2t8YB7gW/ROOZlKTAifjLoP7s/Lz77IqVFpRR2KuzQWHaVtYiYZmbNmuXN\nHx+weKLMp1jA4omy+++/P1KJiIhQWlRKaVEph/c/PGmZLdu3sHLLSlZ+spLz/3Y+W2Vr4863aExE\nBD7Y/gEDfzUQBLrkd2k4d2lRKb2Lejd5nbitZ+ee5ObktnvMieJjXh6e83BKx1uLSIhvLSLGGGOi\nZa9D92Jl+crGFpEwhX6P9uOuh+9iU9WmZo+NVRsbnm+u2ozS9PotCD0699hpwtKwrbg3Jfklu3T3\n5CZjXopibvUuaxExxhhjomnsCWO5/f0WZgEtz+GMr5zRZAXaltTH6tmyfUuT5KQhYancyKbP3PM3\nP3qzYVtFTUWz8+Tl5LU6aYk/D3cZXXvDtY1jXj5I7WdiiYgxxhjTQaZdN42nxzzNUk0+C2jqHVNb\ndZ7cnFx6FfWiV1GvVn/v6rpqNn+2uWnCEk5iPnPb3t38bkMLTE19TbPzFOcVNyQnbz/2NrGzmidV\nbWGJiDHGGNNB2nsW0I4UdCqgf0l/+pf0b1V5VaWytrJ5whIkKRsrN/JW3lvJu5naIGfXDjdRNn78\n+ExXIa0snujyKRaweKLMh1hKSkqYMX0GKxat4MsHfJkVi1YwY/qMyK0fIiJ0ye/CXj32YtSeozhp\n35P41kHfYmLZRH52/M/4Q/kf6JvfF3ZxqKklIh4bM2ZMpquQVhZPdPkUC1g8UeZTLJD98Yw9YSw5\n7+9aKmGzZkJs1owxxhjTeumYNWMtIsYYY4xJSXzMy4T+E+j3bL+UzmGJiDHGGGNSFh/z8vhfH0/p\neEtEPLZgwYJMVyGtLJ7o8ikWsHiizKdYwL94UmGJiMduvvnmTFchrSye6PIpFrB4osynWMC/eFJh\ng1VDfBusWlVVRVFRUaarkTYWT3T5FAtYPFHmUyzgVzyp3vTOWkQ85suHO87iiS6fYgGLJ8p8igX8\niycVlogYY4wxJmMsETHGGGNMxlgi4rFJkyZlugppZfFEl0+xgMUTZT7FAv7FkwpLRDw2aNCgTFch\nrSye6PIpFrB4osynWMC/eFJhs2ZCfJs1Y4wxxnQUmzVjjDHGmKxjiYgxxhhjMsYSEY8tW7Ys01VI\nK4snunyKBSyeKPMpFvAvnlRYIuKxq666KtNVSCuLJ7p8igUsnijzKRbwL55U2GDVEN8Gq65evdqr\nEdkWT3T5FAtYPFHmUyzgVzw2WNU048uHO87iiS6fYgGLJ8p8igX8iycVlogYY4wxJmMsETHGGGNM\nxlgi4rHp06dnugppZfFEl0+xgMUTZT7FAv7FkwpLRDxWVVWV6SqklcUTXT7FAhZPlPkUC/gXTyoi\nM2tGRC4DfgD0BV4HLlfVl3dS/jJgCLAK+Jmq/jmhzBnAT4My7wLXqOo/dnBOr2bNGGOMMR0lq2fN\niMg44JfA9cAhuERknoiUtlD+/4BpwI+B/YEpwO0ickqozFHAfcAfgIOBx4DZIrJ/+0VijDHGmLaI\nRCICTATuVNV7VXUZcAlQBVzQQvlvBuUfVtWVqvoA8Hvg6lCZK4B/qOqvVPUdVf0xsBiY0H5hGGOM\nMaYtMp6IiEgecBjwVHybuv6iJ4GyFg4rALYnbNsOjBaR3OB1WXCOsHk7OKd3Nm3alOkqpJXFE10+\nxQIWT5T5FAv4F08qMp6IAKVALrAhYfsG3HiRZOYBFwVjOhCRw4ELgbzgfATHtuWc3rnggpYalLKT\nxRNdPsUCFk+U+RQL+BdPKqKQiKTiBuAfwEIRqQUeBe4J9sV29eQnn3wy5eXlTR5lZWXMnj27Sbn5\n8+dTXl7e7PjLLruMmTNnNtm2ePFiysvLm2W/119/fbPpW6tXr6a8vLzZzZBuvfVWJk2a1GRbVVUV\n5eXlLFiwoMn2WbNmJY1t3LhxWRfH+PHjAZgyZYoXccRNmTLFizgAtm/f7kUc8fcj/FnL5jjipkyZ\n4kUcACeeeKIXccTfj/BnLZvimDVrVsO1sW/fvpSXlzNx4sRmx7RGxmfNBF0zVcD/U9U5oe33AN1U\n9fQdHJsL9AE+BL4D3KSq3YN9q4BfquotofJTgK+q6iEtnM9mzRhjjDEpyNpZM6paCywCjo9vExEJ\nXj+/k2PrVfWDYEzJmcDc0O6F4XMGvhxsN8YYY0wEdMp0BQK/Au4RkUXAS7hZNEUE3S0iciPQX1XP\nC17vC4wGXgR6AlcCnwPODZ1zBvAfEbkSeAI4Czco9tsdEI8xxhhjWiHjLSIAqvogbjGznwKvAiOB\nE1V1Y1CkLzAwdEgu8H3gNdzA1XzgKFVdHTrnQuBs4OKg3Ndw3TJL2jea6EjsZ8x2Fk90+RQLWDxR\n5lMs4F88qYhEIgKgqneo6hBV7ayqZar6SmjfeFX9Uuj1MlU9VFW7qGoPVf2aqv4vyTkfUdXhwTlH\nquq8joonChYvbnUXXVaweKLLp1jA4okyn2IB/+JJRcYHq0aJDVY1xhhjUpO1g1WNMcYYs/uyRMQY\nY4wxGWOJiDHGGGMyxhIRjyVbfTCbWTzR5VMsYPFEmU+xgH/xpMISEY9NmODXjYYtnujyKRaweKLM\np1jAv3hSYbNmQmzWjDHGGJMamzVjjDHGmKxjiYgxxhhjMsYSEY8l3nY621k80eVTLGDxRJlPsYB/\n8aTCEhGPzZo1K9NVSCuLJ7p8igUsnijzKRbwL55U2GDVEBusaowxxqTGBqsaY4wxJutYImKMMcaY\njLFExBhjjDEZY4mIx8aPH5/pKqSVxRNdPsUCFk+U+RQL+BdPKiwR8diYMWMyXYW0sniiy6dYwOKJ\nMp9iAf/iSYXNmgmxWTPGGGNMamzWjDHGGGOyjiUixhhjjMkYS0Q8tmDBgkxXIa0snujyKRaweKLM\np1jAv3hSYYmIx26++eZMVyGtLJ7o8ikWsHiizKdYwL94UmGDVUN8G6xaVVVFUVFRpquRNhZPdPkU\nC1g8UeZTLOBXPDZY1TTjy4c7zuKJLp9iAYsnynyKBfyLJxWWiBhjjDEmYywRMcYYY0zGWCLisUmT\nJmW6Cmll8USXT7GAxRNlPsUC/sWTCktEPDZo0KBMVyGtLJ7o8ikWsHiizKdYwL94UrFLs2ZEpFBV\nt6exPhnl26wZY4wxpqN02KwZEckRketEZB3wqYjsHWy/QUQubOv5jDHGGLP7SqVrZjJwPnAVUBPa\n/hZwURrqZIwxxpjdRCqJyLnAxar6V6A+tP11YHhaamXSYtmyZZmuQlpZPNHlUyxg8USZT7GAf/Gk\nIpVEZE/gvRbOlbdr1THpdNVVV2W6Cmll8USXT7GAxRNlPsUC/sWTilQSkSXAsUm2fx14NdWKiMhl\nIrJCRD4TkRdEZNROyp8jIq+JSKWIfCAiM0WkZ2j/eSISE5H64GtMRKpSrV82uu222zJdhbSyeKLL\np1jA4okyn2IB/+JJRacUjvkp8CcR2ROXyHxNRIbhumxOTaUSIjIO+CVwMfASMBGYJyL7qeqmJOWP\nBv4EfBd4HNdKcyfwe1xCFLcV2A+Q4PVudWMd36aFWTzR5VMsYPFEmU+xgH/xpKLNLSKq+hgwFjgB\nqMQlJiOAsar6rxTrMRG4U1XvVdVlwCVAFXBBC+WPBFao6u2qukpVn8clIqObV1c3qupHwWNjivUz\nxhhjTDtoUyIiIrki8nngTVX9sqruoapFqnqMqs5PpQIikgccBjwV36ZucZMngbIWDlsIDBSRk4Jz\n9AHOAJ5IKNdFRFaKyGoRmS0i+6dSR2OMMca0jzYlIqpaD8wHeqSxDqVALrAhYfsGoG8L9Xge+Cbw\ngIjUAB8CnwATQsXewbWolAPn4GJ9XkT6p7HukTZ9+vRMVyGtLJ7o8ikWsHiizKdYwL94UpHKYNW3\ngL3TXZG2CFo2ZgBTgEOBE4G9cN0zAKjqC6r6F1V9Q1WfA74GbAS+s7Pzn3zyyZSXlzd5lJWVMXv2\n7Cbl5s+fT3l5ebPjL7vsMmbOnNlk2+LFiykvL2fTpqZDXq6//vpmH8TVq1dTXl7ebFrXrbfe2uy+\nBFVVVZSXl7NgwYIm22fNmsWsWbOa1W3cuHFZF8f48eMbyvgQR7isD3EA3HvvvV7EEX8/wp+1bI4j\nfLwPcQC8+eabXsQRfz/Cn7VsimPWrFkN18a+fftSXl7OxIkTmx3TGm1e4l1EvgLcCFwHLMKNE2mg\nqtvaeL483HiQ/6eqc0Lb7wG6qerpSY65FyhU1W+Eth0NPAf0U9XE1pV4mQeBWlU9p4X9tsS7McYY\nk4IOW+Id+DtwEDAHWIvrEvkE2BJ8bRNVrcUlNMfHt4mIBK+fb+GwIqAuYVsMNytGmhd3S9MDB+K6\ncYwxxhgTAalM3/1i2msBvwLuEZFFNE7fLQLuARCRG4H+qnpeUH4u8HsRuQSYB/QHfg28qKrrg2Ou\nA17ALb7WHbck/SDgrnaovzHGGGNS0OZERFWfSXclVPVBESnFTQXuA7wGnBiabtsXGBgq/ycR6QJc\nBvwC1xrzFHBN6LQ9cOuK9MW11CwCyoLpwbuFTZs2UVpamulqpI3FE10+xQIWT5T5FAv4F08qUuma\nQUS6i8j3ReSu4DFRRLrtSkVU9Q5VHaKqnVW1TFVfCe0br6pfSih/u6oeqKpdVHWAqp6nqh+G9l+p\nqnsF5+uvqmNV9Y1dqWO2ueCClpZhyU4WT3T5FAtYPFHmUyzgXzypaHMiIiKHA8tx3Sc9g8eVwPJg\nsKeJiClTpmS6Cmll8USXT7GAxRNlPsUC/sWTilRmzTyHG3fxbVWtC7Z1wo292FtVP5/2WnYQmzVj\njDHGpCbVWTOpDFY9nFASAqCqdSJyM/BKy4cZY4wxxjSVyhiRbbjZJ4kGAhW7Vh1jjDHG7E5SSUQe\nAGaKyDgRGRg8zsR1zTRfytNkTOLqfNnO4okun2IBiyfKfIoF/IsnFakkIj8A/gbcC6wMHvcADwNX\np6leJg0WL251F11WsHiiy6dYwOKJMp9iAf/iSUWbB6s2HChSBAwNXi5X1aodlc8GNljVGGOMSU2H\nDVYN1gvJVdWPgTdD23sCdW2914wxxhhjdl+pdM3cD3wjyfZvBPuMMcYYY1ollUTkCODfSbb/J9hn\njDHGGNMqqSQiBUB+ku15QOddq45Jp/Ly8kxXIa0snujyKRaweKLMp1jAv3hSkUoi8hJwcZLtl+Bu\nLGciYsKECZmuQlpZPNHlUyxg8USZT7GAf/GkIpUl3o8GngRext3xFuB4YBQwRlWfS2sNO5DNmjHG\nGGNSk+qsmTa3iKjqf4EyYA1ugOpY3L1nRmZzEmKMMcaYjpfKvWZQ1deAc9JcF2OMMcbsZtrcIiIi\nh4rIgaHXXxWR2SLyMxFJNojVZMjs2bMzXYW0sniiy6dYwOKJMp9iAf/iSUUqg1XvBPYDEJG9cfee\nqQLOAG5OX9XMrpo1y69b/1g80eVTLGDxRJlPsYB/8aQilcGqW4FDVXW5iFwNfElVTwwGsd6vqgPb\no6IdwQarGmOMManpsMGqgISOOwH4e/B8DVCawvmMMcYYs5tKJRF5BZgsIt8CjgOeCLbvBWxIV8WM\nMcYY479UEpHvAYcCtwHTVPW9YPvXgefTVTFjjDHG+C+VdUTeUNUDVbWbqv4ktGsScF76qmZ21fjx\n4zNdhbSyeKLLp1jA4okyn2IB/+JJRUrriCSjqtvTdS6THmPGjMl0FdLK4okun2IBiyfKfIoF/Isn\nFW2eNeMzmzVjjDHGpKYjZ80YY4wxxqSFJSLGGGOMyRhLRDy2YMGCTFchrSye6PIpFrB4osynWMC/\neFKRyr1mHhGRSUm2XyUiD6WnWiYdbr7ZrxX3LZ7o8ikWsHiizKdYwL94UpHKEu8bgS+o6tsJ2w8E\nnlTVPmmsX4fybbBqVVUVRUVFma5G2lg80eVTLGDxRJlPsYBf8XTkYNUuQF2S7bVA1xTOZ9qJLx/u\nOIsnunyKBSyeKPMpFvAvnlSkkoi8CYxLsv1MYMmuVccYY4wxu5NUFjS7AfibiAwFng62HQ+cBZyR\nrooZY4wxxn+pLPE+FzgN2Ae4A/glMAA4QVVnp7d6ZldMmtRsTHFWs3iiy6dYwOKJMp9iAf/iSUVK\n03dV9QlVPVpVi1W1VFW/pKrP7EpFROQyEVkhIp+JyAsiMmon5c8RkddEpFJEPhCRmSLSM6HMGSKy\nNDjn6yJy0q7UMdsMGjQo01VIK4snunyKBSyeKPMpFvAvnlREYol3ERkH/Am4GHgJmIjr5tlPVTcl\nKX808AzwXeBxYE/gTuAdVf16UOaooMzVwBPAOcHzQ1Q16VgW32bNGGOMMR2lw2bNiEhMROpberT1\nfIGJwJ2qeq+qLgMuAaqAC1oofySwQlVvV9VVqvo8LhEZHSpzBfAPVf2Vqr6jqj8GFgMTUqyjMcYY\nY9IslcGqpye8zgMOAc4Drm/ryUQkDzgM+Fl8m6qqiDwJlLVw2EJgmoicpKr/EJE+uBaUJ0JlynDj\nV8LmAV9tax2NMcYY0z5SGaz6WMLjYVW9FrgKKE+hDqVALrAhYfsGoG8LdXge+CbwgIjUAB8Cn9C0\ntaNvW87po2XLlmW6Cmll8USXT7GAxRNlPsUC/sWTinTea+YF3DTedici+wMzgCnAocCJwF647hkT\nuOqqqzJdhbSyeKLLp1jA4okyn2IB/+JJRVoSERHpjBuTsS6FwzcB9UDi0vB9gPUtHHMN8N9g/Mdb\nqvov4FLggqCbhuDYtpyzwcknn0x5eXmTR1lZGbNnN52dPH/+fMrLmzcCXXbZZcycObPJtsWLF1Ne\nXs6mTU3H3l5//fVMnz69ybbVq1dTXl7eLFO+9dZbm031qqqqory8vNmNk2bNmkVhYWGzuo0bNy7r\n4hg/fjwAt912mxdxxN12221exAGgql7EEX8/wp+1bI4j7rbbbvMiDoBvfOMbXsQRfz/Cn7VsimPW\nrFkN18a+fftSXl7OxIkTmx3TGqnca+YTIHyQACW4waXfVNU5ba6EyAvAi6r63eC1AKuBW1T150nK\nPwzUqOrZoW1lwAJgT1VdLyL3A51V9auhMv8FXlfVS1uoh82aMcYYY1KQ6qyZVAarfi/hdQzYiEsk\nPknhfAC/Au4RkUU0Tt8tAu4BEJEbgf6qel5Qfi7wexG5BDcAtT/w66AO8RaPGcB/RORK3CDWs3CD\nYr+dYh2NMcYYk2ZtTkRU9U8t7RORA1T1rRTO+aCIlAI/xXWfvAacqKobgyJ9gYHhOohIF+Ay4BfA\nFuApXJdNvMxCETkbmBY8/gd8taU1RIwxxhjT8XZ5jIiIlIjIxSLyEvB6qudR1TtUdYiqdlbVMlV9\nJbRvvKp+KaH87ap6oKp2UdUBqnqeqn6YUOYRVR0enHOkqs5LtX7ZKLHvMNtZPNHlUyxg8USZT7GA\nf/GkIuVEREQ+LyJ/wk2d/QHuBnhHpqtiZtdVVVVlugppZfFEl0+xgMUTZT7FAv7Fk4o2DVYVkb7A\n+cCFQFfgQdwqqAf50OVhg1WNMcaY1LT7Eu8iMhd4BxiJG7DaX1Uvb2tFjTHGGGPi2jJY9STgFuC3\nqvq/dqqPMcYYY3YjbRkjcgxuvZBFIvKiiEwIZrqYiEpc/CbbWTzR5VMsYPFEmU+xgB/xVFRU8KMf\nXcH48aemdHyrExFVfUFVvw30wy2lfibwQXCOL4tISUo1MO3mggtaunlxdrJ4osunWMDiiTKfYoHs\nj6eiooJTTy2je/fbufTSD3d+QBJtXlm1ycEiw3ADV78FdAf+paqp3PguEnwbrLp48WIv4oizeKLL\np1jA4okyn2KB7I/nRz+6gu7db2f06Bjvvgvf+Q7QxsGqu5SINJxEJBcYC1xgiYgxxkSLquLunGFM\n+qgqRx89mGnT1iBCyolIKku8J6tMPTA7eBhjjMmwiooKbrzxWp55Zi4FBbVUV+dx3HFj+eEPp1FS\nYj3pZudUY9TUfMj27SvZvn1V8DX8fBUi1exqjpuWRMQYY0x0xPvtTzllKVOnxhABVXj55ds59dSn\nefzxhZaMGGKxOmpqPmgx0aiuXo1qbUP5Tp16UVg4mMLCIfTqdQqFhUOIxaah+tEuJSOWiHhs5syZ\nXHjhhZmuRtpYPNHlUyyQ/fHceOO1nHLKUkaPjgHwxBNwyikEr5dy002TmTZtRmYrmaJsf28S3XXX\nXVx00UXtcu5YrJbq6rVJEox40rEGqG8on5e3R0OiUVJyCIWFQygsHEJBwWAKCwfTqVPz5PWLX3yP\nl1++veGzlgpLRDy2ePFir35hLZ7o8ikWiH48qkpd3RZqaj6kpuZDqqs/bHheU/Mh//rXo9x8c+OF\n4X+hlZ9GjYpxzTV/5IIL8snL24O8vN7k5/cOPd+D3NyiDETVOosWLYr0e9Ma4W6zVas2cvfd01Lq\nNovFqtm+fU1DotGYYMQTjnVA4+cgP79fKNE4oiHRcNsGp/S+//CH0zj11KeBpXTrlloykpbBqr6w\nwarGGMjc4E7VGLW1m6mp+aBZcpH4Ohbb3uTY3NwS8vP7kZ/fjyuueIkbbvisxe9z3XV5/Pzng6it\n3Uh9/bZm+3NyihqSkry83k2eJ9vW3omLT+Ndwt1mo0aFu81yeOKJEU26zerrP6O6enWSsRnueU3N\nh0D8Gi4UFOwZtF4MaUg44s8LCgaRm1vYbjHddNNknnjiIV5//UPIxKwZX1giYszuqz0vdrFYHbW1\nHwUJxQctJhc1NetRrWtybKdOPcnP70dBQT/y8/uHnjc+Cgr6kZtb3HDM0UfvxdSpK5P226vC5MlD\n+O9/VwR1q6a2dhM1NR9RW7uR2tqNDc+TbUueuBQHrSquZaX586YtL7m5nVv9s2vLhTsbhKe7Jnrx\nRWHlyqFcdFFPtm9fRW3thtDeHAoKBjZLMBoTjYHk5OR3XCBJpHqvGeuaMcbssmyfHprq4M5YrJqa\nmvWhhCJ5S0Zt7Uc0/ucKIOTl7dGQUBQXj6RnzxObJRf5+X3JySloczzHHTe2xX77l1/O4QtfaFxl\nISenIPhPes9WndvFvDFIUD4KnscTF/e8qmoZNTXPBolLRbNzuMSlectKspaXn/3spibjXQBEdj7e\nRVVRrSUWq0G1mljMPVRrdvp8x+WqicVqWnhenfD9mj+fN287v/hF8p/t6NHKI4+s5bvfPTYYDDq4\nYYxGQcGe5OTkteo9yjaWiBhjUuJTc3ni4E5ovNipLmXy5FO54oojm7Vk1NV93OQ8Ip3Iz+/bkEx0\n7Wka0HEAACAASURBVHpkkuSiH3l5e5CT035/fsP99slbEaamfO6cnAIKCwdQWDigVeXr67c3a1VJ\nfF5ZuaQhsamv/7TJ8fPn0+KFe9SoGFdd9VvGjn08aTKwK0TyyMkpQCSfnJyCJM8LyMnJb3iem1sU\n2u72JT4XyadLl+sR2dLC94QuXXoxbNjMrE7s28oSEY+Vl5czZ86cTFcjbSye6EhsQZg8GaZOzcz0\nUFUlFqumvv5T6usrkn6tq0u+Pf51/vwX+fnPG5OQa6+FadPc89GjY0ya9CxnnbWmIaHo0WN40m6S\nvLxeiLTlFl7to6SkhMcfX8hNN03muuvmsGzZeoYP78txx5Xz+ONTOzRRzM0tJDd3IIWFA1tVvjFx\n+Yjq6o/o0uXsJhfu8HsjAp07F9C79xnk5BQ2SQyaPk+eGCRPMPKD/e3zPtbV/RrVLQ3dZuF4VKG6\nOm+3SkLAEhGvTZgwIdNVSCuLJzoSWxBOO611zeUAqvXU11e2OkloeX/j88RxFYlE8snNLSE3twud\nOrmvja97U1T0KiKN6yWcdlr4WOjadU+OOGJ5Vl0gSkpKgvdgBvPmzePEE0/MdJVaJZy4lJRAXV33\nJhfu8HujCvX1pQwdelNmKpuCxG6zcDyJ3Wa7C0tEPDZmzJhMVyGtLJ7oeOaZuUyd2tiCMGoUoecx\nrrnmLsaNezdpIhGLVe3k7BIkCo3JQvyrm1q6d7PtzZOM8L4uOx3EV1//HKqVDRe7cDw+/JeaLUlI\nMokX7vB7k40X7sRus1Gj0tdtlq0sETHG7FBd3adUVS2jqmoJlZVLqKx8G5E1La6kKAL5+fWIdKaw\ncI8WWyJaSiJycjp3ePdGWwZ3mo7VnuNdMiGx2yw/v5aamryMdJtFhSUixhgA6uq2UVW1lMrKJU2S\njurqVQ1lCgoGU1y8P7W1xahua3F6aCzWjwMP/FsH1n7X+Hax84mPF+5wt1m2zzhLB0tEPDZ79mxO\nC3dAZjmLJz1qaz9pSDTCX6ur1wYlhMLCvSgu/hx77HEmxcX7U1S0P0VFw+nUqQsAxx9/RZMWhAUL\n4Jhj3NHZ2IKQeLHbunUb3bp1zeqLXVi2/+6EL9yPPvoop59+eqarlDaPPfZYVr836WCJiMdmzZrl\n1Qfc4mmbmppNCQnH21RVLaGmZn1QIofOnfehuHh/+vQ5N5RwDNvpSpmJLQhPPQVHH53dLQjhi924\nceN44IEHMl2ltPHpd+f+++/3KhHx6b1Jla2sGmIrq5qOlI4mWVWlpmZD0haO2tqNgFvbonPnfSkq\n2p/i4s81JBydO++7S0s+x5d1fuaZps3l11yT/S0Ixpi2s5VVjfd86EtNdREwl3B80KR1I/68ru4T\nwE1RLSoaRlHR/vTo8aUg8difzp33aZeln62f2xiTDpaImEjzafXO1iwj3qVLMdXVa5q1blRWLmm4\nx0dOTiFFRcMpKtqfXr1OoqjItXIUFu7drqt17oglIcaYVFkiYiIr1ft/RNWOlxF/myuv3IdvfauS\nWKwScHdAjXejlJae1tDCUVg4BJHcTIVhjDFpZYmIx8aPH8/dd9+d6WqkLPHCPX06XH114+qd06Z9\nl+uvvxbVOmKxWlRrUa1r8tVtr0vYVxs6pum21u9r+/eaP//DJsuIx+MBGD0a/va3KoYMmUJxsRvL\nUVAwMBLLhbdGtn/WElk80eVTLOBfPKmwRMRj2bhyp2qM7dtXUln5Fk899Wduuqnxwn344Y3lRo2K\nMWnS3XzlK7vyC5yDSB4incjJyWt4Hv7qtse3Nd2fk5NHTk7hTo5xX6ETxcW/RqTxTqTheESguLgb\nAwd+Pyu7ObLxs7YjFk90+RQL+BdPKiwR8dhZZ52V6Sq0SFWprl5LZeVbwcDLt6iqcgMwY7EqVCEv\nT5osmHX88Y3P43epHDnygeAGVZ12kFSEt4cThI5tbairuxfVioaYwvFk+zLiUf6spcLiiS6fYgH/\n4kmFJSKmXTVOL327SdJRWfl2aPBlcTCt9ED22OOs4PkBqB6D6soWV++srS2hZ8/jm++MKFtG3Bhj\nmrNExKRNbe3mINGIJxsu4air2wyASAHFxSMoLj6AXr3KKS4+gOLiz1FYODhp64RvF25bRtwYY5qz\nRMRjzz33HMcee2zaz1tXty24D8lboS6VtxpW7HQLaA2juPhz9OhxQkMLR1unlyZeuN96Cw44IHsv\n3InLiFdVfUpRURcvlhFfsGAB/7+9O4+zuWwfOP65zmDGLJrJyJZtyBJaSEWUJWQZjza0vApPRSVZ\nSpaSRGjxiKge9Ch+isdTnpAlS4sSoVIZHlvIkiwxZgbDXL8/vmemc2Yzxsycpev9ep2XOfe5v/f3\nvmaOOdfc3/v+3k3T7/EeBCwe/xVMsUDwxZMffnNnVRF5HHgKKAf8ADyhqt/mUPdfwIOAAp4D9z+r\nan13nQeBf2Wqc0pVc7x3dTDcWdXzvhtbtx6kVq1y+b7vxrlzye5N0LwvqZw+vcddw0XJktXdIxv1\nMhKOkiWvKLAbaHnevXPLloPUrl0uaO7e2alTJz7++GNfd6NABFMsYPH4s2CKBYIrnvzeWdUvEhER\n6Qq8CzwCrAP6A3cDNVX1cDb1o4CSHkXFgE3A66r6orvOg8AEoCZ/JiKqqr/n0o+ATkQ877vRqFEa\np09DaKjnCEL2991ISztNcvJWr2QjKeknTp3aiZPHQVhYVSIi6rlvnuUkHeHhtQkJKZmlvcKSlJRE\nREREkZ2vsCUnJxMenvueLoEimGIBi8efBVMsEFzxBPot3vsDb6vqewAi0hvoAPQEXs5cWVUTgYx1\nkCLSGYgGZmStmnPiEWwy33cjzL2NSPp9N8aMGcqwYY97XU5xNkL7H3AOgBIlKhIRUZfY2L9ljHCE\nh1+ZseuqLwVTEgIEzS8fCK5YwOLxZ8EUCwRfPPnh80REnJssNAReSi9TVRWR5UDjPDbTE1iuqnsz\nlUeKyC+AC9gIDFXVzRffa//0+ecLGDUq68ROSL/vxhu0afMGAMWLlyEioh4xMa24/PInMxKO4sVj\nirLLxhhj/uJ8nogAsUAI8Fum8t+AWuc7WETKA+2Abple2oqToGwCLgGeBr4WkStVdf/FdtrfqCqh\noanZLnWFP2+YddVVHxIZWY8SJS4r2g4aY4wx2QiM+0fnrjtwDPivZ6GqfqOqs1R1k6p+CdwB/A70\nKvouFj4R4fTp4nhO+XnrrT+/VoWzZ2O49NKWAZuEPP30077uQoEKpniCKRawePxZMMUCwRdPfvhD\nInIYZ4JC2UzlZYGDeTi+B/Ceqp7NrZL79e+AGudrsH379nTq1Mnr0bhxY+bPn+9Vb9myZXTqlPVe\nFo8//jjTp0/3Ktu4cSOdOnXi8GHvubfPP/8848aN8yrbs2cPnTp1YsuWLV7lkyZNyvKmTU5OplOn\nTqxevdp93w3nR7piBWza9Ge99PtudO3a1e/j8PT+++/To0cPACpXrpxRHshxpKtcuXJQxAGwYsWK\noIgj/efh+V4L5DjSVa5cOSjiAGfSejDEkf7z8HyvBVIc77//fsZnY7ly5ejUqRP9+/fPckxe+Muq\nmW+Atar6pPu5AHuAiar6Si7HNQdWAPVUNeE853ABPwOLVPWpHOoE9KqZP/44RJs2lbnzztNcfz3Z\n3DArsHarNcYYEzgCfdXMeGCGiGzgz+W74bhXwYjIGKCCqj6Y6bi/4yQwWZIQEXkO+AbYjrOiZhBQ\nGZhWSDH43O+/v8gLLyirVt3Hc899RYkSqZw5UzwobphljDEmOPlFIqKqc0UkFhiJc0nme6Ctx9Lb\nckAlz2NEpBRwO9A3h2ZjgH+6jz0GbAAaq+qWHOoHtCNHFrFv3xvUrz+Jdu36AM4E1kDdRM0YY8xf\ngz/MEQFAVaeoalVVLamqjVV1vcdrPVS1Zab6J1Q1UlXfyaG9Aapazd1eBVWNV9VN2dUNdGfO/MaW\nLT249NIOVKz4eEb51q1bfdirgpf5mmegC6Z4gikWsHj8WTDFAsEXT374TSJi8kdV2bKlByDUrv2O\n1wjIoEGDfNexQmDx+K9gigUsHn8WTLFA8MWTH34xWdVfBOJk1V9/ncj27U9Sv/4nlC7dzuu1PXv2\nZJmRHcgsHv8VTLGAxePPgikWCK548jtZ1UZEAtjJkz+yY8cgKlbsmyUJAYLmzZ3O4vFfwRQLWDz+\nLJhigeCLJz8sEQlQ586lkJBwL+HhVxAXN+78BxhjjDF+yC9WzZgLt3PnMyQnb6Nhw/WEhIT5ujvG\nGGNMvtiISAA6cuQT9u2bRPXqrxAZWS/HepnvuBfoLB7/FUyxgMXjz4IpFgi+ePLDEpEA8+dS3fZU\nrNgn17rJyclF1KuiYfH4r2CKBSwefxZMsUDwxZMftmrGg7+vmlFVfvyxI4mJ62nUaBMlSmTenscY\nY4zxjUC/xbvJg3373uDo0U+oX3+RJSHGGGOCgl2aCRAnT/7Ejh1PU7HiE5Qu3d7X3THGGGMKhCUi\nAeDcuVMkJNxzwUt1M28ZHegsHv8VTLGAxePPgikWCL548sMSkQCQvlS3Tp3ZhISUzPNxPXv2LMRe\nFT2Lx38FUyxg8fizYIoFgi+e/LA5In7uyJHF7Ns3kRo1Xicysv4FHTtixIjC6ZSPWDz+K5hiAYvH\nnwVTLBB88eSHrZrx4G+rZs6cOcS3315FVNS11K//ideGdsYYY4w/sb1mgsyfu+qmUbv2DEtCjDHG\nBCW7NOOn9u+f4l6qu9CW6hpjjAlaNiLih5KSfmbHjqeoWLEPpUt3yHc706dPL8Be+Z7F47+CKRaw\nePxZMMUCwRdPftiIiJ85d+4UmzffQ1hYdeLiXr6otjZu3Mjf//73AuqZ71k8/iunWPbs2ROQyxOX\nLl3Ktdde6+tuFJhgiieYYoHAiic2NpbKlSsXeLs2WdWDP0xW3b69P/v2vUnDhuuIjLzKJ30wpiDs\n2bOHOnXq2F4axgSJ8PBwEhISckxG7BbvQeDIkSX8+usEatSYYEmICXiHDx8mOTmZWbNmUadOHV93\nxxhzERISErj//vs5fPhwgY+KWCLiJ86cOcSWLd2JiWlLxYpP+Lo7xhSYOnXq+MVyeGOMf7LJqn5A\nVdm69e/8uVTXfizGGGP+GuwTzw/s3/8mR44spFatdwgNLVdg7Xbq1KnA2vIHFo//CqZYjDFFyxIR\nH3OW6g6kQoXHiI3tWKBt9+nTp0Db8zWLx38FUyzGmKJliYgPOUt17yUsLI7q1V8t8PbbtGlT4G36\nksXjv4IpFmNM0bJExId27RpKcvIWrrzy/QvaVdcY89eydetWXC4Xc+fOveBjT58+jcvl4uWXL+6+\nRMYUFktEfOTo0WX8+us/iIsbZ0t1jQkwLpfrvI+QkBC++OKLAjvnxew3JSK2X5XxW7Z81wfOnPmd\nLVseJCamDZdf3rfQzjN//nw6d+5caO0XNYvHfxVULKpaaB+YBdn2rFmzvJ6/++67LF++nFmzZuF5\nk8iCun9KrVq1SElJoUSJEhd8bGhoKCkpKRQvXrxA+mJMQbNEpIilL9VVPVvoS3Xff//9oPmgA4vH\nn11MLImJiQx7cRgLli8gNSSV4ueKE39rPKOfG01UVNRF9auw2r733nu9nq9Zs4bly5dzzz335On4\nU6dOERYWdkHnzE8SUhDHBrr8fK9N0bJLM0Vs//63OHJkgXupbvlCPdecOXMKtf2iZvH4r/zGkpiY\nSOM2jZl8YDK/dPqFfR338UunX5h8cDKN2zQmMTEx330qzLYvxNKlS3G5XHz00Uc888wzVKxYkcjI\nSM6cOcPhw4fp378/9erVIzIykujoaOLj49m8ebNXG9nNEenWrRtlypRh7969dOzYkaioKMqWLcuw\nYcO8js1ujsjgwYNxuVzs3buX+++/n+joaC699FJ69erFmTNnvI5PTk7mscceo3Tp0pQqVYq77rqL\n3bt353neyfjx47nyyiuJiIjg0ksv5YYbbuDDDz/0qrN37166d+9O+fLlKVmyJDVq1OCJJ57wGl3a\nvn07d9xxBzExMURERHDTTTfx6aef5vl7DXD06FH69OlDpUqVCA0NpWbNmowfP/68MZjCZSMiRSgp\naTM7dgygQoVHiY2N93V3jPG5YS8OI6FGAmk10v4sFEirnkaCJvDsqGd5fdzrftd2fjz33HNERETw\nzDPPkJSUREhICFu3bmXJkiXcddddVKlShQMHDvDWW2/RvHlzNm/eTGxsbI7tiQipqam0bt2a5s2b\n8+qrr7JkyRLGjh1LzZo1efDBB3M9VkTo3LkzNWvWZNy4caxbt45p06ZRoUIFnn/++Yy699xzDwsX\nLqRnz540bNiQ5cuX07lz5zxd5po0aRJPPfUU9913HwMGDCAlJYXvv/+etWvXcscddwBOEtKoUSNS\nUlLo1asXNWvWZM+ePcydO5fU1FRKlCjBvn37aNy4MWlpafTr149LLrmEd955h/bt27NgwQJuu+22\n836vT548SdOmTTl69Ci9e/emYsWKfPHFFzz11FMcPnyYl1566bzxmMJhiUgRSUs77V6qW61Qluoa\nE4gWLF9AWqe0bF9Lq57GvPnzeLBfzh+ouZm3dB5pt+fc9scLPuZ1ii4RUVW++uorihX789duo0aN\nSEhI8Kp3zz33ULduXd59910GDhyYa5uJiYkMHz6cAQMGANCrVy/q1avH9OnTc01E0vtz0003MXHi\nxIxjDx48yPTp0zMSkTVr1rBgwQKGDh3KqFGjAOjduzf33nsvmzZtOm/Mn3zyCddddx0zZ87Msc5T\nTz3FsWPH2LhxI3Xr1s0oHzlyZMbXo0aN4tixY6xbty5ju4CePXtSt25dBgwYkCURye57PXbsWA4c\nOMCmTZuoVKkSAA8//DBlypRh/Pjx9O/fnzJlypw3JlPwLBEpIjt3DiU5OYEGDdYSEhLu6+4Y43Oq\nSmpIKuT0h7XA/lP7afh2w5zr5Ng4cJpc2051pRbq5NjMevbs6fXBCN5zN86dO8fx48eJjo6mWrVq\nbNyYt81LH3nkEa/nTZs2ZeHChec9TkTo1auXV1mzZs1YunQpqampFC9enCVLliAiPProo171nnji\nCT744IPzniM6OpoNGzbwww8/cPXVV2d5/ezZsyxcuJC77rrLKwnJbPHixTRr1sxrz6JSpUrx0EMP\nMXLkSHbu3ElcXFzGa9l9r+fNm0fLli0JDw/nyJEjGeW33norEyZMYPXq1dx+++3njckUPEtEioCz\nVHc81au/RlTUNUV23h49evCvf/2ryM5X2Cwe/5WfWESE4ueKO0lDdrmAQvnQ8izsdf4P1ex0/Kgj\nB/RAjm0XP1e8SJe0Vq1aNUtZWloar776Km+//Ta7d+8mLc0ZwRERatSocd42o6OjiYyM9CqLiYnh\n2LFjeepT5l1UY2JiUFX++OMPypQpw+7duwkNDaVixYpe9fLSN4ChQ4fyxRdfcO2111KzZk3atm3L\nfffdx/XXXw/A/v37SUlJyTUJUVX27t2bZdQD/lyVtHv3bq9EJLvv9fbt29m2bRsfffRRltdEhEOH\nDuUpJlPw/CYREZHHgaeAcsAPwBOq+m0Odf8FPEjWX2E/q2p9j3p3AyOBqsD/gMGqurhQAsiB91Ld\nfkV56qC726XF47/yG0v8rfFM3jmZtOpZL6G4dri4+7a7aVA+fzv33tX2rlzb7tS6aPfHKVky600L\nhw8fzksvvUTv3r1p0aIFMTExuFwuHn300YykJDchISHZlntO8izM48+nfv36/O9//2PhwoUsWbKE\nuXPnMmnSJMaMGcMzzzxTIOfITubvdXo8HTp0oF+/7H8P165du9D6Y3LnF4mIiHQFXgMeAdYB/YGl\nIlJTVQ9nc0hfwPNdXAzYBGRMKReRJsBsd71FwH3AfBG5VlW9p6QXEmep7kOkpaX6ZFfdvC4lDBQW\nj//KbyyjnxvNyjYrSdAEJ2EQQJ1Eoc72OoyaMirffSrMtgvKf/7zH9q3b8+UKVO8yo8ePUr16tV9\n1Ks/ValShdOnT7Nv3z6vUZFt27bluY2IiAi6du1K165dSU1NpUOHDrzwwgsMGjSIChUqULJkSX76\n6accjxcRKlWqxNatW7O8lj6/pkqVKrn2QUSoWrUqycnJtGzZMs99N0XDX5bv9gfeVtX3VHUL0BtI\nBnpmV1lVE1X1UPoDuB6IBmZ4VOsLLFbV8aq6VVWHAxuBItuda//+tzly5GNq1y78pbrGBKKoqCjW\nLFtDnwp9qLqgKhUXVqTqgqr0qdCHNcvWXNS9Pgqz7QuV0yWgkJCQLKMPM2fO9JrD4Ett27ZFVbMk\nSpMmTcrTZa2jR496PS9evDi1a9fm3LlzpKamUqxYMeLj4/nPf/6TazLSvn17vvzyS77//vuMshMn\nTjBt2jRq167tdVkmp3516dKFzz77LNu73R47dqzARoHMhfP5iIiIFAcaAhlrp1RVRWQ50DiPzfQE\nlqvqXo+yxjijLJ6WAn+7iO7mWVJSgnupbm9iY22LdGNyEhUVxevjXud1Xi/wyaOF2faFyOlDrmPH\njrzyyis88sgjNGrUiB9++IE5c+ZkO8fBF5o0aUKHDh0YO3YsBw8e5LrrrmPFihXs2rULOP9t52+5\n5RaqV6/OjTfeyGWXXcaPP/7I22+/zR133JExUXfcuHF89tlnNGnShF69elGrVi1+/fVX5s6dy/ff\nf0+JEiUYNmwY8+bNo1WrVvTt25dSpUrxzjvvcPDgQaZNm+Z1zpy+10OHDmXRokW0bt2anj17cs01\n15CYmMimTZv48MMPOXToEOHhtpDAF/xhRCQWCAF+y1T+G858kVyJSHmgHTA100vl8tvmxUpLO01C\nwr2EhVWlevXMuVDRWb16tc/OXRgsHv9VULEUZqJQ2ElIbu3n9NqIESPo27cvixYtYsCAAWzevJll\ny5ZRrly5LMdk10ZO7WZ3bF7ay86cOXPo1asX8+fPZ/DgwYgIM2fORFXPe8fSRx99lKNHjzJ+/Hj6\n9OnDJ598wqBBg3jnnXcy6lSpUoW1a9fSuXNn3nvvPfr27cvs2bO57bbbMm5LX7FiRb7++muaN2/O\nhAkTePbZZ7nkkktYvHgxbdu2zVNckZGRfPXVV/Tv35/ly5fTr18/Xn31VXbv3s2YMWOyncNjioiq\n+vQBlAfSgBsylY8D1uTh+CHAIaBYpvLTQNdMZY8CB3JpqwGgZcuW1fj4eK/HjTfeqB999JF6Wrp0\nqcbHx2tm9913tT79dIieOPFdRtmGDRs0Pj5ef//9d6+6w4cP17Fjx3qV7d69W+Pj4zUhIcGrfOLE\nifrUU095lSUlJWl8fLx++eWXXuWzZ8/WSpUqZelbly5d8hzHY489ptOmTfMq80Uc3bt3V1X16mMg\nx5EuPj4+KOJQVa1QoUKWON544w0FdMOGDVnqm8C2Zs0aFRH98MMPfd0VU0Q2bNiggL766quq6vwu\nSP9sTP/MvPnmmxVnEUkDvZA84EIqF8YDKA6kAp0ylc8APsrD8f8DXs2mfDfQN1PZCOC7XNpqcLG/\nOI8cWaarVqF79rya7zYKSlJSkq+7UKAsHv+VXSzpv7gsEQlsKSkpWcq6deumJUqU0EOHDvmgR8YX\n8vL/Ob3OhSYiPp8joqqpIrIBaAV8DCDO2ForYGJux4pIc6A6MD2bl9dk00Zrd3mhOHPmsHupbmsu\nv7x/YZ0mz4LteqfF47+CKRbj7cUXX2TLli3cfPPNiAgLFy5kxYoVPPnkk3YnUlMgfJ6IuI0HZrgT\nkvTlu+G4V8GIyBiggqpmvmfx34G1qppAVq8Dn4nIAJzlu/fgTIp9uDACUE1fqnvGJ0t1jTGmMDRt\n2pTPPvuMkSNHkpSURJUqVRg9enSh3gfE/LX4RSKiqnNFJBbn5mNlge+Btqr6u7tKOaCS5zEiUgq4\nHWeZbnZtrhGRe4HR7sc24G9aSPcQOXDgnxw58l/q1ZtPaGiFwjiFMcYUuXbt2tGuXTtfd8MEMb/5\ns11Vp6hqVVUtqaqNVXW9x2s9VLVlpvonVDVSVd/J2lpGnf+oam13m1ep6tLC6HtS0ha2b+9P+fK9\niI0tktXBefL000/7ugsFyuLxX8EUizGmaPlNIhKo0pfqhoZWpkaN8b7ujpfM+0gEOovHfwVTLMaY\nouUXl2YC2a5dz5KU9JNf7qr7xBNP+LoLBcri8V/BFIsxpmhZInIRjh5dzt69rxIX9wpRUdf6ujvG\nGGNMwLFLM/mUmnqELVseIDq6FZUqDfB1d4wxxpiAZIlIPvy5VPc0deq867dLdbds2eLrLhQoi8d/\nBVMsxpii5Z+foH7uwIGpHD48n1q1phMaWvH8B/jIoEGDfN2FAmXx+K9gisUYU7QsEblAzlLdfpQv\n/whlynT2dXdy9cYbb/i6CwXK4vFfwRSLKRgjRozA5fL+iKlatSo9e/Y877EzZszA5XKxZ8+eAuvP\n7t27cblcvPfeewXWpikYlohcgLS0MyQk3OeXS3WzE2xLKi0e/xVMsVyonTt30qtXL6pXr07JkiW5\n5JJLaNq0KRMnTuTUqVO+7p7PZLfjr8vlytOuv9kdm1fvv/8+r7/+eo7tGv9jq2YugLNU90caNPiG\nkJAIX3fHGONjixYtokuXLoSFhfHAAw9Qr149zpw5w+rVqxk0aBCbN2/mrbfe8nU3/cbWrVuzjJIU\ntNmzZ/Pzzz/z5JNPepVXqVKFlJQUihcvXqjnNxfOEpE8OnZsBXv3vkJc3MtERTXwdXeMCUqqWmh/\ntRZ027/88gv33HMP1apVY+XKlVx22WUZrz366KO8+OKLLFq0KNf+nDlzhtDQ0ALrk7/zdRJQokQJ\nn57fl1JSUihZsqSvu5EtuzSTB6mpR0hIeIDo6JZUqjTQ193Js3Hjxvm6CwXK4vFfFxNLYmIiQ4f2\n5aabqtGqVSVuuqkaQ4f2JTEx8aL7VZhtjxs3jqSkJKZPn+6VhKSLi4vzutGby+Wib9++zJ49E4dx\nkwAAIABJREFUm3r16hEWFsbSpc6uE8nJyQwcOJDKlSsTFhZG7dq1ee2117K0+emnn9KsWTNiYmKI\nioqidu3aDBs2zKvOpEmTqFevHhEREVx66aU0atSIDz74INdYXnvtNVwuF3v37s3y2pAhQwgNDeX4\n8eMArF69mi5dulClShXCwsKoXLkyAwYMyNNlqOzmiGzevJmWLVsSHh5OpUqVGD16NGlpaVmO/fjj\nj+nYsSMVK1YkLCyMGjVqMGrUKK+6LVq0YNGiRRnzQVwuF3FxcUDOc0RWrlxJs2bNiIyMJCYmhs6d\nO2dZBZY+32XHjh10796dmJgYoqOj6dmzZ57i3r59O3feeSfly5enZMmSVKpUiXvuuSfL+3DWrFnc\ncMMNGT+7W265heXLl3vVmTJlSsb7p2LFivTp0yfjZ5OuefPmXHXVVWzcuJGbb76ZiIgIr/fJ4sWL\nufnmm4mMjKRUqVJ07NiRzZsLZRu2PLERkfNwluo+TFraKerUec9vl+pmJzk52dddKFAWj//KbyyJ\niYl07NiYDh0SGDUqDRFQhW+/nUzHjitZuHANUVFRftc2wMKFC4mLi+OGG27I8zErVqxg7ty59OnT\nh9jYWKpWrQpAfHw8n3/+OQ899BBXX301S5cu5emnn2b//v0ZCcnmzZuJj4/nmmuu4cUXXyQ0NJTt\n27fz9ddfZ7Q/depUnnzySbp06UK/fv04deoUmzZtYu3atXTr1i3HfnXp0oVBgwYxd+5cBg70/mPr\n3//+N7fddhuXXHJJxvOUlBQee+wxSpcuzbp165g0aRL79u1jzpw5ucafeUTqt99+o3nz5qSlpTF0\n6FDCw8P55z//SVhYWJZjZ8yYQVRUFAMHDiQyMpKVK1cyfPhwEhMTMxLhZ599luPHj7Nv3z4mTJiA\nqhIZGZljf5YvX0779u2pXr06L7zwAikpKUycOJGmTZuycePGjLlP6f3u0qULcXFxjB07lo0bNzJt\n2jTKli3LmDFjcjxHamoqbdq0ITU1lb59+1KuXDn27dvHwoUL+eOPPzLegy+88AIvvPACN910Ey++\n+CIlSpRg7dq1rFy5kltvvRVwEqKRI0fSpk0bHnvsMbZu3cqUKVNYv349X331FSEhIRn9PXz4MO3b\nt6dbt2488MADlC1bFoCZM2fSvXt3brvtNl5++WWSk5N58803adasGd99951v5nupqj3cD6ABoBs2\nbNB0+/ZN1VWr0EOHPlRjTN5t2LBBM/9/ymzIkCd03DiXrlpFlse4cS4dOrRvvs9fmG2fOHFCRURv\nv/32PB8jIlqsWDHdsmWLV/n8+fNVRHTMmDFe5XfffbeGhITozp07VVV1woQJ6nK59OjRozmeo3Pn\nzlq/fv0LiORPTZo00UaNGnmVrVu3TkVE/+///i+j7NSpU1mOHTt2rIaEhOjevXszykaMGKEul8ur\nXtWqVbVHjx4Zz/v166cul0vXr1+fUXb48GGNjo5Wl8ulu3fvzvW8vXv31sjISD1z5kxGWceOHbVa\ntWpZ6v7yyy8qIvruu+9mlF1zzTVarlw5/eOPPzLKNm3apCEhIdq9e3evWEREH374Ya8277jjDi1T\npkyWc3n6/vvvVUT0ww9z/gzZvn27hoSE6F133ZVjnd9//11DQ0O1Xbt2XuWTJ09Wl8ulM2bMyChr\n3ry5ulwunTp1qlfdkydPakxMjPbu3dur/NChQxodHa29evXK8fx5+f+cXgdooBfw2Rs4f977QHLy\nVrZvf5Ly5R+mTJnbfd0dY4LO558voFGjrMPwAI0apbFq1TwSEzfm67Fq1bxc2/7884/z3e8TJ04A\nXPCISvPmzalVq5ZX2eLFiylWrFiW/XoGDhxIWloaixcvBiA6OhqAjz76KP0Ppyyio6P59ddfWb9+\nfbav56Zr165s2LCBXbt2ZZTNmTOHsLAwOnXqlFHmOaclOTmZI0eO0LhxY9LS0vjuu+8u6JyLFy/m\nxhtvpGHDhhllpUuX5r777stS1/O8J0+e5MiRIzRt2pTk5OR83VDv4MGD/PDDD/To0SNjtAegfv36\ntG7dmk8++cSrvojQq1cvr7JmzZpx5MgRTp48meN50ttesmQJKSkp2dZJ/5kOHz48x3aWL19Oamoq\n/fr18yp/+OGHiYqKyjIfKTQ0lO7du3uVffrppxw/fpxu3bpx5MiRjIeIcMMNN7Bq1aocz1+Y7NJM\nDtLSzrB5832Ehl5OjRr/8HV3jAk6qkpoaCo5zR8VAZH9rF/fMMc6ObcNLhe5tl2iRGq+J7CWKlUK\n4ILnmqRfivG0e/duKlSoQESE90q8OnXqZLwOTqIwffp0Hn74YQYPHkyrVq244447uOuuuzJieOaZ\nZ1ixYgXXX389NWrUoE2bNtx77700adIEcC4THD161Os8ZcqUweVycffddzNgwADmzJnD4MGDAZg3\nbx7t27f3uryxd+9ennvuORYsWMCxY8cyykUky1yF89m9ezc33nhjlvLMyRo4l6aGDRvGqlWrMhLB\n/J43/dwANWvWzPJanTp1WLZsWZYJnpkvW8TExABw7NixHC8BVa1alYEDBzJ+/HhmzZpFs2bN6NSp\nE/fff3/G+2jnzp24XK6Mn/mF9Ld48eLExcVlvJ6uYsWKFCvm/RG/bds2VJUWLVpkaV9EvBKyomSJ\nSA527XqOpKRNNGiwJmCX6h4+fJjY2Fhfd6PAWDz+Kz+xiAinTxdHNfuEQRXS0spz3XUL89WntLSO\nqB7Ise3Tp4vnexVNVFQUFSpU4Keffrqg4y5m1UJYWBhffPEFq1atYtGiRSxZsoQ5c+bQqlUrli1b\nhohQu3Zttm7dysKFC1myZAkffvghU6ZM4fnnn+f555/n66+/pkWLFohIRhK2a9cuKleuTPny5WnW\nrBlz585l8ODBrFmzhj179vDKK69k9CEtLY1bb72VP/74gyFDhlCrVi0iIiLYt28fDz74YLaTTAvC\n8ePHufnmm4mOjmbUqFHExcURFhbGhg0bGDx4cKGdN7P0ORiZ5TRCle6VV16he/fu/Pe//2XZsmX0\n7duXMWPGsHbtWipUqFAYXc32vZaWloaIMGvWrIw5I54yJy5FxRKRbJw4sY4TJ14hLm4sUVENz3+A\nn+rZsycff5z/4Wd/Y/H4r/zGcsst8Xz77WSuvz7rB8m337po2fLufC+Xb9Hirlzbbt68UzZH5V3H\njh2ZOnUqa9euvaAJq5lVqVKFFStWkJSU5DUqkpCQkPG6pxYtWtCiRQteffVVxowZw7PPPsuqVato\n2bIl4HwA3X333dx9992cPXuW22+/ndGjRzNkyBCuvvrqLKswypUrl/F1165defzxx9m2bRtz5swh\nIiKCjh07Zrz+448/sm3bNmbOnOl1+SRzmxcS+7Zt27KUZ77U8tlnn3Hs2DH++9//ctNNN2WU79ix\nI8uxeU0u07+vW7duzfb8sbGxBbrctW7dutStW5ehQ4fyzTff0KRJE9566y1GjhxJ9erVSUtLY/Pm\nzVx11VXn7a/nyFpqaiq7du2idevW5+1D9erVUVXKlCmT8X7xBzZHJBuPPfY4s2dXIDq61/kr+7ER\nI0b4ugsFyuLxX/mNZciQ0SxaVId161yk/1GpCuvWuVi0qA6DB4/Kd58Ks21w9tcJDw/noYce4tCh\nQ1le37FjBxMnTjxvO+3bt+fs2bNZbpP/j3/8A5fLRbt27QC8LoOku/rqq1FVTp8+DZDlskuxYsWo\nU6cOqkpqairR0dG0bNnS6+F5b40777wTl8vF7NmzmTdvHh07dvT6ME4fEcg8AjFhwoR8jS61b9+e\nb775xmtOy++//87s2bO96oWEhKCqXuc9c+YMU6ZMydJmREREni7VlCtXjmuuuYZ3333X61LPTz/9\nxLJly+jQocMFx5OdxMREzp0751VWt25dXC5Xxs+tc+fOiAgjR47McXTl1ltvpXjx4lneU9OmTePE\niRNeCWNO2rZtS6lSpXjppZc4e/ZsltcPHz6c17AKlI2IZOPJJ9M4fvwA8fE3XfQSP19q0CC4brxm\n8fiv/MYSFRXFwoVrGDv2WZ577mNKlEjlzJni3HJLJxYuHHVR//cKs21w7hMye/ZsunXrRp06dbzu\nrPrVV18xb948evTocd524uPjadGiBcOGDWPXrl0Zy3cXLFhA//79qVatGgAjR47kiy++oEOHDlSp\nUoXffvuNN998k8qVK9O0aVMA2rRpQ7ly5bjpppsoW7YsmzdvZvLkyXTs2DHLHJTslClThhYtWjB+\n/HhOnjxJ165dvV6vXbs21atXZ+DAgfz666+UKlWK//znP/zxxx/5+A46ydzMmTNp27YtTz75JOHh\n4UydOpWqVauyadOmjHpNmjQhJiaGBx54gL59+wLOPTeyS34aNmyYsQy5UaNGREZG5vgh/corr9C+\nfXtuvPFG/v73v5OcnMwbb7xBTEwMzz//fL5iymzlypX06dOHu+++m5o1a3L27Fnee+89ihUrxp13\n3gk4IxXDhg1j1KhRNGvWjDvuuIPQ0FC+/fZbKlasyOjRo4mNjWXIkCGMHDmS2267jU6dOrFlyxbe\nfPNNrr/++mwn+GYWFRXFm2++yQMPPECDBg3o1q0bZcqUYc+ePSxatChja4IidyFLbIL9gXv57ttv\nF8wSP2P+yvKy3C+ztLS0QutPYbW9fft27dWrl8bFxWlYWJiWKlVKmzRpopMmTdLTp09n1HO5XNq3\nb/a/T5KSknTgwIF6+eWXa2hoqNaqVUvHjx/vVWfVqlV6++236+WXX65hYWF6+eWX6/3336/bt2/P\nqDN16lRt3ry5lilTRkuWLKlXXHGFDh48WBMTE/Mcz7Rp09Tlcml0dLRX/9Nt2bJF27Rpo6VKldLL\nLrtMe/furT/++KO6XC6vpbEjRozQkJAQr2OrVaumPXv29Cr76aeftEWLFhoeHq6VKlXSl156Sd95\n550sy3fXrFmjTZo00YiICL388st1yJAh+umnn6rL5dLPP//c63t5//3366WXXqoulytjKe8vv/yS\npY+qqitXrtRmzZppRESERkdHa+fOnbMssU5finzkyBGv8hkzZmTpZ2a7du3Shx56SK+44goNDw/X\n2NhYbdWqla5atSpL3RkzZmjDhg21ZMmSWrp0aW3RooWuWLHCq86UKVP0yiuv1NDQUC1fvrz26dNH\njx8/7lWnefPmetVVV+XYp88//1zbtWunMTExGh4erldccYX27NlTN27cmOMxhbl8V/Q8k2z+SkSk\nAbDh7behZk1nGPe556qyevWu8x5rjPG2ceNGGjZsyIYNG4Jq9MeYv6K8/H9OrwM0VNWNeW3b5ojk\nwnOJXyCaPn26r7tQoCwe/xVMsRhjipYlIrm42CV+vrZxY54T0oBg8fivYIrFGFO0LBHJRUEs8fOl\nyZMn+7oLBcri8V/BFIsxpmjZqplseC7xW7jw4pb4GWOMMSZnNiKSjTffLM/x430CeumuMcYYEwhs\nRCQb77yz0Gb5G2OMMUXARkSCmOeOmcHA4vFfwRSLMaZoWSISxPr06ePrLhQoi8d/BVMsxpiiZZdm\nglibNm183YUCZfH4r9xiSd+8zRgTuArz/7ElIsaYQhEbG0t4eDj333+/r7tijCkA4eHhxMbGFni7\nlogYYwpF5cqVSUhI8NmOnsaYghUbG0vlypULvF1LRILY/Pnz6dy5s6+7UWAsHv+VUyyVK1culF9c\nhS2YfjYQXPEEUywQfPHkh99MVhWRx0Vkl4ikiMg3ItLoPPVLiMhoEflFRE6JyE4R6e7x+oMikiYi\n59z/polIcqEH4kfGjRvn6y4UKIvHfwVTLGDx+LNgigWCL5788IsRERHpCrwGPAKsA/oDS0Wkpqrm\nNK77b6AM0APYAZQna2J1HKgJpG8WE5i71+VTmTJlfN2FAmXx+K9gigUsHn8WTLFA8MWTH36RiOAk\nHm+r6nsAItIb6AD0BF7OXFlEbgOaAXGq+oe7eE827aqq/l44XTbGGGPMxfL5pRkRKQ40BFakl6mq\nAsuBxjkcFg+sB54RkV9FZKuIvCIiYZnqRbov3ewRkfkicmVhxGCMMcaY/PGHEZFYIAT4LVP5b0Ct\nHI6JwxkROQV0drfxJnAp8Hd3na04IyqbgEuAp4GvReRKVd1fkAEYY4wxJn/8IRHJDxeQBtyrqicB\nRGQA8G8ReUxVT6vqN8A36QeIyBogAegFPJ9Du2EArVu3pl69el4vHD16lO7du9OiRYuMsjVr1jB3\n7lz+8Y9/eNUdO3YstWvX9poJnZCQwD//+U+GDx9OTExMRvlbb71FWFgY3bt3zyg7cOAAL7/8Mn37\n9qVatWoZ5R988AEHDx6kX79+GWUpKSkMHTqUBx54gGuvvTajfMmSJXz66ads3LjRq2+DBw+mbdu2\nARXHN998w4gRI1i3bl1GPIEcR7p169bRpk2bgI8DYNmyZbz22msBH0f6+8rzvRbIcaT/PNatW8fk\nyZMDPg6AL7/8kltuuSXg40h/X23dutXr93SgxLFkyRKWLl3K0aNH2b9/P/Xq1SMxMTG9auarE7kS\n5yqI77gvzSQDd6rqxx7lM4BLVPX2bI6ZATRR1ZoeZbWBn4Gaqrojh3PNBVJV9b4cXr8X+L/8R2OM\nMcb85d2nqrPzWtnnIyKqmioiG4BWwMcAIiLu5xNzOOwr4C4RCVfV9CW5tXBGSX7N7gARcQH1gUW5\ndGcpcB/wC85lH2OMMcbkTRhQFeezNM98PiICICJdgBlAb/5cvnsXUFtVfxeRMUAFVX3QXT8C2Ixz\n6WUEzjLeqcAqVe3trvOc+/XtQDQwCOgENFTVLUUWnDHGGGNy5PMREQBVnSsiscBIoCzwPdDWY+lt\nOaCSR/0kEWkNTAK+BY4Ac4DnPJqNAf7pPvYYsAFobEmIMcYY4z/8YkTEGGOMMX9NPr+PiDHGGGP+\nuiwRMcYYY4zPWCLidqGb7vkrEWkmIh+LyD73Rn+dfN2n/BKRISKyTkROiMhvIvKRiNQ8/5H+SUR6\ni8gPInLc/fjavV1BwBORwe7323hf9yU/ROR5j80x0x+bfd2viyEiFURkpogcFpFk93uvga/7lR/u\n382Zfz5pIjLJ1327UCLiEpEX3Ru1JovIdhF51tf9uhgiEikiE9x3Mk8WkdUicl1ej7dEBK9N954H\nrgV+wNl0L9anHcufCJzJvo8R+Jv8NcOZkHwDcCtQHFgmIiV92qv82ws8AzTA2dZgJfBfEanj015d\nJHfS/gjO/5tA9hPOZPly7kdT33Yn/0QkGuc2B6eBtkAdYCDOxP1AdB1//lzKAa1xfr/N9WWn8mkw\nzo01HwNq46zoHCQifXzaq4szHeeWG/cB9YBPgeUiUj4vB9tkVUBEvgHWquqT7ueC86ExUVWzbLoX\nKEQkDejseaO4QOZODA8BN6vqal/3pyCIyBHgKVX9l6/7kh8iEomzIu1RnFVr36nqAN/26sKJyPPA\n31Q1IEcMMhORsTirBG/xdV8Kg4hMANp73tQyUIjIAuCgqj7sUTYPSFbVB3zXs/xx7/GWCMSr6hKP\n8vXAJ6o6/Hxt/OVHRPK56Z7xjWicv4KO+rojF8s9PNsNCAfW+Lo/F2EysEBVV/q6IwXgCvclzR0i\nMktEKp3/EL8VD6wXkbnuy5obReQhX3eqILh/Z9+H81d4IPoaaCUiVwCIyNXATcAnPu1V/hXD2S/u\ndKbyFPI4qugX9xHxsfxsumeKmHuUagKwWlUD9tq9iNTDSTzS/4q4PVDvbeNOpK7BGTYPdN8A3XE2\nyyyPc6PEL0Sknqom+bBf+RWHM0r1GjAauB6YKCKnVXWmT3t28W7H2cj0XV93JJ/GAqWALSJyDmdA\nYJiqfuDbbuWPqp507+X2nIhswfnsvBfnD/lteWnDEhETKKYAV+L85RDItgBX4/wivQt4T0RuDrRk\nREQux0kMb1XVVF/352KpquctqX8SkXXAbqALEIiXzVzAOlVNv8njD+4kuDcQ6IlIT2Cxqh70dUfy\nqSvOB3U3nDuEXwO8LiL7AzhJvB94B9gHnAU2ArNxrjaclyUicBg4hzNJzVNZIFDf6EFFRN4A2gPN\nVPWAr/tzMVT1LLDT/fQ7EbkeeBLnr9dA0hBna4WN7tEqcEYWb3ZPugvVAJ6ApqrHReR/QA1f9yWf\nDuDsNu4pAbjDB30pMCJSGWfieufz1fVjLwNjVPXf7uc/i0hVYAgBmiSq6i6ghXshQSlV/U1EPuDP\n33W5+svPEXH/NZe+6R7gtene177ql3G4k5C/AS1UdY+v+1MIXECorzuRD8txNpG8BmeE52pgPTAL\nuDqQkxDImIRbA+cDPRB9RdZLy7VwRnkCWU+cof9AnU8Bzrywc5nK0giCz2NVTXEnITE4q7Xm5+U4\nGxFxjAdmiLMLcPqme+E4G/EFFPeGgDWA9L9S49yToY6q6l7f9ezCicgU4B6czQqTRCR91Oq4qgbc\n7sgi8hKwGNgDROFMuLsFaOPLfuWHe96E11wdEUkCjqhq5r/E/Z6IvAIswPmgrgi8AKQC7/uyXxfh\nH8BXIjIEZ4nrDcBDwMO5HuXH3H8gdgdmqGqaj7tzMRYAz4rIr8DPOMv5+wPTfNqriyAibXA+c7YC\nV+CM+mwmj5+hloiQp033Asl1wCqc1SWKM1kNnIldPX3VqXzqjRPDZ5nKewDvFXlvLt5lOD+H8sBx\nYBPQJkhWnEBg37fmcpxr2qWB34HVwI2qesSnvconVV0vIrfjTIx8DtgFPBmoEyLdbsXZ/DQQ5+x4\n6gO8iLPi7DJgP/CmuyxQXQKMwUnijwLzgGdVNfPIT7bsPiLGGGOM8ZmAvyZljDHGmMBliYgxxhhj\nfMYSEWOMMcb4jCUixhhjjPEZS0SMMcYY4zOWiBhjjDHGZywRMcYYY4zPWCJijDHGGJ+xRMQYk0FE\nDojIIxdQv62InBOREoXZr0AnIu+LyGxf98MYf2SJiDEBRETS3B/8adk8zonI8Is8RT2c29Dn1Qqg\nvKqeucjzGmP+omyvGWMCSzmPr7vhbM5Wkz83OTyZ3UEiEpKXfR8udG8VVT0LHLqQY4wxxpONiBgT\nQFT1UPoDZ+M8VdXfPcqT3ZdL0kSktYh8JyKngYYiUktEFojIbyJyQkTWiMgtnu17XpoRkVB3Ow+4\nj0sSkS0icptH/fRzlXA/7+Vuo4O77gn3saU9jikuIm+KyHF3X57Py6ULEWkhIl+JSLKI/CIir4pI\nmPu1eiKSIiKdPeo/ICKJIlLd/byxiCwXkcMicsz9dX2P+unx9hCRxe54fxSR9O/dlyJyUkS+EJFK\nHseNcX8v+4jIr+46s9w7YecUi0tEhovILvd5NohIJ4/XS4vIByLyuzveBBG5J7fvjzGByhIRY4LX\nS0A/oA6wBYgEPgJuwdl6/HNggYiUPU87I3B2PK2Ps7PzbBGJ9Hg9886Z0cDjQFegOVALZxfYdMOB\n24F7gJtxdlRtl1sHRKQO8DEwC6gL3IezG+trAKr6EzAEmCoi5USkKjAR6KeqO9zNRAJTgRuBJsCv\nwCciEprpdMOBt4CrgT3A/+HslDocaASUBCZkOqYu0AFo6/63STZ1PL0A3ImzI3ZdYAowR0Sud78+\nDqgKtAZqA0/g7GpqTPBRVXvYwx4B+AAeBI5mU94WOAfcmoc2tgE9PZ4fAB5xfx0KpAGDPV6PcZfd\nnOlcJdzPe7mfl/M4pj+w0+P5UeBRj+fFgH3A7Fz6ORP4R6ayVsBpwOVRthRYBnwBfHie2IsDyUDL\nXOK9xV3WNafvO8725ylAaY+yv7n7Fu1+/n56fECE+7xXZxPjNI84Jvv6PWYPexTFw+aIGBO8Nng+\nEZFSwEic5KEcTgIQBlQ+Tzs/pn+hqsdE5AxwWS71j6rqQY/nB9Lri8hlOCMm33q0eVZEvj9PH64G\naojIQ54hASE4Iyq73WU9cEZ/knFGXf6sLFIeGA00c/fHBZQga/w/enz9G86Iz0+Zyi4RkWLqzJEB\n2KHe82vW4CQ6V3jG6lYL5/v+pYiIR3lx4Gv311OAD0TkBuBTnKQqczvGBAVLRIwJXkmZnk8EbgAG\nATtx/opfiPNhnJvUTM+V3C/rXmj9vIgEJgFvZ/Parx5fN8D5kC8GlAU8k4P3cT7sHwf24oxYfEfW\n+D37r7mU5TemSHcbrTL1D+AUgKr+V0Qq41zmuRX4QkReUdWLXRVljN+xRMSYv44mwNuqugBARKJx\nRhOKjKoeEpE/cOZarHf3oxhwDc6clZxsBK5U1Z05VXCPtkwDnsWZFzNbRBqpanoS0Ri4V1WXuevX\nAKIuMqR01UWktMeoSGPgLM6lr8x+dL9WObdRDlX9HZgBzBCRdcBQnHkqxgQVS0SM+evYBtwtIstw\n/u+PwpnPUdTeAJ4Xkd3ADmAgEE7WSa+eXgK+EpHxOB/OKTj3PLlZVfu760wDElT1ZRGJAr7Hmb/x\nlPv17cCDIvIjEAu8jHsE4jzk/FU4A7wrIkOA0jiTaN9T1T8yV3Rf3poIvOFe9bMG53JVU+CQqn4g\nIqPd5Ztx5pS0c39tTNCxRMSYv46+OB/Wa3Du/TEaZ/Kpp8zJQHbJQW4JQ168iJMIzMb5AH8LZ3Jp\njkmBqm4UkeY4ydNqdx+246xoQUQexpn7cZW7fqKIPACsFJFFqroKZ5LpmzgJyi/AM2S91JPfeH/G\nmWC6FCgFzMdZsZRTPE+LyH6c0ZtqwDGcOT2j3FXO4iRKVXAusX0OPJyHfhgTcET1Yn+nGGNM/omI\nCyepmKqqY3zdnwslImOAW1S1ia/7YkwgshERY0yREpE4nGWxX+JckumPs4rnA1/2yxjjG3ZDM2NM\nUVOcywzrcS45xAEtVHWXT3tljPEJuzRjjDHGGJ+xERFjjDHG+IwlIsYYY4zxGUtEjDHGGOMzlogY\nY4wxxmcsETHGGGOMz1giYowxxhifsUTEGGOMMT5jiYgxxhhjfMYSEWOMMcb4zP8DQxGzv6qCAAAA\nAklEQVRKlm8BLEQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10abf70940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# XGBCRegressor\n",
    "draw_learning_curves(X_train, y_train, model_A,\n",
    "                     title=\"Learning Curves of XGBRegressor\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[learning_curve] Training set sizes: [ 2808  5616  8424 11232 14040 16848 19656 22464 25272 28080]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 16 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   1 out of   1 | elapsed:    6.2s remaining:    0.0s\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of  50 | elapsed:   12.1s remaining:  3.2min\n",
      "[Parallel(n_jobs=-1)]: Done   9 out of  50 | elapsed:   22.3s remaining:  1.7min\n",
      "[Parallel(n_jobs=-1)]: Done  15 out of  50 | elapsed:   32.7s remaining:  1.3min\n",
      "[Parallel(n_jobs=-1)]: Done  21 out of  50 | elapsed:   40.0s remaining:   55.3s\n",
      "[Parallel(n_jobs=-1)]: Done  27 out of  50 | elapsed:   51.5s remaining:   43.9s\n",
      "[Parallel(n_jobs=-1)]: Done  33 out of  50 | elapsed:  1.0min remaining:   31.5s\n",
      "[Parallel(n_jobs=-1)]: Done  39 out of  50 | elapsed:  1.2min remaining:   20.1s\n",
      "[Parallel(n_jobs=-1)]: Done  45 out of  50 | elapsed:  1.4min remaining:    9.1s\n",
      "[Parallel(n_jobs=-1)]: Done  50 out of  50 | elapsed:  1.6min finished\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhoAAAGHCAYAAAD2qfsmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10aaa0a518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# RandomForestRegressor\n",
    "draw_learning_curves(X_train, y_train, model_B,\n",
    "                     title=\"Learning Curves of RandomForestRegressor\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discusion \n",
    "\n",
    "For XGBRegressor, the training score keeps decreasing, and the testing score keeps increasing. While for the RandomForestRegressor, the training score remains constant at a high level and the cross-validation score is low. It is clear that the best model is XGBRegressor because the RandomForestRegressor overfits the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Step 4:  Parameter Tuning\n",
    "\n",
    "#### Approach for Parameter Tuning :\n",
    " - 1. Choose a relatively high learning rate. Generally, a learning rate of 0.1 works but somewhere between 0.05 to 0.3 should work for different problems. Determine the optimum number of trees for this learning rate. XGBoost has a handy function called “cv” which performs cross-validation at each boosting iteration and thus returns the optimum number of trees required.\n",
    " - 2. Tune tree-specific parameters ( max_depth, min_child_weight, gamma, subsample, colsample_bytree) for decided learning rate and the number of trees.\n",
    " - 3. Tune regularization parameters (lambda, alpha) for xgboost which can help reduce model complexity and enhance performance.\n",
    " - 4. Lower the learning rate and decide the optimal parameters.\n",
    " \n",
    "This approach  inspired from  [Complete Guide to Parameter Tuning in XGBoost](https://www.analyticsvidhya.com/blog/2016/03/complete-guide-parameter-tuning-xgboost-with-codes-python/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# function for modeling and cross-validation\n",
    "\n",
    "\n",
    "def modelval(alg, X_train, y_train, useTrainCV=True,\n",
    "             cv_folds=5, early_stopping_rounds=50):\n",
    "    \n",
    "    if useTrainCV:\n",
    "        xgb_param = alg.get_xgb_params()\n",
    "        xgtrain = xgb.DMatrix(X_train, label=y_train)\n",
    "        \n",
    "        cvresult = xgb.cv(xgb_param, xgtrain,\n",
    "                          num_boost_round=alg.get_params()[\n",
    "                              'n_estimators'], nfold=cv_folds,\n",
    "                          metrics='auc', verbose_eval=20,\n",
    "                          early_stopping_rounds=early_stopping_rounds,\n",
    "                          shuffle=True)\n",
    "        \n",
    "        alg.set_params(n_estimators=cvresult.shape[0])\n",
    "        \n",
    "        print('n_estimators=', cvresult.shape[0])\n",
    "        \n",
    "    # cross-validation\n",
    "    all_auc = cross_val_score(alg, X_train, y_train,\n",
    "                              scoring='roc_auc', cv=5,\n",
    "                              n_jobs=-1, verbose=10\n",
    "                             )\n",
    "    \n",
    "    auc = np.average(all_auc)\n",
    "    print(\"AUC Score : %f\" % auc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.1- Find the number of estimators "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-auc:0.762176+0.00427215\ttest-auc:0.755931+0.0131404\n",
      "[20]\ttrain-auc:0.861178+0.00656747\ttest-auc:0.774002+0.0172957\n",
      "[40]\ttrain-auc:0.925636+0.00414626\ttest-auc:0.774429+0.0183072\n",
      "[60]\ttrain-auc:0.971717+0.00412828\ttest-auc:0.775152+0.0166882\n",
      "n_estimators= 13\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   1 tasks      | elapsed:    7.7s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.767045\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:   12.1s remaining:    8.1s\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:   12.3s remaining:    0.0s\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:   12.3s finished\n"
     ]
    }
   ],
   "source": [
    "xgb1 = XGBRegressor(\n",
    "        learning_rate=0.1,\n",
    "        n_estimators=1000,\n",
    "        max_depth=5,\n",
    "        min_child_weight=1,\n",
    "        gamma=0,\n",
    "        subsample=0.8,\n",
    "        colsample_bytree=0.8,\n",
    "        objective='binary:logistic',\n",
    "        nthread=4,\n",
    "        scale_pos_weight=1,\n",
    "        seed=27,\n",
    "        random_state=42)\n",
    "\n",
    "modelval(xgb1, X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.2: Tune max_depth and min_child_weight"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 35 candidates, totalling 175 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   1 tasks      | elapsed:    7.2s\n",
      "[Parallel(n_jobs=-1)]: Done   2 tasks      | elapsed:   13.0s\n",
      "[Parallel(n_jobs=-1)]: Done  13 tasks      | elapsed:   26.0s\n",
      "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:   26.3s\n",
      "[Parallel(n_jobs=-1)]: Done  37 tasks      | elapsed:   43.4s\n",
      "[Parallel(n_jobs=-1)]: Done  50 tasks      | elapsed:   45.1s\n",
      "[Parallel(n_jobs=-1)]: Done  65 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=-1)]: Done  80 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done  97 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done 114 tasks      | elapsed:  1.9min\n",
      "[Parallel(n_jobs=-1)]: Done 146 out of 175 | elapsed:  2.5min remaining:   30.1s\n",
      "[Parallel(n_jobs=-1)]: Done 164 out of 175 | elapsed:  3.0min remaining:   12.3s\n",
      "[Parallel(n_jobs=-1)]: Done 175 out of 175 | elapsed:  3.1min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise-deprecating',\n",
       "       estimator=XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "       colsample_bytree=0.8, gamma=0, importance_type='gain',\n",
       "       learning_rate=0.1, max_delta_step=0, max_depth=5,\n",
       "       min_child_weight=1, missing=None, n_estimators=13, n_jobs=1,\n",
       "       nthread=4, objective='binary:logistic', random_state=0, reg_alpha=0,\n",
       "       reg_lambda=1, scale_pos_weight=1, seed=27, silent=True,\n",
       "       subsample=0.8),\n",
       "       fit_params=None, iid=False, n_jobs=-1,\n",
       "       param_grid={'min_child_weight': range(1, 6), 'max_depth': range(3, 10)},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring='roc_auc', verbose=10)"
      ]
     },
     "execution_count": 289,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grid seach on subsample and max_features\n",
    "\n",
    "param_test1 = {\n",
    "    'max_depth':range(3,10,1),\n",
    "    'min_child_weight':range(1,6,1)\n",
    "}\n",
    "\n",
    "gsearch1 = GridSearchCV(estimator = XGBRegressor(learning_rate=0.1, n_estimators=13,\n",
    "                                                  max_depth=5, min_child_weight=1, gamma=0,\n",
    "                                                  subsample=0.8, colsample_bytree=0.8,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test1, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch1.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.7761149044220699, {'max_depth': 3, 'min_child_weight': 5})"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gsearch1.best_score_, gsearch1.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 2 candidates, totalling 10 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of  10 | elapsed:    4.0s remaining:    9.4s\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of  10 | elapsed:    4.2s remaining:    4.2s\n",
      "[Parallel(n_jobs=-1)]: Done   7 out of  10 | elapsed:    4.4s remaining:    1.9s\n",
      "[Parallel(n_jobs=-1)]: Done  10 out of  10 | elapsed:    4.9s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7761149044220699\n",
      "Best params:  {'max_depth': 3}\n"
     ]
    }
   ],
   "source": [
    "param_test2 = {\n",
    " 'max_depth':[2,3],\n",
    "}\n",
    "\n",
    "gsearch2 = GridSearchCV(estimator = XGBRegressor(learning_rate=0.1, n_estimators=13,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0,\n",
    "                                                  subsample=0.8, colsample_bytree=0.8,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test2, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch2.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch2.best_score_)\n",
    "print('Best params: ', gsearch2.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- max_depth: 3 \n",
    "- min_child_weight: 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.3: Tune gamma"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 5 candidates, totalling 25 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 out of  25 | elapsed:   16.9s remaining:  3.2min\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of  25 | elapsed:   17.5s remaining:  1.2min\n",
      "[Parallel(n_jobs=-1)]: Done   8 out of  25 | elapsed:   17.8s remaining:   37.8s\n",
      "[Parallel(n_jobs=-1)]: Done  11 out of  25 | elapsed:   19.6s remaining:   25.0s\n",
      "[Parallel(n_jobs=-1)]: Done  14 out of  25 | elapsed:   20.3s remaining:   16.0s\n",
      "[Parallel(n_jobs=-1)]: Done  17 out of  25 | elapsed:   20.4s remaining:    9.6s\n",
      "[Parallel(n_jobs=-1)]: Done  20 out of  25 | elapsed:   20.5s remaining:    5.1s\n",
      "[Parallel(n_jobs=-1)]: Done  23 out of  25 | elapsed:   20.6s remaining:    1.8s\n",
      "[Parallel(n_jobs=-1)]: Done  25 out of  25 | elapsed:   20.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7766101185611339\n",
      "Best params:  {'gamma': 0.4}\n"
     ]
    }
   ],
   "source": [
    "param_test3 = {\n",
    "    'gamma':[i/10.0 for i in range(0,5)]\n",
    "}\n",
    "\n",
    "gsearch3 = GridSearchCV(estimator = XGBRegressor(learning_rate=0.1, n_estimators=13,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0,\n",
    "                                                  subsample=0.8, colsample_bytree=0.8,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test3, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch3.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch3.best_score_)\n",
    "print('Best params: ', gsearch3.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- gamma': 0.4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.4: re-calibrate the number of boosting rounds for the updated parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-auc:0.761442+0.00457465\ttest-auc:0.755995+0.0131639\n",
      "[20]\ttrain-auc:0.81019+0.0054802\ttest-auc:0.77589+0.0185268\n",
      "[40]\ttrain-auc:0.844858+0.00448446\ttest-auc:0.779018+0.0155921\n",
      "[60]\ttrain-auc:0.874066+0.00241827\ttest-auc:0.778594+0.0176325\n",
      "[80]\ttrain-auc:0.901496+0.00159629\ttest-auc:0.780378+0.0195711\n",
      "n_estimators= 45\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    4.8s remaining:    7.3s\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    5.0s remaining:    3.3s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.771492\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    5.3s remaining:    0.0s\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    5.3s finished\n"
     ]
    }
   ],
   "source": [
    "xgb2 = XGBRegressor(\n",
    "        learning_rate=0.1,\n",
    "        n_estimators=1000,\n",
    "        max_depth=3,\n",
    "        min_child_weight=5,\n",
    "        gamma=0.4,\n",
    "        subsample=0.8,\n",
    "        colsample_bytree=0.8,\n",
    "        objective='binary:logistic',\n",
    "        nthread=4,\n",
    "        scale_pos_weight=1,\n",
    "        seed=27,\n",
    "        random_state=42)\n",
    "\n",
    "modelval(xgb2, X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.5: Tune subsample and colsample_bytree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 16 candidates, totalling 80 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 tasks      | elapsed:   45.0s\n",
      "[Parallel(n_jobs=-1)]: Done  13 tasks      | elapsed:   50.4s\n",
      "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:   55.0s\n",
      "[Parallel(n_jobs=-1)]: Done  42 out of  80 | elapsed:  1.6min remaining:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done  51 out of  80 | elapsed:  2.2min remaining:  1.3min\n",
      "[Parallel(n_jobs=-1)]: Done  60 out of  80 | elapsed:  2.4min remaining:   47.3s\n",
      "[Parallel(n_jobs=-1)]: Done  69 out of  80 | elapsed:  2.4min remaining:   23.1s\n",
      "[Parallel(n_jobs=-1)]: Done  78 out of  80 | elapsed:  2.5min remaining:    3.9s\n",
      "[Parallel(n_jobs=-1)]: Done  80 out of  80 | elapsed:  2.5min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7784973500671042\n",
      "Best params:  {'colsample_bytree': 0.6, 'subsample': 0.6}\n"
     ]
    }
   ],
   "source": [
    "param_test4 = {\n",
    "    'subsample':[i/10.0 for i in range(6,10)],\n",
    "    'colsample_bytree':[i/10.0 for i in range(6,10)]\n",
    "}\n",
    "\n",
    "gsearch4 = GridSearchCV(estimator = XGBRegressor(learning_rate=0.1, n_estimators=45,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0.4,\n",
    "                                                  subsample=0.8, colsample_bytree=0.8,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test4, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch4.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch4.best_score_)\n",
    "print('Best params: ', gsearch4.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 36 candidates, totalling 180 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 tasks      | elapsed:   32.0s\n",
      "[Parallel(n_jobs=-1)]: Done  13 tasks      | elapsed:   38.3s\n",
      "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:   41.5s\n",
      "[Parallel(n_jobs=-1)]: Done  37 tasks      | elapsed:  1.3min\n",
      "[Parallel(n_jobs=-1)]: Done  50 tasks      | elapsed:  1.8min\n",
      "[Parallel(n_jobs=-1)]: Done  65 tasks      | elapsed:  2.0min\n",
      "[Parallel(n_jobs=-1)]: Done  80 tasks      | elapsed:  2.6min\n",
      "[Parallel(n_jobs=-1)]: Done  97 tasks      | elapsed:  3.1min\n",
      "[Parallel(n_jobs=-1)]: Done 114 tasks      | elapsed:  3.4min\n",
      "[Parallel(n_jobs=-1)]: Done 133 tasks      | elapsed:  4.1min\n",
      "[Parallel(n_jobs=-1)]: Done 152 out of 180 | elapsed:  4.7min remaining:   51.5s\n",
      "[Parallel(n_jobs=-1)]: Done 171 out of 180 | elapsed:  5.1min remaining:   16.0s\n",
      "[Parallel(n_jobs=-1)]: Done 180 out of 180 | elapsed:  5.1min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7843527633228524\n",
      "Best params:  {'colsample_bytree': 0.5, 'subsample': 0.45}\n"
     ]
    }
   ],
   "source": [
    "param_test5 = {\n",
    " 'subsample':[i/100.0 for i in range(40, 70,5)],\n",
    " 'colsample_bytree':[i/100.0 for i in range(40,70,5)]\n",
    "}\n",
    "\n",
    "gsearch5 = GridSearchCV(estimator = XGBRegressor(learning_rate=0.1, n_estimators=45,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0.4,\n",
    "                                                  subsample=0.6, colsample_bytree=0.6,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test5, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch5.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch5.best_score_)\n",
    "print('Best params: ', gsearch5.best_params_)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- colsample_bytree: 0.5\n",
    "- subsample: 0.45"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.6: Tuning Regularization Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 5 candidates, totalling 25 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 out of  25 | elapsed:   23.2s remaining:  4.4min\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of  25 | elapsed:   25.9s remaining:  1.7min\n",
      "[Parallel(n_jobs=-1)]: Done   8 out of  25 | elapsed:   33.3s remaining:  1.2min\n",
      "[Parallel(n_jobs=-1)]: Done  11 out of  25 | elapsed:   33.9s remaining:   43.1s\n",
      "[Parallel(n_jobs=-1)]: Done  14 out of  25 | elapsed:   35.4s remaining:   27.8s\n",
      "[Parallel(n_jobs=-1)]: Done  17 out of  25 | elapsed:   35.8s remaining:   16.9s\n",
      "[Parallel(n_jobs=-1)]: Done  20 out of  25 | elapsed:   36.0s remaining:    9.0s\n",
      "[Parallel(n_jobs=-1)]: Done  23 out of  25 | elapsed:   36.2s remaining:    3.1s\n",
      "[Parallel(n_jobs=-1)]: Done  25 out of  25 | elapsed:   36.6s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7843522678113103\n",
      "Best params:  {'reg_alpha': 1e-05}\n"
     ]
    }
   ],
   "source": [
    "param_test6 = {\n",
    " 'reg_alpha':[1e-5, 1e-2, 0.1, 1, 100]\n",
    "}\n",
    "\n",
    "gsearch6 = GridSearchCV(estimator = XGBRegressor(learning_rate =0.1, n_estimators=45,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0.4,\n",
    "                                                  subsample=0.45, colsample_bytree=0.5,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test6, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch6.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch6.best_score_)\n",
    "print('Best params: ', gsearch6.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0]\ttrain-auc:0.5+0\ttest-auc:0.5+0\n",
      "[20]\ttrain-auc:0.808729+0.0072023\ttest-auc:0.776281+0.0184808\n",
      "[40]\ttrain-auc:0.841839+0.00386715\ttest-auc:0.785332+0.0118756\n",
      "[60]\ttrain-auc:0.869725+0.00484593\ttest-auc:0.787045+0.012746\n",
      "[80]\ttrain-auc:0.890851+0.00274643\ttest-auc:0.784556+0.0163174\n",
      "[100]\ttrain-auc:0.907865+0.00316484\ttest-auc:0.785494+0.016355\n",
      "n_estimators= 52\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    4.3s remaining:    6.5s\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    4.5s remaining:    3.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC Score : 0.783332\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    4.6s remaining:    0.0s\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    4.6s finished\n"
     ]
    }
   ],
   "source": [
    "xgb3 = XGBRegressor(\n",
    "        learning_rate=0.1,\n",
    "        n_estimators=1000,\n",
    "        max_depth=3,\n",
    "        min_child_weight=5,\n",
    "        gamma=0.4,\n",
    "        subsample=0.45,\n",
    "        colsample_bytree=0.5,\n",
    "        objective='binary:logistic',\n",
    "        nthread=4,\n",
    "        scale_pos_weight=1,\n",
    "        seed=27,\n",
    "        random_state=42)\n",
    "\n",
    "modelval(xgb3, X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step 4.7: Reducing Learning Rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 6 candidates, totalling 30 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 24 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   3 out of  30 | elapsed:   43.0s remaining:  6.5min\n",
      "[Parallel(n_jobs=-1)]: Done   7 out of  30 | elapsed:   49.0s remaining:  2.7min\n",
      "[Parallel(n_jobs=-1)]: Done  11 out of  30 | elapsed:   51.0s remaining:  1.5min\n",
      "[Parallel(n_jobs=-1)]: Done  15 out of  30 | elapsed:   52.1s remaining:   52.1s\n",
      "[Parallel(n_jobs=-1)]: Done  19 out of  30 | elapsed:   53.5s remaining:   31.0s\n",
      "[Parallel(n_jobs=-1)]: Done  23 out of  30 | elapsed:   54.1s remaining:   16.5s\n",
      "[Parallel(n_jobs=-1)]: Done  27 out of  30 | elapsed:   57.3s remaining:    6.4s\n",
      "[Parallel(n_jobs=-1)]: Done  30 out of  30 | elapsed:   57.7s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best score:  0.7857534160662417\n",
      "Best params:  {'learning_rate': 0.08}\n"
     ]
    }
   ],
   "source": [
    "param_test7 = {\n",
    " 'learning_rate':[0.05, 0.06, 0.07, 0.08, 0.09, 0.1]\n",
    "}\n",
    "\n",
    "gsearch7 = GridSearchCV(estimator = XGBRegressor(learning_rate =0.1, n_estimators=52,\n",
    "                                                  max_depth=3, min_child_weight=5, gamma=0.4,\n",
    "                                                  subsample=0.45, colsample_bytree=0.5,\n",
    "                                                  objective='binary:logistic', nthread=4,\n",
    "                                                  scale_pos_weight=1, seed=27),\n",
    "                        param_grid = param_test7, scoring='roc_auc', n_jobs=-1,\n",
    "                        iid=False, cv=5, verbose=10)\n",
    "gsearch7.fit(X_train,y_train)\n",
    "\n",
    "print('Best score: ', gsearch7.best_score_)\n",
    "print('Best params: ', gsearch7.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "XGBRegressor(base_score=0.5, booster='gbtree', colsample_bylevel=1,\n",
       "       colsample_bytree=0.5, gamma=0.4, importance_type='gain',\n",
       "       learning_rate=0.08, max_delta_step=0, max_depth=3,\n",
       "       min_child_weight=5, missing=None, n_estimators=52, n_jobs=1,\n",
       "       nthread=4, objective='binary:logistic', random_state=42,\n",
       "       reg_alpha=0, reg_lambda=1, scale_pos_weight=1, seed=27, silent=True,\n",
       "       subsample=0.45)"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# final model\n",
    "model = XGBRegressor(\n",
    "        learning_rate=0.08,\n",
    "        n_estimators=52,\n",
    "        max_depth=3,\n",
    "        min_child_weight=5,\n",
    "        gamma=0.4,\n",
    "        subsample=0.45,\n",
    "        colsample_bytree=0.5,\n",
    "        objective='binary:logistic',\n",
    "        nthread=4,\n",
    "        scale_pos_weight=1,\n",
    "        seed=27,\n",
    "        random_state=42)\n",
    "\n",
    "# train the model\n",
    "model.fit(X_train, y_train, eval_metric='auc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.figure(figsize=(8, 6))\n",
    "sns.barplot(x=\"RESPONSE\", y=\"% of data\", data=df_sp)\n",
    "matplotlib.rc('xtick', labelsize=25)  \n",
    "matplotlib.rc('ytick', labelsize=25)\n",
    "plt.xlabel('RESPONSE', fontsize=15)\n",
    "plt.ylabel('% of data', fontsize=15)\n",
    "plt.title('Distribution of customers responses', fontsize=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjMAAAJqCAYAAADe08H9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZpiRJElDzTAjSZKG2v8D4mqBqwH+b4QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f10ab8d93c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# feature importance\n",
    "feat_imp = pd.Series(model.feature_importances_,\n",
    "                     index=features.columns).sort_values(ascending=False)\n",
    "# plot 20 important features\n",
    "fig = plt.figure()\n",
    "feat_imp.iloc[:10].plot(kind='bar') #, title='Feature Importances')\n",
    "matplotlib.rc('xtick', labelsize=12)  \n",
    "matplotlib.rc('ytick', labelsize=12)\n",
    "plt.xlabel('Features', fontsize=12)\n",
    "plt.ylabel('Weight', fontsize=12)\n",
    "plt.title('Normalized weight for the most predictive features', fontsize=12)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  <a class=\"anchor\" id=\"part3\">Part 3: Kaggle Competition</a>\n",
    "\n",
    "Now that I've created a model to predict which individuals are most likely to respond to a mailout campaign, it's time to test that model in competition through [Kaggle](http://www.kaggle.com/t/21e6d45d4c574c7fa2d868f0e8c83140)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/IPython/core/interactiveshell.py:2717: DtypeWarning: Columns (18,19) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\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>LNR</th>\n",
       "      <th>AGER_TYP</th>\n",
       "      <th>AKT_DAT_KL</th>\n",
       "      <th>ALTER_HH</th>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <th>ANZ_HAUSHALTE_AKTIV</th>\n",
       "      <th>...</th>\n",
       "      <th>VHN</th>\n",
       "      <th>VK_DHT4A</th>\n",
       "      <th>VK_DISTANZ</th>\n",
       "      <th>VK_ZG11</th>\n",
       "      <th>W_KEIT_KIND_HH</th>\n",
       "      <th>WOHNDAUER_2008</th>\n",
       "      <th>WOHNLAGE</th>\n",
       "      <th>ZABEOTYP</th>\n",
       "      <th>ANREDE_KZ</th>\n",
       "      <th>ALTERSKATEGORIE_GROB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1754</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1770</td>\n",
       "      <td>-1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1465</td>\n",
       "      <td>2</td>\n",
       "      <td>9.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1470</td>\n",
       "      <td>-1</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1478</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 366 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    LNR  AGER_TYP  AKT_DAT_KL  ALTER_HH  ALTER_KIND1  ALTER_KIND2  \\\n",
       "0  1754         2         1.0       7.0          NaN          NaN   \n",
       "1  1770        -1         1.0       0.0          NaN          NaN   \n",
       "2  1465         2         9.0      16.0          NaN          NaN   \n",
       "3  1470        -1         7.0       0.0          NaN          NaN   \n",
       "4  1478         1         1.0      21.0          NaN          NaN   \n",
       "\n",
       "   ALTER_KIND3  ALTER_KIND4  ALTERSKATEGORIE_FEIN  ANZ_HAUSHALTE_AKTIV  ...  \\\n",
       "0          NaN          NaN                   6.0                  2.0  ...   \n",
       "1          NaN          NaN                   0.0                 20.0  ...   \n",
       "2          NaN          NaN                  11.0                  2.0  ...   \n",
       "3          NaN          NaN                   0.0                  1.0  ...   \n",
       "4          NaN          NaN                  13.0                  1.0  ...   \n",
       "\n",
       "   VHN  VK_DHT4A  VK_DISTANZ  VK_ZG11  W_KEIT_KIND_HH  WOHNDAUER_2008  \\\n",
       "0  4.0       5.0         6.0      3.0             6.0             9.0   \n",
       "1  1.0       5.0         2.0      1.0             6.0             9.0   \n",
       "2  3.0       9.0         6.0      3.0             2.0             9.0   \n",
       "3  2.0       6.0         6.0      3.0             NaN             9.0   \n",
       "4  1.0       2.0         4.0      3.0             3.0             9.0   \n",
       "\n",
       "   WOHNLAGE ZABEOTYP ANREDE_KZ ALTERSKATEGORIE_GROB  \n",
       "0       3.0        3         1                    4  \n",
       "1       5.0        3         1                    4  \n",
       "2       4.0        3         2                    4  \n",
       "3       2.0        3         2                    4  \n",
       "4       7.0        4         2                    4  \n",
       "\n",
       "[5 rows x 366 columns]"
      ]
     },
     "execution_count": 317,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load data\n",
    "mailout_test = pd.read_csv('Arvato_data/Udacity_MAILOUT_052018_TEST.csv',\n",
    "                           sep=';')\n",
    "mailout_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jaouadeddadsi2016/anaconda3/lib/python3.5/site-packages/pandas/core/ops.py:1649: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  result = method(y)\n"
     ]
    }
   ],
   "source": [
    "# Clean data\n",
    "clean_test = clean_data(mailout_test, drop_row=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Drop LNR\n",
    "test_LNR = clean_test['LNR']\n",
    "\n",
    "clean_test.drop('LNR', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 321,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Columns to add\n",
    "mask_1 = features.columns.isin(clean_test.columns)\n",
    "col_to_add = list(features.columns[~mask_1])\n",
    "col_to_add"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 322,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Columns to drop\n",
    "mask_2 = clean_test.columns.isin(features.columns)\n",
    "col_to_drop = list(clean_test.columns[~mask_2])\n",
    "col_to_drop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {
    "collapsed": 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>CJT_GESAMTTYP_1.0</th>\n",
       "      <th>CJT_GESAMTTYP_2.0</th>\n",
       "      <th>CJT_GESAMTTYP_3.0</th>\n",
       "      <th>CJT_GESAMTTYP_4.0</th>\n",
       "      <th>CJT_GESAMTTYP_5.0</th>\n",
       "      <th>CJT_GESAMTTYP_6.0</th>\n",
       "      <th>FINANZTYP_1.0</th>\n",
       "      <th>FINANZTYP_2.0</th>\n",
       "      <th>FINANZTYP_3.0</th>\n",
       "      <th>FINANZTYP_4.0</th>\n",
       "      <th>...</th>\n",
       "      <th>WOHNLAGE_3.0</th>\n",
       "      <th>WOHNLAGE_4.0</th>\n",
       "      <th>WOHNLAGE_5.0</th>\n",
       "      <th>WOHNLAGE_7.0</th>\n",
       "      <th>WOHNLAGE_8.0</th>\n",
       "      <th>PLZ8_BAUMAX_1.0</th>\n",
       "      <th>PLZ8_BAUMAX_2.0</th>\n",
       "      <th>PLZ8_BAUMAX_3.0</th>\n",
       "      <th>PLZ8_BAUMAX_4.0</th>\n",
       "      <th>PLZ8_BAUMAX_5.0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\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",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</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",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</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",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 483 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   CJT_GESAMTTYP_1.0  CJT_GESAMTTYP_2.0  CJT_GESAMTTYP_3.0  CJT_GESAMTTYP_4.0  \\\n",
       "0                0.0                0.0                0.0                0.0   \n",
       "1                1.0                0.0                0.0                0.0   \n",
       "2                0.0                1.0                0.0                0.0   \n",
       "3                0.0                0.0                0.0                1.0   \n",
       "4                0.0                0.0                0.0                0.0   \n",
       "\n",
       "   CJT_GESAMTTYP_5.0  CJT_GESAMTTYP_6.0  FINANZTYP_1.0  FINANZTYP_2.0  \\\n",
       "0                1.0                0.0            0.0            0.0   \n",
       "1                0.0                0.0            0.0            0.0   \n",
       "2                0.0                0.0            0.0            1.0   \n",
       "3                0.0                0.0            0.0            0.0   \n",
       "4                0.0                1.0            0.0            0.0   \n",
       "\n",
       "   FINANZTYP_3.0  FINANZTYP_4.0  ...  WOHNLAGE_3.0  WOHNLAGE_4.0  \\\n",
       "0            0.0            0.0  ...           1.0           0.0   \n",
       "1            0.0            0.0  ...           0.0           0.0   \n",
       "2            0.0            0.0  ...           0.0           1.0   \n",
       "3            0.0            0.0  ...           0.0           0.0   \n",
       "4            0.0            0.0  ...           0.0           0.0   \n",
       "\n",
       "   WOHNLAGE_5.0  WOHNLAGE_7.0  WOHNLAGE_8.0  PLZ8_BAUMAX_1.0  PLZ8_BAUMAX_2.0  \\\n",
       "0           0.0           0.0           0.0              0.0              0.0   \n",
       "1           1.0           0.0           0.0              0.0              0.0   \n",
       "2           0.0           0.0           0.0              0.0              0.0   \n",
       "3           0.0           0.0           0.0              0.0              0.0   \n",
       "4           0.0           1.0           0.0              0.0              0.0   \n",
       "\n",
       "   PLZ8_BAUMAX_3.0  PLZ8_BAUMAX_4.0  PLZ8_BAUMAX_5.0  \n",
       "0              0.0              0.0              0.0  \n",
       "1              0.0              0.0              0.0  \n",
       "2              0.0              0.0              0.0  \n",
       "3              0.0              0.0              0.0  \n",
       "4              0.0              0.0              0.0  \n",
       "\n",
       "[5 rows x 483 columns]"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Imputer\n",
    "new_test = pd.DataFrame(Imp.fit_transform(clean_test.values),\n",
    "                        columns= clean_test.columns)\n",
    "\n",
    "# feature scaling\n",
    "new_test = pd.DataFrame(scaler.fit_transform(new_test.values),\n",
    "                        columns= new_test.columns)\n",
    "\n",
    "new_test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {
    "collapsed": 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>RESPONSE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LNR</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1754</th>\n",
       "      <td>0.035629</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1770</th>\n",
       "      <td>0.032007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1465</th>\n",
       "      <td>0.012915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1470</th>\n",
       "      <td>0.017483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1478</th>\n",
       "      <td>0.014195</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      RESPONSE\n",
       "LNR           \n",
       "1754  0.035629\n",
       "1770  0.032007\n",
       "1465  0.012915\n",
       "1470  0.017483\n",
       "1478  0.014195"
      ]
     },
     "execution_count": 324,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Prediction\n",
    "test_label = model.predict(new_test.values)\n",
    "\n",
    "submission = pd.DataFrame(test_label, index=test_LNR.astype('int32'),\n",
    "                          columns=[\"RESPONSE\"])\n",
    "\n",
    "submission.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Save file to submit\n",
    "submission.to_csv('submission.csv')"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}