Arvato Project Workbook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c105f898",
   "metadata": {},
   "source": [
    "# Capstone Project: Create a Customer Segmentation Report for Arvato Financial Services\n",
    "\n",
    "In this project, you will analyze demographics data for customers of a mail-order sales company in Germany, comparing it against demographics information for the general population. You'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, you'll apply what you'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 you will use has been provided by our partners at Bertelsmann Arvato Analytics, and represents a real-life data science task.\n",
    "\n",
    "If you completed the first term of this program, you will be familiar with the first part of this project, from the unsupervised learning project. The versions of those two datasets used in this project will include many more features and has not been pre-cleaned. You are also free to choose whatever approach you'd like to analyzing the data rather than follow pre-determined steps. In your work on this project, make sure that you carefully document your steps and decisions, since your main deliverable for this project will be a blog post reporting your findings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "id": "14bff09c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries here; add more as necessary\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from datetime import datetime\n",
    "import time\n",
    "\n",
    "# magic word for producing visualizations in notebook\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder, StandardScaler\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import accuracy_score, roc_auc_score\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import SVC\n",
    "from xgboost import XGBClassifier\n",
    "import xgboost as xgb\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier, GradientBoostingClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9531f17b",
   "metadata": {},
   "source": [
    "## Part 0: Get to Know the Data\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. 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 your 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 your final predictions 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, you can 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",
    "\n",
    "In the below cell, we've provided some initial code to load in the first two datasets. Note for all of the `.csv` data files in this project that they're semicolon (`;`) delimited, so an additional argument in the [`read_csv()`](https://pandas.pydata.org/pandas-docs/stable/generated/pandas.read_csv.html) call has been included to read in the data properly. Also, considering the size of the datasets, it may take some time for them to load completely.\n",
    "\n",
    "You'll notice when the data is loaded in that a warning message will immediately pop up. Before you really start digging into the modeling and analysis, you're going to need to perform some cleaning. Take some time to browse the structure of the data and look over the informational spreadsheets to understand the data values. Make some decisions on which features to keep, which features to drop, and if any revisions need to be made on data formats. It'll be a good idea to create a function with pre-processing steps, since you'll need to clean all of the datasets before you work with them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "81f29b0d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3444: DtypeWarning: Columns (18,19) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  exec(code_obj, self.user_global_ns, self.user_ns)\n"
     ]
    }
   ],
   "source": [
    "# load in the data\n",
    "azdias = pd.read_csv('E:\\\\udct\\\\Arvato Project\\\\Udacity_AZDIAS_052018.csv', sep=';')\n",
    "customers = pd.read_csv('E:\\\\udct\\\\Arvato Project\\\\Udacity_CUSTOMERS_052018.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "cf2d4f96",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Index(['CAMEO_DEUG_2015', 'CAMEO_INTL_2015'], dtype='object')\n",
      "Index(['CAMEO_DEUG_2015', 'CAMEO_INTL_2015'], dtype='object')\n"
     ]
    }
   ],
   "source": [
    "# Check for columns showing problem\n",
    "print(azdias.columns[18:20])\n",
    "print(customers.columns[18:20])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e311c1d",
   "metadata": {},
   "source": [
    "### Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "210c5207",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(891221, 366) \n",
      "\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": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of azdias dataset\n",
    "print(azdias.shape, \"\\n\")\n",
    "azdias.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "acc43c9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(191652, 369) \n",
      "\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": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of customers dataset\n",
    "print(customers.shape, \"\\n\")\n",
    "customers.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8ca7938c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'CUSTOMER_GROUP', 'ONLINE_PURCHASE', 'PRODUCT_GROUP'}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find the extra features in customers data\n",
    "extra_features = (set(customers.columns) - set(azdias.columns))\n",
    "extra_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a2ce8245",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "float64    267\n",
       "int64       93\n",
       "object       6\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find data types for azdias\n",
    "azdias.dtypes.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ce048a7c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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>8.0</td>\n",
       "      <td>51.0</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>4.0</td>\n",
       "      <td>24.0</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>2.0</td>\n",
       "      <td>12.0</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>6.0</td>\n",
       "      <td>43.0</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             8.0            51.0                      NaN   \n",
       "2             4C             4.0            24.0            D19_UNBEKANNT   \n",
       "3             2A             2.0            12.0            D19_UNBEKANNT   \n",
       "4             6B             6.0            43.0               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": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find categorical features in azdias\n",
    "azdias.select_dtypes(['object']).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "86fa45ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "float64    267\n",
       "int64       94\n",
       "object       8\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find data types for customers\n",
    "customers.dtypes.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f5daeb75",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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",
       "      <th>PRODUCT_GROUP</th>\n",
       "      <th>CUSTOMER_GROUP</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1A</td>\n",
       "      <td>1.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "      <td>W</td>\n",
       "      <td>COSMETIC_AND_FOOD</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>D19_BANKEN_GROSS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>FOOD</td>\n",
       "      <td>SINGLE_BUYER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5D</td>\n",
       "      <td>5.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>D19_UNBEKANNT</td>\n",
       "      <td>1992-02-10 00:00:00</td>\n",
       "      <td>W</td>\n",
       "      <td>COSMETIC_AND_FOOD</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4C</td>\n",
       "      <td>4.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>D19_NAHRUNGSERGAENZUNG</td>\n",
       "      <td>1992-02-10 00:00:00</td>\n",
       "      <td>W</td>\n",
       "      <td>COSMETIC</td>\n",
       "      <td>MULTI_BUYER</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7B</td>\n",
       "      <td>7.0</td>\n",
       "      <td>41.0</td>\n",
       "      <td>D19_SCHUHE</td>\n",
       "      <td>1992-02-12 00:00:00</td>\n",
       "      <td>W</td>\n",
       "      <td>FOOD</td>\n",
       "      <td>MULTI_BUYER</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             1A             1.0            13.0            D19_UNBEKANNT   \n",
       "1            NaN             NaN             NaN         D19_BANKEN_GROSS   \n",
       "2             5D             5.0            34.0            D19_UNBEKANNT   \n",
       "3             4C             4.0            24.0   D19_NAHRUNGSERGAENZUNG   \n",
       "4             7B             7.0            41.0               D19_SCHUHE   \n",
       "\n",
       "         EINGEFUEGT_AM OST_WEST_KZ      PRODUCT_GROUP CUSTOMER_GROUP  \n",
       "0  1992-02-12 00:00:00           W  COSMETIC_AND_FOOD    MULTI_BUYER  \n",
       "1                  NaN         NaN               FOOD   SINGLE_BUYER  \n",
       "2  1992-02-10 00:00:00           W  COSMETIC_AND_FOOD    MULTI_BUYER  \n",
       "3  1992-02-10 00:00:00           W           COSMETIC    MULTI_BUYER  \n",
       "4  1992-02-12 00:00:00           W               FOOD    MULTI_BUYER  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find categorical features in customers\n",
    "customers.select_dtypes(['object']).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc061937",
   "metadata": {},
   "source": [
    "#### We can see here that the 6 categorical features in both datasets are common. The customers data has two more columns which are not in the azdias data."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "149ae9fd",
   "metadata": {},
   "source": [
    "#### Earlier while loading the data, there was a warning indiacting some columns store in mixed-types. So, before diving deeper into the analysis of the datasets, lets validate these columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b17c7a87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1.0, nan, 5.0, 4.0, 7.0, 3.0, 9.0, 2.0, 6.0, 8.0, '6', '3', '8',\n",
       "       '9', '2', '4', '1', '7', '5', 'X'], dtype=object)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check unique values for feature CAMEO_DEU_2015\n",
    "customers['CAMEO_DEUG_2015'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f43168a7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CAMEO_DEUG_2015\n",
       "1.0     4280\n",
       "2.0     5910\n",
       "3.0     4805\n",
       "4.0     5606\n",
       "5.0     3042\n",
       "6.0     4709\n",
       "7.0     2680\n",
       "8.0     3333\n",
       "9.0     1661\n",
       "1      12498\n",
       "2      17574\n",
       "3      13585\n",
       "4      16458\n",
       "5       8624\n",
       "6      14008\n",
       "7       7878\n",
       "8       9716\n",
       "9       4731\n",
       "X        126\n",
       "Name: CAMEO_DEUG_2015, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the count for feature\n",
    "customers.groupby(\"CAMEO_DEUG_2015\")[\"CAMEO_DEUG_2015\"].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0e33ce3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([13.0, nan, 34.0, 24.0, 41.0, 23.0, 15.0, 55.0, 14.0, 22.0, 43.0,\n",
       "       51.0, 33.0, 25.0, 44.0, 54.0, 32.0, 12.0, 35.0, 31.0, 45.0, 52.0,\n",
       "       '45', '25', '55', '51', '14', '54', '43', '22', '15', '24', '35',\n",
       "       '23', '12', '44', '41', '52', '31', '13', '34', '32', '33', 'XX'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check unique values for feature CAMEO_INTL_2015\n",
    "customers['CAMEO_INTL_2015'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "54c5ed4e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CAMEO_INTL_2015\n",
       "12.0      924\n",
       "13.0     1955\n",
       "14.0     4939\n",
       "15.0     2372\n",
       "22.0     1941\n",
       "23.0     1494\n",
       "24.0     4504\n",
       "25.0     2472\n",
       "31.0      681\n",
       "32.0      440\n",
       "33.0      396\n",
       "34.0     1423\n",
       "35.0      553\n",
       "41.0     2859\n",
       "43.0     2476\n",
       "44.0      688\n",
       "45.0     1352\n",
       "51.0     2126\n",
       "52.0      253\n",
       "54.0     1258\n",
       "55.0      920\n",
       "12       2791\n",
       "13       5728\n",
       "14      14708\n",
       "15       6845\n",
       "22       5566\n",
       "23       4276\n",
       "24      13301\n",
       "25       6900\n",
       "31       2050\n",
       "32       1256\n",
       "33       1178\n",
       "34       3945\n",
       "35       1741\n",
       "41       8461\n",
       "43       7158\n",
       "44       2144\n",
       "45       3936\n",
       "51       5987\n",
       "52        770\n",
       "54       3537\n",
       "55       2794\n",
       "XX        126\n",
       "Name: CAMEO_INTL_2015, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the count for feature\n",
    "customers.groupby(\"CAMEO_INTL_2015\")[\"CAMEO_INTL_2015\"].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b7c7ed1",
   "metadata": {},
   "source": [
    "#### We can see that the problem of mixed types in columns is due to a mix of float, int and string values. Similar situation will be there with azdias dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca5e4573",
   "metadata": {},
   "source": [
    "#### Let's explore excel spreadsheets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "52852c11",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load the excel data\n",
    "attributes_values = pd.read_excel('E:\\\\udct\\\\Arvato Project\\\\DIAS Attributes - Values 2017.xlsx').drop('Unnamed: 0', axis=1)\n",
    "attributes_info = pd.read_excel('E:\\\\udct\\\\Arvato Project\\\\DIAS Information Levels - Attributes 2017.xlsx').drop('Unnamed: 0', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "26931d06",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the first row as header\n",
    "attributes_info = attributes_info.rename(columns=attributes_info.iloc[0]).drop(attributes_info.index[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "05c814c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(313, 4) \n",
      "\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>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>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>in cooperation with Kantar TNS; the informatio...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Person</td>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>age through prename analysis</td>\n",
       "      <td>modelled on millions of first name-age-referen...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>ANREDE_KZ</td>\n",
       "      <td>gender</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>CJT_GESAMTTYP</td>\n",
       "      <td>Customer-Journey-Typology relating to the pref...</td>\n",
       "      <td>relating to the preferred information, marketi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_MINIMALIST</td>\n",
       "      <td>financial typology: low financial interest</td>\n",
       "      <td>Gfk-Typology based on a representative househo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_SPARER</td>\n",
       "      <td>financial typology: money saver</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_VORSORGER</td>\n",
       "      <td>financial typology: be prepared</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_ANLEGER</td>\n",
       "      <td>financial typology: investor</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_UNAUFFAELLIGER</td>\n",
       "      <td>financial typology: unremarkable</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NaN</td>\n",
       "      <td>FINANZ_HAUSBAUER</td>\n",
       "      <td>financial typology: main focus is the own house</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Information level              Attribute  \\\n",
       "1                NaN               AGER_TYP   \n",
       "2             Person   ALTERSKATEGORIE_GROB   \n",
       "3                NaN              ANREDE_KZ   \n",
       "4                NaN          CJT_GESAMTTYP   \n",
       "5                NaN      FINANZ_MINIMALIST   \n",
       "6                NaN          FINANZ_SPARER   \n",
       "7                NaN       FINANZ_VORSORGER   \n",
       "8                NaN         FINANZ_ANLEGER   \n",
       "9                NaN  FINANZ_UNAUFFAELLIGER   \n",
       "10               NaN       FINANZ_HAUSBAUER   \n",
       "\n",
       "                                          Description  \\\n",
       "1                                  best-ager typology   \n",
       "2                       age through prename analysis    \n",
       "3                                              gender   \n",
       "4   Customer-Journey-Typology relating to the pref...   \n",
       "5          financial typology: low financial interest   \n",
       "6                     financial typology: money saver   \n",
       "7                     financial typology: be prepared   \n",
       "8                        financial typology: investor   \n",
       "9                    financial typology: unremarkable   \n",
       "10    financial typology: main focus is the own house   \n",
       "\n",
       "                                     Additional notes  \n",
       "1   in cooperation with Kantar TNS; the informatio...  \n",
       "2   modelled on millions of first name-age-referen...  \n",
       "3                                                 NaN  \n",
       "4   relating to the preferred information, marketi...  \n",
       "5   Gfk-Typology based on a representative househo...  \n",
       "6                                                 NaN  \n",
       "7                                                 NaN  \n",
       "8                                                 NaN  \n",
       "9                                                 NaN  \n",
       "10                                                NaN  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of attributes_info dataset\n",
    "print(attributes_info.shape, \"\\n\")\n",
    "attributes_info.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8005519f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the first row as header\n",
    "attributes_values = attributes_values.rename(columns=attributes_values.iloc[0]).drop(attributes_values.index[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "73eaa2ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2258, 4) \n",
      "\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>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>-1</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>no classification possible</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>passive elderly</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>cultural elderly</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>experience-driven elderly</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ALTERSKATEGORIE_GROB</td>\n",
       "      <td>age classification through prename analysis</td>\n",
       "      <td>-1, 0</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>&lt; 30 years</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>30 - 45 years</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>46 - 60 years</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>&gt; 60 years</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Attribute                                   Description  Value  \\\n",
       "1               AGER_TYP                            best-ager typology     -1   \n",
       "2                    NaN                                           NaN      0   \n",
       "3                    NaN                                           NaN      1   \n",
       "4                    NaN                                           NaN      2   \n",
       "5                    NaN                                           NaN      3   \n",
       "6   ALTERSKATEGORIE_GROB  age classification through prename analysis   -1, 0   \n",
       "7                    NaN                                           NaN      1   \n",
       "8                    NaN                                           NaN      2   \n",
       "9                    NaN                                           NaN      3   \n",
       "10                   NaN                                           NaN      4   \n",
       "\n",
       "                       Meaning  \n",
       "1                      unknown  \n",
       "2   no classification possible  \n",
       "3              passive elderly  \n",
       "4             cultural elderly  \n",
       "5    experience-driven elderly  \n",
       "6                      unknown  \n",
       "7                   < 30 years  \n",
       "8                30 - 45 years  \n",
       "9                46 - 60 years  \n",
       "10                  > 60 years  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of attributes_values dataset\n",
    "print(attributes_values.shape, \"\\n\")\n",
    "attributes_values.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "3915da7a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Attribute</th>\n",
       "      <th>Description</th>\n",
       "      <th>Value</th>\n",
       "      <th>Meaning</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>-1</td>\n",
       "      <td>unknown</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>0</td>\n",
       "      <td>no classification possible</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>1</td>\n",
       "      <td>passive elderly</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>2</td>\n",
       "      <td>cultural elderly</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>best-ager typology</td>\n",
       "      <td>3</td>\n",
       "      <td>experience-driven elderly</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Attribute         Description Value                     Meaning\n",
       "1  AGER_TYP  best-ager typology    -1                     unknown\n",
       "2  AGER_TYP  best-ager typology     0  no classification possible\n",
       "3  AGER_TYP  best-ager typology     1             passive elderly\n",
       "4  AGER_TYP  best-ager typology     2            cultural elderly\n",
       "5  AGER_TYP  best-ager typology     3   experience-driven elderly"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Forward fill the data\n",
    "attributes_values['Attribute'] = attributes_values['Attribute'].ffill()\n",
    "attributes_values['Description'] = attributes_values['Description'].ffill()\n",
    "attributes_values.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "7fdb4938",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No of columns in CUSTOMESRS:  369\n",
      "No of columns in AZDIAS:  366\n",
      "No of columns in Attributes info:  313\n",
      "No of columns in Attributes values:  314\n",
      "No of columns in common:  264\n"
     ]
    }
   ],
   "source": [
    "# Find common info between all datasets\n",
    "customers_unique = set(list(customers.columns))\n",
    "azdias_unique = set(list(azdias.columns))\n",
    "attributes_info_unique = set(list(attributes_info.Attribute))\n",
    "attributes_values_unique = set(list(attributes_values.Attribute))\n",
    "\n",
    "print('No of columns in CUSTOMESRS: ', len(customers_unique))\n",
    "print('No of columns in AZDIAS: ', len(azdias_unique))\n",
    "print('No of columns in Attributes info: ', len(attributes_info_unique))\n",
    "print('No of columns in Attributes values: ', len(attributes_values_unique))\n",
    "\n",
    "common = customers_unique & azdias_unique & attributes_info_unique\n",
    "print('No of columns in common: ', len(common))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b388a1c5",
   "metadata": {},
   "source": [
    "#### After performing a basic exploration of all datasets, we've found the following problems:\n",
    "1. A lot of missing data which is already present in the azdias and customers data\n",
    "2. A mixed data types with X and XX values in columns 18 and 19\n",
    "3. Customers data having 3 extra features\n",
    "4. Unknowns in excel spreadsheet with corresponding encoding for the column Value"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce8ccc06",
   "metadata": {},
   "source": [
    "#### Lets handle these problems one by one"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "867a69b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make a copy of datasets\n",
    "customers_new = customers.copy()\n",
    "azdias_new = azdias.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "1dcbbb54",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to fix mixed data types problem\n",
    "def fix_cameo(df):\n",
    "    '''\n",
    "    This function performs fixing of mixed data type issue.\n",
    "    \n",
    "    Input:\n",
    "    df -  The dataframe to perform cleaning operation\n",
    "    \n",
    "    Output:\n",
    "    df - The transformed dataframe\n",
    "    '''\n",
    "    \n",
    "    cols = [\"CAMEO_DEUG_2015\", \"CAMEO_INTL_2015\"]\n",
    "    df[cols] = df[cols].replace({\"X\": np.nan, \"XX\": np.nan})\n",
    "    df[cols] = df[cols].astype(float)\n",
    "\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "396aef51",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use the function on datasets\n",
    "customers_new = fix_cameo(customers_new)\n",
    "azdias_new = fix_cameo(azdias_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "16970629",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1., nan,  5.,  4.,  7.,  3.,  9.,  2.,  6.,  8.])"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify the fix\n",
    "customers_new['CAMEO_DEUG_2015'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "b424f97a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CAMEO_DEUG_2015\n",
       "1.0    16778\n",
       "2.0    23484\n",
       "3.0    18390\n",
       "4.0    22064\n",
       "5.0    11666\n",
       "6.0    18717\n",
       "7.0    10558\n",
       "8.0    13049\n",
       "9.0     6392\n",
       "Name: CAMEO_DEUG_2015, dtype: int64"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify the fix\n",
    "customers_new.groupby(\"CAMEO_DEUG_2015\")[\"CAMEO_DEUG_2015\"].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "7c473fd4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([13., nan, 34., 24., 41., 23., 15., 55., 14., 22., 43., 51., 33.,\n",
       "       25., 44., 54., 32., 12., 35., 31., 45., 52.])"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify the fix\n",
    "customers_new['CAMEO_INTL_2015'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "8b7c47d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CAMEO_INTL_2015\n",
       "12.0     3715\n",
       "13.0     7683\n",
       "14.0    19647\n",
       "15.0     9217\n",
       "22.0     7507\n",
       "23.0     5770\n",
       "24.0    17805\n",
       "25.0     9372\n",
       "31.0     2731\n",
       "32.0     1696\n",
       "33.0     1574\n",
       "34.0     5368\n",
       "35.0     2294\n",
       "41.0    11320\n",
       "43.0     9634\n",
       "44.0     2832\n",
       "45.0     5288\n",
       "51.0     8113\n",
       "52.0     1023\n",
       "54.0     4795\n",
       "55.0     3714\n",
       "Name: CAMEO_INTL_2015, dtype: int64"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verify the fix\n",
    "customers_new.groupby(\"CAMEO_INTL_2015\")[\"CAMEO_INTL_2015\"].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "9f887857",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove extra columns\n",
    "customers_new = customers_new.drop(columns=['CUSTOMER_GROUP', 'ONLINE_PURCHASE', 'PRODUCT_GROUP'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "bc555352",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dictionary to store unknowns\n",
    "dict_unknown = {}\n",
    "for i, row in attributes_values.iterrows():\n",
    "    if \"unknown\" in str(row['Meaning']):\n",
    "        dict_unknown[row['Attribute']] = [int(num) for num in str(row['Value']).split(', ')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "8cdf18be",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to replace unknown with np.nan\n",
    "def replace_unknown(df, dictionary):\n",
    "    '''\n",
    "    This function replaces unknowns present in the dictionary.\n",
    "    \n",
    "    Input:\n",
    "    df -  The dataframe to perform cleaning operation\n",
    "    dictionary - Contains encodings for unknown values\n",
    "    \n",
    "    Output:\n",
    "    None\n",
    "    '''\n",
    "    for key, value in dictionary.items():\n",
    "        if key in df.columns:\n",
    "            for i in range(0, len(value)):\n",
    "                df.loc[df[key] == value[i], key] = np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "b97bca6f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Apply the function\n",
    "replace_unknown(customers_new, dict_unknown)\n",
    "replace_unknown(azdias_new, dict_unknown)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "7cc439a8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Before replacing the NaN values in customers:  13864522\n",
      "After replacing the NaN values in customers:  14488721\n",
      "Before replacing the NaN values in azdias:  33492923\n",
      "After replacing the NaN values in azdias:  37088263\n"
     ]
    }
   ],
   "source": [
    "# verify the fix\n",
    "print('Before replacing the NaN values in customers: ',customers.isna().sum().sum())\n",
    "print('After replacing the NaN values in customers: ', customers_new.isna().sum().sum())\n",
    "\n",
    "print('Before replacing the NaN values in azdias: ', azdias.isna().sum().sum())\n",
    "print('After replacing the NaN values in azdias: ', azdias_new.isna().sum().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "328f0c31",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to find out maximul missing value\n",
    "def find_max_missing_features(df):\n",
    "    \"\"\"\n",
    "    This function will find the features having maximum percentage of missing values\n",
    "    \n",
    "    Input:\n",
    "    df - Dataframe to find the features for missing values\n",
    "    \n",
    "    Output:\n",
    "    features - Features with maximum missing values\n",
    "    \"\"\"\n",
    "    features = pd.DataFrame(columns=['Attribute', 'Missing_Percentage'])\n",
    "\n",
    "    for col in df.columns:\n",
    "        vals = int(len(df[col].unique()))\n",
    "        percentage = round((np.sum(df[col].isna()) / df.shape[0])*100, 2)\n",
    "        features.loc[col] = [vals, percentage]\n",
    "    \n",
    "    features = features.sort_values(['Missing_Percentage', 'Attribute'], ascending=False)\n",
    "    \n",
    "    return features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "09feea52",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Attribute</th>\n",
       "      <th>Missing_Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <td>11.0</td>\n",
       "      <td>99.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <td>15.0</td>\n",
       "      <td>99.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TITEL_KZ</th>\n",
       "      <td>5.0</td>\n",
       "      <td>98.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <td>18.0</td>\n",
       "      <td>97.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <td>18.0</td>\n",
       "      <td>93.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KK_KUNDENTYP</th>\n",
       "      <td>7.0</td>\n",
       "      <td>58.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KBA05_BAUMAX</th>\n",
       "      <td>6.0</td>\n",
       "      <td>57.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGER_TYP</th>\n",
       "      <td>5.0</td>\n",
       "      <td>48.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EXTSEL992</th>\n",
       "      <td>57.0</td>\n",
       "      <td>44.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_HH</th>\n",
       "      <td>21.0</td>\n",
       "      <td>35.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>REGIOTYP</th>\n",
       "      <td>8.0</td>\n",
       "      <td>31.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KKK</th>\n",
       "      <td>5.0</td>\n",
       "      <td>31.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>W_KEIT_KIND_HH</th>\n",
       "      <td>7.0</td>\n",
       "      <td>29.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KBA05_ALTER4</th>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KBA05_HERST1</th>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Attribute  Missing_Percentage\n",
       "ALTER_KIND4          11.0               99.88\n",
       "ALTER_KIND3          15.0               99.33\n",
       "TITEL_KZ              5.0               98.79\n",
       "ALTER_KIND2          18.0               97.34\n",
       "ALTER_KIND1          18.0               93.86\n",
       "KK_KUNDENTYP          7.0               58.41\n",
       "KBA05_BAUMAX          6.0               57.15\n",
       "AGER_TYP              5.0               48.06\n",
       "EXTSEL992            57.0               44.50\n",
       "ALTER_HH             21.0               35.87\n",
       "REGIOTYP              8.0               31.34\n",
       "KKK                   5.0               31.34\n",
       "W_KEIT_KIND_HH        7.0               29.71\n",
       "KBA05_ALTER4          7.0               29.70\n",
       "KBA05_HERST1          7.0               29.70"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply function on datasets\n",
    "feature_customers = find_max_missing_features(customers_new)\n",
    "feature_customers.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "2f9e4495",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot a graph\n",
    "feature_customers[feature_customers.Missing_Percentage > 30].Missing_Percentage.plot(kind='bar');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "00b6a603",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Attribute</th>\n",
       "      <th>Missing_Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND4</th>\n",
       "      <td>13.0</td>\n",
       "      <td>99.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TITEL_KZ</th>\n",
       "      <td>6.0</td>\n",
       "      <td>99.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND3</th>\n",
       "      <td>16.0</td>\n",
       "      <td>99.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND2</th>\n",
       "      <td>18.0</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_KIND1</th>\n",
       "      <td>18.0</td>\n",
       "      <td>90.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AGER_TYP</th>\n",
       "      <td>5.0</td>\n",
       "      <td>76.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EXTSEL992</th>\n",
       "      <td>57.0</td>\n",
       "      <td>73.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KK_KUNDENTYP</th>\n",
       "      <td>7.0</td>\n",
       "      <td>65.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KBA05_BAUMAX</th>\n",
       "      <td>6.0</td>\n",
       "      <td>53.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTER_HH</th>\n",
       "      <td>22.0</td>\n",
       "      <td>34.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ALTERSKATEGORIE_FEIN</th>\n",
       "      <td>27.0</td>\n",
       "      <td>29.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D19_LETZTER_KAUF_BRANCHE</th>\n",
       "      <td>36.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D19_BANKEN_ONLINE_QUOTE_12</th>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D19_GESAMT_ONLINE_QUOTE_12</th>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>D19_VERSAND_ONLINE_QUOTE_12</th>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             Attribute  Missing_Percentage\n",
       "ALTER_KIND4                       13.0               99.86\n",
       "TITEL_KZ                           6.0               99.76\n",
       "ALTER_KIND3                       16.0               99.31\n",
       "ALTER_KIND2                       18.0               96.69\n",
       "ALTER_KIND1                       18.0               90.90\n",
       "AGER_TYP                           5.0               76.02\n",
       "EXTSEL992                         57.0               73.40\n",
       "KK_KUNDENTYP                       7.0               65.60\n",
       "KBA05_BAUMAX                       6.0               53.47\n",
       "ALTER_HH                          22.0               34.81\n",
       "ALTERSKATEGORIE_FEIN              27.0               29.50\n",
       "D19_LETZTER_KAUF_BRANCHE          36.0               28.85\n",
       "D19_BANKEN_ONLINE_QUOTE_12        12.0               28.85\n",
       "D19_GESAMT_ONLINE_QUOTE_12        12.0               28.85\n",
       "D19_VERSAND_ONLINE_QUOTE_12       12.0               28.85"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply function on datasets\n",
    "feature_azdias = find_max_missing_features(azdias_new)\n",
    "feature_azdias.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "6f6d4330",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAFACAYAAABHvzzrAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAhd0lEQVR4nO3debhkVX3u8e9LMzlAgEtDCIMtBgcwTmnEKSpinFAhMUxG00lI0KBRNF6DXBM0CXlInMfEFhP7BocQEMExYAuaxLERUBEQL7O20BgTFeMAvPePvQ9dVNc5ffpU7VWrNu/nec5zau99Tq1fr1P9q1Vrr0G2iYiIftlq2gFERMTkJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T00NbTDgBg11139YoVK6YdRkTETLnoootusb181LUqkvuKFStYt27dtMOIiJgpkq6b71q6ZSIieijJPSKih5LcIyJ6KMk9IqKHNpvcJf2DpJslfX3g3C6Szpd0Vft954Frr5L0LUlXSnpqV4FHRMT8FtNyfy/wtKFzJwJrbe8HrG2PkbQ/cDRwQPs775S0bGLRRkTEomw2udv+LPCfQ6cPA9a0j9cAhw+c/6Dtn9q+BvgW8MjJhBoREYu11D733W2vB2i/79ae3xO4YeDnbmzPRUREQZOexKQR50buBiLpOOA4gH322WezT7zixI+NFdi1px461u9PIoZJxRERsTlLTe43SdrD9npJewA3t+dvBPYe+Lm9gO+MegLbq4HVACtXrsx2UFughje6iKjbUrtlzgVWtY9XAecMnD9a0naS7gvsB3xpvBAjImJLbbblLukDwBOBXSXdCJwMnAqcIelY4HrgCADbl0k6A/gGcBvwItu3dxR7RETMY7PJ3fYx81w6ZJ6fPwU4ZZygIiJiPJmhGhHRQ0nuERE9VMV67jF7Miw0om5puUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQ1paJmZZdqSJGS8s9IqKHktwjInooyT0iooeS3CMieijJPSKih5LcIyJ6KMk9IqKHktwjInooyT0iooeS3CMieijJPSKih5LcIyJ6KMk9IqKHktwjInooyT0iooeS3CMieijJPSKih5LcIyJ6aKzkLullki6T9HVJH5C0vaRdJJ0v6ar2+86TCjYiIhZnycld0p7AS4CVth8MLAOOBk4E1treD1jbHkdEREHjdstsDdxD0tbAPYHvAIcBa9rra4DDxywjIiK20JKTu+1vA68HrgfWA/9t+zxgd9vr259ZD+w26vclHSdpnaR1GzZsWGoYERExwjjdMjvTtNLvC/wScC9Jz1vs79tebXul7ZXLly9fahgRETHCON0yTwausb3B9s+BDwGPAW6StAdA+/3m8cOMiIgtMU5yvx54lKR7ShJwCHA5cC6wqv2ZVcA544UYERFbauul/qLtL0o6E/gKcBtwMbAauDdwhqRjad4AjphEoBERsXhLTu4Atk8GTh46/VOaVnxERExJZqhGRPRQkntERA8luUdE9NBYfe4RAStO/NjYz3HtqYdOIJKIjdJyj4jooST3iIgeSnKPiOihJPeIiB7KDdWInhj3xm5u6vZLWu4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZT13CNiYrJZeD3Sco+I6KEk94iIHkpyj4jooST3iIgeSnKPiOihJPeIiB5Kco+I6KGxkruknSSdKekKSZdLerSkXSSdL+mq9vvOkwo2IiIWZ9yW+1uAT9p+IPBQ4HLgRGCt7f2Ate1xREQUtOTkLmlH4PHAewBs/8z2fwGHAWvaH1sDHD5eiBERsaXGabnvC2wA/lHSxZJOk3QvYHfb6wHa77tNIM6IiNgC4yT3rYFHAH9n++HArWxBF4yk4yStk7Ruw4YNY4QRERHDxknuNwI32v5ie3wmTbK/SdIeAO33m0f9su3VtlfaXrl8+fIxwoiIiGFLTu62vwvcIOkB7alDgG8A5wKr2nOrgHPGijAiIrbYuEv+/jHwPknbAlcDv0fzhnGGpGOB64EjxiwjIiK20FjJ3fYlwMoRlw4Z53kjImI8maEaEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET007gzViIjqrDjxY2P9/rWnHjqhSKYnLfeIiB5Kco+I6KEk94iIHkpyj4jooST3iIgeSnKPiOihJPeIiB5Kco+I6KEk94iIHkpyj4jooST3iIgeSnKPiOihJPeIiB5Kco+I6KEk94iIHkpyj4jooST3iIgeSnKPiOihJPeIiB5Kco+I6KEk94iIHkpyj4jooST3iIgeSnKPiOihsZO7pGWSLpb00fZ4F0nnS7qq/b7z+GFGRMSWmETL/aXA5QPHJwJrbe8HrG2PIyKioLGSu6S9gEOB0wZOHwasaR+vAQ4fp4yIiNhy47bc3wy8Erhj4NzuttcDtN93G/WLko6TtE7Sug0bNowZRkREDFpycpf0TOBm2xct5fdtr7a90vbK5cuXLzWMiIgYYesxfvexwLMlPQPYHthR0unATZL2sL1e0h7AzZMINCIiFm/JLXfbr7K9l+0VwNHAp20/DzgXWNX+2CrgnLGjjIiILdLFOPdTgV+XdBXw6+1xREQUNE63zJ1sXwhc2D7+HnDIJJ43IiKWJjNUIyJ6KMk9IqKHktwjInooyT0iooeS3CMiemgio2UiIuKuVpz4sbGf49pTD13y76blHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9lOQeEdFDSe4RET2U5B4R0UNJ7hERPZTkHhHRQ0nuERE9tOTkLmlvSRdIulzSZZJe2p7fRdL5kq5qv+88uXAjImIxxmm53wb8ie0HAY8CXiRpf+BEYK3t/YC17XFERBS05ORue73tr7SPfwhcDuwJHAasaX9sDXD4mDFGRMQWmkifu6QVwMOBLwK7214PzRsAsNskyoiIiMUbO7lLujdwFnCC7R9swe8dJ2mdpHUbNmwYN4yIiBgwVnKXtA1NYn+f7Q+1p2+StEd7fQ/g5lG/a3u17ZW2Vy5fvnycMCIiYsg4o2UEvAe43PYbBy6dC6xqH68Czll6eBERsRRbj/G7jwWeD3xN0iXtuZOAU4EzJB0LXA8cMVaEERGxxZac3G3/O6B5Lh+y1OeNiIjxZYZqREQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA8luUdE9FCSe0REDyW5R0T0UJJ7REQPJblHRPRQkntERA91ltwlPU3SlZK+JenErsqJiIhNdZLcJS0D3gE8HdgfOEbS/l2UFRERm+qq5f5I4Fu2r7b9M+CDwGEdlRUREUNke/JPKv0W8DTbf9AePx84yPaLB37mOOC49vABwJVjFrsrcMuYzzEJNcRRQwxQRxyJYaMa4qghBqgjjknEcB/by0dd2HrMJ56PRpy7y7uI7dXA6okVKK2zvXJSzzfLcdQQQy1xJIa64qghhlri6DqGrrplbgT2HjjeC/hOR2VFRMSQrpL7l4H9JN1X0rbA0cC5HZUVERFDOumWsX2bpBcD/wosA/7B9mVdlDVgYl08Y6ohjhpigDriSAwb1RBHDTFAHXF0GkMnN1QjImK6MkM1IqKHktwjInooyT0iooeS3JdI0gsk3WOeay8tHU/UQ9JySSsl7TTtWGpQc31I6mquz9T1JrlLOr5wkW8GLpL0sBHXVpUMRNJWkrZqH28r6RGSdikZQ1v2NiPO7Vo6jlEkPbBQOX8AXAa8DbhC0rNLlLtAPFP9m9RQH5I+Iuk+I84/GbikUAyPWOirizJn8l1L0suHTwGvkrQ9gO03FgjjCuCVwIclvd3264fiKULS4cC7gDskvRA4CbgVuL+kP7L9kQIxHAz8E7CdpIuB42xf214+D+jkxbuFzgP2KVDOCcABtjdI2hd4H1OY41HR3+QEpl8fHwQukPQe4G+B5TSNs30o1xB7w8DjXwXWsTFPGHjSpAucyeQOvBb4OE2LYK6ClgE7FIzBts+XtBI4TdLTgefZXs/QUgsdOxl4KHAP4FLgQNtXti2Vs4DOkzvNf5in2r6sXVfofEnPt/0Fyr7RvXW+S8BOhcL4me0NALavlrRdoXKHVfE3oYL6sP0+SR+lqZPLgW2AU4B3u9BYcNsHzz2WdLHtiSfzYbOa3A8A3gjcC3it7R9LWmX7taUDsX0LcHjbav6ipJdNIYbvAki63vaV7bnr5rpqCth2bpKa7TMlXQ58qF3Hv+Qb3e8BfwL8dMS1YwrFsNfQm8xdjm2/pFActfxNaqmP/WlWq/0SsBLYnSb//bxQ+YOK1P9MJnfb1wO/JekwmhbJm6YQxl1aP7b/XtKFNB87DygaiLSV7TuA3x84twzYtlAIP5f0i3NvMm1r8RDgo8D9CsUAzbIXX7f9ueELkl5TKIb/PXR8UaFyh9XyN5l6fUg6jaYb6njbn5d0L5pP/5dKOsH2eaVjKmHmZ6hKuifNH+og248vWO4jbX9pxPltgCNsv79QHAcCX7P9k6HzK4DH2T69QAxPBjbYvnTo/E7Ai2yf0nUMbXm7AD+x/eMS5c0Tw5G2z5hW+QNx1PI3mXp9tJ+m32r79qHzvwK8o0TekPQ2NrbYj6a5D3CnLj7B9CG5LweY69crWO5pc+vVD53fC/ik7QeXjKcteyp1UVsM04yj7dvdmqaVeHXJsmtUc320AzCeXeLNR9KCN25tr5l4obZn7oumS+Q1NAvd/2f7tQH484IxvBc4Hdhq4Nz+wDXAqinVxfeA75euizaOkyuIoZa6OJzmxt2f0WzIsMvcV8EY7g38Bc2gg/9u6+ELwO+WrIta6mMglmU023/+X+Am4MzSMZT6msmWe/sx6xk0w7uuac/tC/wdTau58z54SaIZgrgzzcesg4B/Bl5o+2Ndlz8QRw11MfUYaoqjLfehwGdp3mDm/pPZ9r6Fyj8HOBv4FHAkzeCDDwKvBr5t+6QScQzEM+36eDzwXOBQmpuqjwX2daEuPEkfYYEbqbYnP/5/2u8uS3z3vRjYdcT55cDFhWN5C/BvwPXAo+6OdVFDDLXEAWwH/CVNS/WZpV8PA3FcOnT85fb7VsAVBeOYen3QbB70OeD5wA7tuWsKx/CE9uuJwFUDx08AntBFmTM5WgbYxs0QxLtwM1Fikxl5XRi4QSKa7pivAM+V9Nw2llJDvKZeF5XEUEscX6WZX/AI2/9TqMxRbpX0ONv/LulZNF2X2L6j/dRZSg31cRZN19BRwO3tp5qiXRa2PzP3WNKPBo+7MqvJ/WdLvDZJ6+Z5XFoNdVFDDJsrq1Qcl7pwl8c8Xkgzue4BwNeAY+HOG83vKBjH1OvD9kslnQAcTDPf4XXAjpKOBD5u+0elQypRyKz2ud9OM8V+k0vA9rZLthYXJOlttv+4w+efel3UEEMtcUj6iu0alluoQo310X6KezrNvbKn2O58rR3dda2nC2i6Z+78BGX7Pyde5iwm91lS44s7uiPpUob+4w7q4j/xArHcD/gNms3qb6Pp6/2A7f8uGEM19TGKpHuU6C6SdA0bu3GH2R3cWJ7VbpmIWj2QZhbmyP/EQKnRIS8BngV8BjiQZvXDvYHPSzre9oUl4qCC+pD01c38yEO6jsH2fRfzc5IO8IT2m57JlrukH7Lpu6Bp3qy2tV3Nm1bXLfca6qKGGGqJo10U6uFdl7OIOL4GPMz27e0s7o/bfqKkfYBzSsVYQ31IuoTmdfB+moX07tJSt33dFMIaaZL5opokuCVs32X1R0k7AMcDL6AZ21uTTkcm1FAXNcRQUxwV2Rq4nWY44g7QrMtUeATT1Nl+mJr1/I+hSfDfaL+fZ/u2qQa3qYnli5nerEPSTu2CUJfSvHgPtP0n040KJA2u7f6WQmVOvS5qiKGCOBb1926H0nbpNODLklYDnwfe3pa7nHZYZCFFXv+bY/sK2ye3reKP0MxQLb6C6yJMrCtlVrtldqVZ2vUo4B+At5W8SbQ5apbeLbExRBV1UUMMNcWxGCVutEs6AHgQzUqZV3RZ1gIx/CPzJyzbPrZQHHvSjI75DZpZsmcAZ09hGOSCJvm6mNXkfivNWhn/CPxw+LrL7MQ0L0k32N67UFlTr4saYqgpjsWY5iiqwo2P54w4vQ/NDk3LbO9VIIbP0HyCOwM4k6FPLtMesTNI0hdsP2oSzzWTfe40kxDm3pVK7r50J82/R6nouJ99yNTropIYaoqjdsVen7bPurPQZp2fk4DHA6cC7ykUxn1oXhcvAI4bOC8KjmAapZ1k9grbfwgwqcQOM9pyr8Fmxq0ueuhT3D1NcxRJyZZ7W96DgP8DPJzmDfj0Cm9kdkrSQ4DXA78EfJhmw/B30iw4+AZ3sKjdTLbchxa+30SJdV1qSd411EUNMdQUx3wk3Wdg2F2nNxq16Sbyd16iWQ64CEn/QrOt3etpbmDeTjP1HyjTJSJpuPvLwC22b+i67AHvplmd9PPA02jWono/8Nse2mhnUmYyuTPdtVwAkPQ8t7scSXqs7f8YuPZi228vFMrU64I6YoBK4pD0aGBP4LO2b25bbScCv0YzkQjb7+04jIW6pUqOYDmQJpm+guZmN2z8tFuqS+QNI87tImlb4BjblxSIYbuBv/mVkl4BnOih3aEmaVaT+yUe2j5sjqQ/KhTDy2k264DmI9Zg6+D3aYeeFVBDXdQQQxVxSHod8EyaGaF/qmYnouOBv2Zgj9uueQqbxY9ie0UFMRw86ryklcBbae4BdG17SQ9n4xvbj4CHzK3Qafsrky5wVse5ny3pV4dPSnot8IeFYtA8j0cdd6mGuqghhlriOBR4uO1jgKfQtNgfZ/stXX38HkXSGQOP/2boWrENoSU9b+DxY4euvbhUHKPYXke5LqrvAm+k+RTxhqHj1y/we0vnKSyeP+4X8KvA1cCj22MBfw98GtixUAxfGfV41PHdoC6mHkMtcQAXDR1fUurfP1TuxQOPh1+fF5eIYbjsaf4/mSe23Yf/Xn36msluGdsXSTqcpqX2Ija2yp5mu9S63Q9sFyQScL+BxYlEwaFVNdRFDTFUFMf9JJ07cLxi8NhdbKc22kLD4EoOkZv6J9x5brTvAjwGeGmhGF5p+2/bx0fY/peBa3/tDta8n8mhkANjzPenGVb0KeDFwB1Q7A78fRa67kKLEVVSF1OPoZY4JD1hoesusANPG8cVNGupbEVzb+i5bJyDcbrtBxWK487JWsMTt0pN5JK0auiUaTZQ/7Ltm7suv42heD3ManKfG2MOd73zLgptuivpPNtP6bqcRcRRQ11MPYZa4pC0o+0fzHNtH9vXdx1DW9aFLDwsdORNxg7i+DHwLdpPuO1j2uN9bd+rRBzzxLY3cLTt1xUo62K38xqG5zh0NedhVrtliq+NPMLyjp53i9RQFzXEUFEcF9KOnJK01vYhA9c+zF1HVXXG9hNLlLMIRT4hLFa7/tARNJ9q9qTcaqGe5/Go44mYyeS+Bf6J7v4z/YKk35zvou0PdVTuUnVZF7MUA3Qbx2A/8vASFcVGUUk6ELjB9nfb498BngNcB7ymVFfZXPekpJ2A/drT33TZ3aB2oFkw7LnA/WkS+r4usK7NgIdK+gHNa+Ae7WPa4+27KLDvyb3L/0y/QDOeeb4dZmpL7iWHZ86nhhig2ziKt9Dm8S7gyQCS5tZy+WPgYcBq4LdKBNFOFFoNHA5cQ1P395F0NvDCQje6bwa+BLwa+HfblvQbBcodtL3tn5cssO/Jvcv/TNfZLjYpZQJquLlSQwzQbRy7tVP/NfCY9rhkV96ygdb5UcBqN4t4naVmZ6JSXg1sA+xt+4dwZ0v6HcCftV9dO4lmud+/A94v6Z8LlDnsixT+1Dqrk5hqUEsrNOrybpqp//ceeDx3fFrBOJZJmmu8HUIz1n9OyUbdbwJ/OJfYAdrHx9N0lXTO9ptsHwQ8m+b/7YeBX5L0p5LuXyIGppAv+t5y7/Ij3/M2/yNVKTbefAE1xAAdxuFKpv0DHwA+I+kWmj1D/w1A0i8DJTcwucP2j4dP2v6RpKKf5GxfDZwCnCLpV2j64D9BM4qna8sXWMwNd7DXwEwOhZzP8NrIHZd1K80Kd5tcohl2t2PXMSykZF3UHEPpONTsfnQ/2+e2x2+iuT8D8HZ3sIbIPHFsTbMa4x40e4Xe2p6/P3DvgnFcCjyR0S3XC2w/tEQcw9pRM99zoQQoaT1Nt9DIe3S2/2LihU5jWuy4X8BDgPOArwN/RTON+CzgRuBlhWK4eNr1UFFdTD2GWuKg2Z/zMQPH36AZpfJ84MMF62KqU/sH4riWZkmIa0Z8XV0ohkfRDFH9EM2a8l+nWdvlZprZy1P9ewDP6aLMWe2WKb428gi1fOSpoS5qiKGWOPaw/bmB4x+43Y1I0gsKxQCV3BNyBatC0qzQehLNJ6hPA0+3/QVJD6TpvvpkgRgW+nu8iaYRMtkC23eOmSLpEtsPGzi+AVjhDtdGHhHDjTSruo3kcvuG1lAXU4+hljgkXWn7AfNc+6btIjfwKnp9LrjjkwvM2B18XUi63ANLL3Q1O3REDLt4nrkF6mjP5VltuRdfG3mEZTQjIKbdQqqhLmqIoZY4viPpINtfHDwp6VHAdwqUP6eW1+fH2HQ7StMMC92NJs6u3THw+H+GrhVp3c6X2LuMYVZb7hcyf4XY9pMKxDC13euH4riQ6dfF1GOoJQ5JjwT+GXgvTbcQNEsRrwKOsv2lrmNo46ji9TlM0grgT2kmWL3V9tsKlHk7cCvt7FBgbvSOaCYXbVMghq8x+rUp4P62t5t4mbOY3GtQ6uNczB5JuwMvAg6g+Q99GXA+zZZuLyoUQ1WvT0n70WySfRDNBhVrXHjG5uZI2tn29zt67lGryArYCzjJ9jMmXeZMTmKS9MqBx0cMXfvrQmEcsvkf6V4NdVFDDDXFYfsm239OM2LnappW+18Cl5eKgWZbv01I2lZSiVmhc+U9WNIHaG4Yfgp4sO3TakvsrbVdPbHt6+a+gJ1p3vwvpHldfLyLMmey5T6NtZFrVUNd1BBDLXG048iPpll18Hs0XTSvsL3g+v8dxPGvNH3Nx9u+pj33dJqRGZ+0fUKhOG4HbqDpe9/kxrbtl5SIYzG6/LQzjdfFrN5QnfruLhWpoS5qiKGWOK6gmQ36LNvfApD0skJl38n2UyUdA3xK0vuBB9PcxDzK82wi3pFjqWfY8OZ0GWfx18WsJvdaVt6rQQ11UUMMtcTxHJoW2gWSPgl8kOk1OM6g6fd/GfBfwJNsf7NkALbfW7K8ihV/Xcxqt8zU737Xooa6qCGGmuJoY7kXzTK3xwBPAtYAZ9s+r1D5jwPeCfwHzQSeJwB/Q9MdcIrtnxaK4yNs+qZ7C83SA6eXiGGxStyELvm6mNXkvk2lN2SKq6EuaoihpjiGqdnb9QiaLpFSw0LX0fS3f2ng3D2Bk4HDbD+wUByj9pTdhWbhvatsn1gijmGjJhUtNNGoqxjo8HUxq8n9bnXTdCE11EUNMdQURw0kbWX7jnmuPch2yZE7o2JYBlw0OKO4w7Jebfuv2sdzm6dvQ/OJ7qjhCWd9MZNDIbn73TRdSA11UUMMUE8cNXjF3IPhYaE0i5hNVeGlKQa3w3wd8FI3++0eSTN6qJdmteVexboZNaihLmqIoaY4alDDsNC2rOF9ZKEZ5/07wC/b/u0CMQzWxV361Wub7DVJszpaZqF1M2bv3Wo8NdRFDTHUFEcNahgWCnARd11bZu6G6oXACwvFsK+kc9sY9pJ0T2/cQKS3gy9mNbmv9zyL20t6TulgpqyGuqghhpriqEENw0IBHm/7hlEXJD2LZv37rh02dLxVW/7uNEtE99KsdsvM+1FK0vW2F1xmtE9qqIsaYqgpjhrUMixU0pXAU21fO3T+94BX2y6xxd3d0qy23Bda1+XudlOthrqoIQaoJ46ps11iKd3FeBlwvqRn2L4KQNKraPYvHTVMcuLUbDl4LM348j1pPrl8BzgHeE+Nw2cnYSaT+2bGos7eR5Ex1FAXNcRQUxyxke2PS/op8AlJhwN/ABxI013TyQqMI/wTzQzd19JsuQjNaoyrgNOBowrFUdSsdssUXxu5VjXURQ0x1BRHbKqdMfth4HPAkS64/aIq2R2rtJlsuQPPHHHuzrWRC8cybTXURQ0x1BRHtCT9kI2jZbaj6Tq7WZJoNlDZsUAY32/H+p81N7FL0lY0s0NLfXoobiaTe7smMgCSHkbTf3ckzY7qE99otmY11EUNMdQUR2xke4fF/FyXG2XQLNj1N8A7Jc2VsRNwQXutl2Yyuc+zNrJsHzzVwKaghrqoIYaa4oglWQt0MrGqHalzFICk/0XzmrilPe7tOPdZ7XO/g2Zt5GMH1ka+2va+042svBrqooYYaoojtlzJmaJtl9DBNJ/snmV79xLlljara8s8B/guzdrI75Z0CHezoW4DaqiLGmKoKY7Ycp23MiUdJOktwHXAuTQNgSKrY07DTLbc50x7zeya1FAXNcRQUxyxeF2udyPpFJp7L9cDHwDOBta1i4f11kwn90HTWDO7VjXURQ0x1BRHbErStrZ/1j7urFtG0gbgSuDNwEdt/+Tu0F3Xm+QeEfWR9Ge2/3LE+V8AzrH9xPa4s40y2rXjn8LGT3IXAE8G9rZ9Wxdl1mBW+9wjYjb8WtstcidJvwh8Fvj03Lkud0CyfbvtT9j+HeCXaZYd+BzwbTWbh/dSWu4R0RlJ2wNnAt+0/XJJ+wGfAF5n+11Tjm0H4Ddtr5lmHF1Jco+ITrVjyT8I/Bx4NHCC7bMLx/AE4Pu2vyrpSODxwP8D3ulCm4WXluQeEZ2R9PL24TbAK2mGH3527nqhncLeATyEZvmDb9Js6PJJ4DHAshK7QU3DTM5QjYiZMbj8wFuHzpVqWR5se/+2i+jbwG62b5f0LuCrhWIoLsk9Irr0Hts3jrrQ7sRUwk8A2iGQ17ndnNu2JfVyLXdIco+Ibq2VNO9OTJTZZm+3tntIA49pj5cXKH8q0uceEZ2R9AzgLcConZiePl+rfsIxnLzQdduv7TqGaUjLPSI6U8lOTDV0DRWXSUwR0Snba4HfBS4E9gUOKZjYoekaWjF8su0aenPBOIpKt0xEdGbETkw/B25vj4vsxFRD19A0pFsmIjqz2J2YOo6hhq6h4tJyj4i7hWlu0j0NSe4R0Ws1dA1NQ5J7REQPZbRMREQPJblHRPRQkntERA8luUdE9FCSe0RED/1/8jwNheWf2roAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot a graph\n",
    "feature_azdias[feature_azdias.Missing_Percentage > 30].Missing_Percentage.plot(kind='bar');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9e0235d",
   "metadata": {},
   "source": [
    "#### To get a better understanding of missing values in both datasets, let's combine them for further analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "2bc93b68",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Value_Customer</th>\n",
       "      <th>Missing_Customers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ALTER_KIND4</td>\n",
       "      <td>11.0</td>\n",
       "      <td>99.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTER_KIND3</td>\n",
       "      <td>15.0</td>\n",
       "      <td>99.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TITEL_KZ</td>\n",
       "      <td>5.0</td>\n",
       "      <td>98.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ALTER_KIND2</td>\n",
       "      <td>18.0</td>\n",
       "      <td>97.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ALTER_KIND1</td>\n",
       "      <td>18.0</td>\n",
       "      <td>93.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KK_KUNDENTYP</td>\n",
       "      <td>7.0</td>\n",
       "      <td>58.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>KBA05_BAUMAX</td>\n",
       "      <td>6.0</td>\n",
       "      <td>57.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>5.0</td>\n",
       "      <td>48.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>EXTSEL992</td>\n",
       "      <td>57.0</td>\n",
       "      <td>44.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ALTER_HH</td>\n",
       "      <td>21.0</td>\n",
       "      <td>35.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>REGIOTYP</td>\n",
       "      <td>8.0</td>\n",
       "      <td>31.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>KKK</td>\n",
       "      <td>5.0</td>\n",
       "      <td>31.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>W_KEIT_KIND_HH</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>KBA05_ALTER4</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>KBA05_HERST1</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             index  Value_Customer  Missing_Customers\n",
       "0      ALTER_KIND4            11.0              99.88\n",
       "1      ALTER_KIND3            15.0              99.33\n",
       "2         TITEL_KZ             5.0              98.79\n",
       "3      ALTER_KIND2            18.0              97.34\n",
       "4      ALTER_KIND1            18.0              93.86\n",
       "5     KK_KUNDENTYP             7.0              58.41\n",
       "6     KBA05_BAUMAX             6.0              57.15\n",
       "7         AGER_TYP             5.0              48.06\n",
       "8        EXTSEL992            57.0              44.50\n",
       "9         ALTER_HH            21.0              35.87\n",
       "10        REGIOTYP             8.0              31.34\n",
       "11             KKK             5.0              31.34\n",
       "12  W_KEIT_KIND_HH             7.0              29.71\n",
       "13    KBA05_ALTER4             7.0              29.70\n",
       "14    KBA05_HERST1             7.0              29.70"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add a customer data in temp\n",
    "temp_c = feature_customers.reset_index()\n",
    "temp_c.rename(columns={'Attribute':'Value_Customer', 'Missing_Percentage':'Missing_Customers'}, inplace=True)\n",
    "temp_c.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "3e1f46e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Value_Azdias</th>\n",
       "      <th>Missing_azdias</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ALTER_KIND4</td>\n",
       "      <td>13.0</td>\n",
       "      <td>99.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>TITEL_KZ</td>\n",
       "      <td>6.0</td>\n",
       "      <td>99.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ALTER_KIND3</td>\n",
       "      <td>16.0</td>\n",
       "      <td>99.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ALTER_KIND2</td>\n",
       "      <td>18.0</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ALTER_KIND1</td>\n",
       "      <td>18.0</td>\n",
       "      <td>90.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>5.0</td>\n",
       "      <td>76.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>EXTSEL992</td>\n",
       "      <td>57.0</td>\n",
       "      <td>73.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>KK_KUNDENTYP</td>\n",
       "      <td>7.0</td>\n",
       "      <td>65.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>KBA05_BAUMAX</td>\n",
       "      <td>6.0</td>\n",
       "      <td>53.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ALTER_HH</td>\n",
       "      <td>22.0</td>\n",
       "      <td>34.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ALTERSKATEGORIE_FEIN</td>\n",
       "      <td>27.0</td>\n",
       "      <td>29.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>D19_LETZTER_KAUF_BRANCHE</td>\n",
       "      <td>36.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>D19_BANKEN_ONLINE_QUOTE_12</td>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>D19_GESAMT_ONLINE_QUOTE_12</td>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>D19_VERSAND_ONLINE_QUOTE_12</td>\n",
       "      <td>12.0</td>\n",
       "      <td>28.85</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          index  Value_Azdias  Missing_azdias\n",
       "0                   ALTER_KIND4          13.0           99.86\n",
       "1                      TITEL_KZ           6.0           99.76\n",
       "2                   ALTER_KIND3          16.0           99.31\n",
       "3                   ALTER_KIND2          18.0           96.69\n",
       "4                   ALTER_KIND1          18.0           90.90\n",
       "5                      AGER_TYP           5.0           76.02\n",
       "6                     EXTSEL992          57.0           73.40\n",
       "7                  KK_KUNDENTYP           7.0           65.60\n",
       "8                  KBA05_BAUMAX           6.0           53.47\n",
       "9                      ALTER_HH          22.0           34.81\n",
       "10         ALTERSKATEGORIE_FEIN          27.0           29.50\n",
       "11     D19_LETZTER_KAUF_BRANCHE          36.0           28.85\n",
       "12   D19_BANKEN_ONLINE_QUOTE_12          12.0           28.85\n",
       "13   D19_GESAMT_ONLINE_QUOTE_12          12.0           28.85\n",
       "14  D19_VERSAND_ONLINE_QUOTE_12          12.0           28.85"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add a azdias data in temp\n",
    "temp_a = feature_azdias.reset_index()\n",
    "temp_a.rename(columns={'Attribute':'Value_Azdias', 'Missing_Percentage':'Missing_azdias'}, inplace=True)\n",
    "temp_a.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "4519dad6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Value_Customer</th>\n",
       "      <th>Missing_Customers</th>\n",
       "      <th>Value_Azdias</th>\n",
       "      <th>Missing_azdias</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ALTER_KIND4</td>\n",
       "      <td>11.0</td>\n",
       "      <td>99.88</td>\n",
       "      <td>13.0</td>\n",
       "      <td>99.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ALTER_KIND3</td>\n",
       "      <td>15.0</td>\n",
       "      <td>99.33</td>\n",
       "      <td>16.0</td>\n",
       "      <td>99.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TITEL_KZ</td>\n",
       "      <td>5.0</td>\n",
       "      <td>98.79</td>\n",
       "      <td>6.0</td>\n",
       "      <td>99.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ALTER_KIND2</td>\n",
       "      <td>18.0</td>\n",
       "      <td>97.34</td>\n",
       "      <td>18.0</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ALTER_KIND1</td>\n",
       "      <td>18.0</td>\n",
       "      <td>93.86</td>\n",
       "      <td>18.0</td>\n",
       "      <td>90.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KK_KUNDENTYP</td>\n",
       "      <td>7.0</td>\n",
       "      <td>58.41</td>\n",
       "      <td>7.0</td>\n",
       "      <td>65.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>KBA05_BAUMAX</td>\n",
       "      <td>6.0</td>\n",
       "      <td>57.15</td>\n",
       "      <td>6.0</td>\n",
       "      <td>53.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AGER_TYP</td>\n",
       "      <td>5.0</td>\n",
       "      <td>48.06</td>\n",
       "      <td>5.0</td>\n",
       "      <td>76.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>EXTSEL992</td>\n",
       "      <td>57.0</td>\n",
       "      <td>44.50</td>\n",
       "      <td>57.0</td>\n",
       "      <td>73.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ALTER_HH</td>\n",
       "      <td>21.0</td>\n",
       "      <td>35.87</td>\n",
       "      <td>22.0</td>\n",
       "      <td>34.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>REGIOTYP</td>\n",
       "      <td>8.0</td>\n",
       "      <td>31.34</td>\n",
       "      <td>8.0</td>\n",
       "      <td>17.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>KKK</td>\n",
       "      <td>5.0</td>\n",
       "      <td>31.34</td>\n",
       "      <td>5.0</td>\n",
       "      <td>17.74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>W_KEIT_KIND_HH</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.71</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>KBA05_ALTER4</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>KBA05_HERST1</td>\n",
       "      <td>7.0</td>\n",
       "      <td>29.70</td>\n",
       "      <td>7.0</td>\n",
       "      <td>16.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             index  Value_Customer  Missing_Customers  Value_Azdias  \\\n",
       "0      ALTER_KIND4            11.0              99.88          13.0   \n",
       "1      ALTER_KIND3            15.0              99.33          16.0   \n",
       "2         TITEL_KZ             5.0              98.79           6.0   \n",
       "3      ALTER_KIND2            18.0              97.34          18.0   \n",
       "4      ALTER_KIND1            18.0              93.86          18.0   \n",
       "5     KK_KUNDENTYP             7.0              58.41           7.0   \n",
       "6     KBA05_BAUMAX             6.0              57.15           6.0   \n",
       "7         AGER_TYP             5.0              48.06           5.0   \n",
       "8        EXTSEL992            57.0              44.50          57.0   \n",
       "9         ALTER_HH            21.0              35.87          22.0   \n",
       "10        REGIOTYP             8.0              31.34           8.0   \n",
       "11             KKK             5.0              31.34           5.0   \n",
       "12  W_KEIT_KIND_HH             7.0              29.71           7.0   \n",
       "13    KBA05_ALTER4             7.0              29.70           7.0   \n",
       "14    KBA05_HERST1             7.0              29.70           7.0   \n",
       "\n",
       "    Missing_azdias  \n",
       "0            99.86  \n",
       "1            99.31  \n",
       "2            99.76  \n",
       "3            96.69  \n",
       "4            90.90  \n",
       "5            65.60  \n",
       "6            53.47  \n",
       "7            76.02  \n",
       "8            73.40  \n",
       "9            34.81  \n",
       "10           17.74  \n",
       "11           17.74  \n",
       "12           16.61  \n",
       "13           16.62  \n",
       "14           16.62  "
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge both temp datasets\n",
    "merged_data = temp_c.merge(temp_a, on='index')\n",
    "merged_data.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "a35bacc4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ALTER_KIND4',\n",
       " 'ALTER_KIND3',\n",
       " 'TITEL_KZ',\n",
       " 'ALTER_KIND2',\n",
       " 'ALTER_KIND1',\n",
       " 'KK_KUNDENTYP',\n",
       " 'KBA05_BAUMAX',\n",
       " 'AGER_TYP',\n",
       " 'EXTSEL992',\n",
       " 'ALTER_HH']"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find highest missing columns which are common in both datasets\n",
    "common_cols = list(merged_data.loc[(merged_data['Missing_Customers'] > 30) & (merged_data['Missing_azdias'] > 30)]['index'])\n",
    "common_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "0c3fd40b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop these columns from both datasets\n",
    "azdias_new = azdias_new.drop(common_cols, axis = 1)\n",
    "customers_new = customers_new.drop(common_cols, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "c9f894e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of datasets after dropping columns:\n",
      "Customers: (191652, 356)\n",
      "Azdias: (891221, 356)\n"
     ]
    }
   ],
   "source": [
    "# Check the shape of data after dropping columns\n",
    "print(\"Shape of datasets after dropping columns:\")\n",
    "print(f\"Customers: {customers_new.shape}\")\n",
    "print(f\"Azdias: {azdias_new.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "1de3433c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to find missing rows\n",
    "def missing_row_hist(df1, df2, bins=30, figsize=(10,6)):\n",
    "    \"\"\"\n",
    "    This function plot a barchart comparing two dataframes for missing rows\n",
    "    and percentage of missing values based on a threshold percentage given.\n",
    "    \n",
    "    Input:\n",
    "    df1 - Dataframe to find missing rows\n",
    "    df2 - Dataframe to find missing rows\n",
    "    \n",
    "    Output:\n",
    "    None\n",
    "    \"\"\"\n",
    "    bins = 30\n",
    "    fig = plt.figure(figsize=figsize)\n",
    "\n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.set_title(\"Features wise Missing Rows\", fontsize = 14)\n",
    "    ax.set_xlabel(\"Features\", fontsize = 12)\n",
    "    ax.set_ylabel(\"Number of Missing Rows\", fontsize = 12)\n",
    "    sns.distplot(df1.isnull().sum(axis=1), bins, kde=False, ax=ax, label=\"Azdias\")\n",
    "    sns.distplot(df2.isnull().sum(axis=1), bins, kde=False, ax=ax, label=\"Customers\")\n",
    "    ax.legend()\n",
    "    \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "5ac57396",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot a graph\n",
    "missing_row_hist(azdias_new, customers_new)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cab89465",
   "metadata": {},
   "source": [
    "#### The figure shows that there are 250 features out 356 contains missing rows. The max missing rows in azdias dataset are more than 500000 and for customers dataset the count is greater that 100000. There are less than 50 missing values in both datasets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "234bbe35",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to drop missing rows\n",
    "def drop_missing_rows(df, threshold):\n",
    "    \"\"\"\n",
    "    This function drops rows with number of missing features as per the given threshold.\n",
    "    \n",
    "    Input:\n",
    "    df - Dataframe to drop the missing rows\n",
    "    threshold - Threshold to delete rows\n",
    "    \n",
    "    Output:\n",
    "    df - The transformed dataframe\n",
    "    \"\"\"\n",
    "    total_rows = df.shape[0]\n",
    "    \n",
    "    df = df.dropna(thresh=df.shape[1]-threshold)\n",
    "    dropped_rows = total_rows - df.shape[0]\n",
    "    \n",
    "    # Reset index\n",
    "    df = df.reset_index()\n",
    "    del df['index']\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "f19d4e9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Apply the function\n",
    "customers_new = drop_missing_rows(customers_new, threshold=50)\n",
    "azdias_new = drop_missing_rows(azdias_new, threshold=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "6ab8f462",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of datasets after dropping rows:\n",
      "Customers: 134246\n",
      "Azdias: 737288\n"
     ]
    }
   ],
   "source": [
    "# Check the shape of data after dropping rows\n",
    "print(\"Shape of datasets after dropping rows:\")\n",
    "print(f\"Customers: {len(customers_new)}\")\n",
    "print(f\"Azdias: {len(azdias_new)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e41a216e",
   "metadata": {},
   "source": [
    "#### We can start feature engineering from this point, since some columns have mixed information and some have too much information packed into one single column. We can address these issues by creating new features or modifying existing features to contain useful information.\n",
    "1. `CAMEO_INTL_2015` into `CAMEO_INTL_2015_wealth_type` and `CAMEO_INTL_2015_family_type`\n",
    "2. `LP_LEBENSPHASE_FEIN` into `LP_LEBENSPHASE_FEIN_family_type`, `LP_LEBENSPHASE_FEIN_earner_type`, `LP_LEBENSPHASE_FEIN_age_group`\n",
    "3. `PRAEGENDE_JUGENDJAHRE` by movent type (avangard or mainstream) and decade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "335122d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to clean the data\n",
    "def data_cleaning(df):\n",
    "    '''\n",
    "    This function deals with designated columns and imputes missing data. \n",
    "    \n",
    "    Input: \n",
    "    df -  Dataframe to perform cleaning operation\n",
    "    \n",
    "    Output:\n",
    "    None\n",
    "    ''' \n",
    "    labelencoder = LabelEncoder()\n",
    "   \n",
    "    # CAMEO_INTL_2015\n",
    "    cameo_wealth_type = {\n",
    "        11:'Wealthy', 12:'Wealthy',13:'Wealthy', 14:'Wealthy', 15:'Wealthy', \n",
    "        21:'Prosperous', 22:'Prosperous', 23:'Prosperous', 24:'Prosperous', 25:'Prosperous',\n",
    "        31:'Comfortable', 32:'Comfortable', 33:'Comfortable', 34:'Comfortable', 35:'Comfortable',\n",
    "        41:'Less Affluent', 42:'Less Affluent', 43:'Less Affluent', 44:'Less Affluent', 45:'Less Affluent',\n",
    "        51:'Poorer', 52:'Poorer', 53:'Poorer', 54:'Poorer', 55:'Poorer'\n",
    "    }\n",
    "    cameo_family_type = {\n",
    "        11:'Pre-Family Couples & Singles',\n",
    "        12:'Young Couples With Children',\n",
    "        13:'Families With School Age Children',\n",
    "        14:'Older Families &  Mature Couples',\n",
    "        15:'Elders In Retirement',\n",
    "        21:'Pre-Family Couples & Singles',\n",
    "        22:'Young Couples With Children',\n",
    "        23:'Families With School Age Children',\n",
    "        24:'Older Families &  Mature Couples',\n",
    "        25:'Elders In Retirement',\n",
    "        31:'Pre-Family Couples & Singles',\n",
    "        32:'Young Couples With Children',\n",
    "        33:'Families With School Age Children',\n",
    "        34:'Older Families &  Mature Couples',\n",
    "        35:'Elders In Retirement',\n",
    "        41:'Pre-Family Couples & Singles',\n",
    "        42:'Young Couples With Children',\n",
    "        43:'Families With School Age Children',\n",
    "        44:'Older Families &  Mature Couples',\n",
    "        45:'Elders In Retirement',\n",
    "        51:'Pre-Family Couples & Singles',\n",
    "        52:'Young Couples With Children',\n",
    "        53:'Families With School Age Children',\n",
    "        54:'Older Families &  Mature Couples',\n",
    "        55:'Elders In Retirement',\n",
    "\n",
    "    }\n",
    "    \n",
    "    # not in customers, but in azdias\n",
    "    if 'CAMEO_INTL_2015' in df.columns:\n",
    "        df['CAMEO_INTL_2015_wealth_type'] = df['CAMEO_INTL_2015'].map(cameo_wealth_type)\n",
    "        df['CAMEO_INTL_2015_family_type'] = df['CAMEO_INTL_2015'].map(cameo_family_type)\n",
    "    \n",
    "    # LP_LEBENSPHASE_FEIN\n",
    "    family_type_dict = {\n",
    "        1: 'single', 2: 'single', 3: 'single', 4: 'single', 5: 'single', 6: 'single', 7: 'single',\n",
    "        8: 'single', 9: 'single', 10:'single', 11:'single', 12:'single', 13:'single', 14:'couples',\n",
    "        15:'couples', 16:'couples', 17:'couples', 18:'couples', 19:'couples', 20:'couples', 21:'single_parent',\n",
    "        22:'single_parent', 23:'single_parent', 24:'families', 25:'families', 26:'families', 27:'families', 28:'families',\n",
    "        29:'mulitperson_households', 30:'mulitperson_households', 31:'mulitperson_households', 32:'mulitperson_households',\n",
    "        33:'mulitperson_households', 34:'mulitperson_households', 35:'mulitperson_households', 36:'mulitperson_households',\n",
    "        37:'mulitperson_households', 38:'mulitperson_households', 39:'mulitperson_households', 40:'mulitperson_households'\n",
    "    }\n",
    "    earner_type_dict = {\n",
    "        1: 'low', 2: 'low', 3: 'average', 4: 'average', 5: 'low', 6: 'low', 7: 'average',\n",
    "        8: 'average', 9: 'independant', 10:'wealthy', 11:'homeowners', 12:'homeowners', 13:'top', 14:'low',\n",
    "        15: 'low', 16:'average', 17: 'independant', 18:'wealthy', 19:'homeowners', 20:'top', 21:'low',\n",
    "        22: 'average', 23:'high', 24:'low', 25:'average', 26: 'independant', 27:'homeowners', 28:'top',\n",
    "        29: 'low', 30:'average', 31:'low', 32:'average',\n",
    "        33: 'independant', 34:'homeowners', 35:'top', 36: 'independant',\n",
    "        37: 'homeowners', 38:'homeowners', 39:'top', 40:'top'\n",
    "    }\n",
    "    age_group_dict = {\n",
    "        1: 'younger_age', 2: 'middle_age', 3: 'younger_age', 4: 'middle_age', 5: 'advanced_age', \n",
    "        6: 'retirement_age', 7: 'advanced_age', 8: 'retirement_age', 9: np.nan, 10: np.nan, \n",
    "        11:'advanced_age', 12:'retirement_age', 13:'higher_age', 14:'younger_age', 15: 'higher_age', \n",
    "        16:'higher_age', 17: np.nan, 18:'younger_age', 19:'higher_age', 20:'higher_age', \n",
    "        21: np.nan, 22: np.nan, 23:np.nan, 24:np.nan, 25:np.nan, \n",
    "        26: np.nan, 27:np.nan, 28:np.nan, 29: 'younger_age', 30:'younger_age', \n",
    "        31:'higher_age', 32:'higher_age', 33: 'younger_age', 34:'younger_age', 35:'younger_age', \n",
    "        36: 'higher_age', 37: 'advanced_age', 38:'retirement_age', 39:'middle_age', 40:'retirement_age'\n",
    "    }\n",
    "    \n",
    "    if 'LP_LEBENSPHASE_FEIN' in df.columns:\n",
    "        df['LP_LEBENSPHASE_FEIN_family_type'] = df['LP_LEBENSPHASE_FEIN'].map(family_type_dict)\n",
    "        df['LP_LEBENSPHASE_FEIN_earner_type'] = df['LP_LEBENSPHASE_FEIN'].map(earner_type_dict)\n",
    "        df['LP_LEBENSPHASE_FEIN_age_group'] = df['LP_LEBENSPHASE_FEIN'].map(age_group_dict)\n",
    "    \n",
    "    # LP_FAMILIE_GROB\n",
    "    family_type_dict = {\n",
    "        1: 'single', 2: 'couple', 3: 'single_parent', 4: 'single_parent', 5: 'single_parent', \n",
    "        6: 'family', 7: 'family', 8: 'family', 9: 'mulitperson_households', 10: 'mulitperson_households', \n",
    "        11: 'mulitperson_households'\n",
    "    }\n",
    "    if 'LP_FAMILIE_GROB' in df.columns:\n",
    "        df['LP_FAMILIE_GROB_family_type'] = df['LP_FAMILIE_GROB'].map(family_type_dict)\n",
    "    \n",
    "    # LP_STATUS_GROB\n",
    "    earner_type_dict = {\n",
    "        1: 'low', 2: 'low', 3: 'average', 4: 'average', 5: 'average', \n",
    "        6: 'independant', 7: 'independant', 8: 'houseowners', 9: 'houseowners', 10: 'top'\n",
    "    }\n",
    "    if 'LP_STATUS_GROB' in df.columns:\n",
    "        df['LP_STATUS_GROB_earner_type'] = df['LP_STATUS_GROB'].map(earner_type_dict)\n",
    "    \n",
    "    \n",
    "    # PRAEGENDE_JUGENDJAHRE\n",
    "    movement_dict = {\n",
    "        1: 'mainstream', 2: 'avantgarde', 3: 'mainstream', 4: 'avantgarde',\n",
    "        5: 'mainstream', 6: 'avantgarde', 7: 'avantgarde', 8: 'mainstream', 9: 'avantgarde', \n",
    "        10: 'mainstream', 11: 'avantgarde', 12: 'mainstream', 13: 'avantgarde', 14: 'mainstream',\n",
    "        15: 'avantgarde'\n",
    "    }\n",
    "    if 'PRAEGENDE_JUGENDJAHRE' in df.columns:\n",
    "        df['PRAEGENDE_JUGENDJAHRE_movement'] = df['PRAEGENDE_JUGENDJAHRE'].map(movement_dict)\n",
    "    \n",
    "    decade_dict = {\n",
    "        1: '40s', 2: '40s', 3: '50s', 4: '50s',\n",
    "        5: '60s', 6: '60s', 7: '60s', 8: '70s', 9: '70s', \n",
    "        10: '80s', 11: '80s', 12: '80s', 13: '80s', 14: '90s',\n",
    "        15: '90s'\n",
    "    }\n",
    "    if 'PRAEGENDE_JUGENDJAHRE' in df.columns:\n",
    "        df['PRAEGENDE_JUGENDJAHRE_decade'] = df['PRAEGENDE_JUGENDJAHRE'].map(decade_dict)\n",
    "    \n",
    "    \n",
    "    # not in customers, but in azdias\n",
    "    if 'EINGEFUEGT_AM' in df.columns:\n",
    "        df['EINGEFUEGT_AM'] = pd.to_datetime(df['EINGEFUEGT_AM']).dt.year\n",
    "    \n",
    "    list_of_obj_cols = ['CAMEO_INTL_2015', 'CAMEO_INTL_2015_wealth_type','CAMEO_INTL_2015_family_type',\n",
    "                        'LP_LEBENSPHASE_FEIN', 'LP_LEBENSPHASE_FEIN_family_type', 'LP_LEBENSPHASE_FEIN_earner_type', \n",
    "                        'LP_LEBENSPHASE_FEIN_age_group', 'LP_FAMILIE_GROB', 'LP_FAMILIE_GROB_family_type', \n",
    "                        'LP_STATUS_GROB', 'LP_STATUS_GROB_earner_type', 'PRAEGENDE_JUGENDJAHRE', \n",
    "                        'PRAEGENDE_JUGENDJAHRE_movement', 'PRAEGENDE_JUGENDJAHRE_decade', 'EINGEFUEGT_AM',\n",
    "                        'CAMEO_INTL_2015', 'OST_WEST_KZ', 'CAMEO_DEU_2015', 'PRODUCT_GROUP', 'CUSTOMER_GROUP', \n",
    "                        'D19_LETZTER_KAUF_BRANCHE']\n",
    "   \n",
    "    # imputing with most frequent\n",
    "    for col in df.columns:\n",
    "        if col in list_of_obj_cols:\n",
    "            df[col].fillna(df[col].value_counts().idxmax(), inplace=True)\n",
    "        else: \n",
    "            print(col)\n",
    "            df[col].fillna(df[col].median(), inplace=True)\n",
    "\n",
    "    \n",
    "    # label encoding\n",
    "    \n",
    "    if 'CAMEO_INTL_2015' in df.columns:\n",
    "        df['CAMEO_INTL_2015_wealth_type'] = labelencoder.fit_transform(df['CAMEO_INTL_2015_wealth_type'])\n",
    "        df['CAMEO_INTL_2015_family_type'] = labelencoder.fit_transform(df['CAMEO_INTL_2015_family_type'])\n",
    "    \n",
    "    if 'LP_LEBENSPHASE_FEIN_family_type' in df.columns:\n",
    "        df['LP_LEBENSPHASE_FEIN_family_type'] = labelencoder.fit_transform(df['LP_LEBENSPHASE_FEIN_family_type'])\n",
    "        df['LP_LEBENSPHASE_FEIN_earner_type'] = labelencoder.fit_transform(df['LP_LEBENSPHASE_FEIN_earner_type'])\n",
    "        df['LP_LEBENSPHASE_FEIN_age_group'] = labelencoder.fit_transform(df['LP_LEBENSPHASE_FEIN_age_group'])\n",
    "\n",
    "    \n",
    "    if 'LP_LEBENSPHASE_FEIN_family_type' in df.columns:\n",
    "        df['LP_FAMILIE_GROB_family_type'] = labelencoder.fit_transform(df['LP_FAMILIE_GROB_family_type'])\n",
    "\n",
    "    \n",
    "    if 'LP_LEBENSPHASE_FEIN_family_type' in df.columns:\n",
    "        df['LP_STATUS_GROB_earner_type'] = labelencoder.fit_transform(df['LP_STATUS_GROB_earner_type'])\n",
    "\n",
    "    drop_cols = ['LP_LEBENSPHASE_FEIN', 'LP_FAMILIE_GROB', 'LP_STATUS_GROB', 'PRAEGENDE_JUGENDJAHRE', \n",
    "                 'CAMEO_INTL_2015', 'D19_LETZTER_KAUF_BRANCHE']\n",
    "    \n",
    "    df.drop(drop_cols, axis = 1, inplace = True, errors='ignore')\n",
    "    \n",
    "    # OST_WEST_KZ\n",
    "    if 'OST_WEST_KZ' in df.columns:\n",
    "        df['OST_WEST_KZ'] = labelencoder.fit_transform(df['OST_WEST_KZ'])\n",
    "        \n",
    "    \n",
    "    # CAMEO_DEU_2015\n",
    "    if 'CAMEO_DEU_2015' in df.columns:\n",
    "        df['CAMEO_DEU_2015'] = labelencoder.fit_transform(df['CAMEO_DEU_2015'])\n",
    "    \n",
    "    \n",
    "    for value in ['PRODUCT_GROUP', 'CUSTOMER_GROUP']:\n",
    "        if value in df.columns:\n",
    "            df[value] = labelencoder.fit_transform(df[value])\n",
    "            \n",
    "    if 'PRAEGENDE_JUGENDJAHRE_movement' in df.columns:   \n",
    "        df['PRAEGENDE_JUGENDJAHRE_movement'] = labelencoder.fit_transform(df['PRAEGENDE_JUGENDJAHRE_movement'])\n",
    "        df['PRAEGENDE_JUGENDJAHRE_decade'] = labelencoder.fit_transform(df['PRAEGENDE_JUGENDJAHRE_decade'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "436e8c60",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LNR\n",
      "AKT_DAT_KL\n",
      "ALTERSKATEGORIE_FEIN\n",
      "ANZ_HAUSHALTE_AKTIV\n",
      "ANZ_HH_TITEL\n",
      "ANZ_KINDER\n",
      "ANZ_PERSONEN\n",
      "ANZ_STATISTISCHE_HAUSHALTE\n",
      "ANZ_TITEL\n",
      "ARBEIT\n",
      "BALLRAUM\n",
      "CAMEO_DEUG_2015\n",
      "CJT_GESAMTTYP\n",
      "CJT_KATALOGNUTZER\n",
      "CJT_TYP_1\n",
      "CJT_TYP_2\n",
      "CJT_TYP_3\n",
      "CJT_TYP_4\n",
      "CJT_TYP_5\n",
      "CJT_TYP_6\n",
      "D19_BANKEN_ANZ_12\n",
      "D19_BANKEN_ANZ_24\n",
      "D19_BANKEN_DATUM\n",
      "D19_BANKEN_DIREKT\n",
      "D19_BANKEN_GROSS\n",
      "D19_BANKEN_LOKAL\n",
      "D19_BANKEN_OFFLINE_DATUM\n",
      "D19_BANKEN_ONLINE_DATUM\n",
      "D19_BANKEN_ONLINE_QUOTE_12\n",
      "D19_BANKEN_REST\n",
      "D19_BEKLEIDUNG_GEH\n",
      "D19_BEKLEIDUNG_REST\n",
      "D19_BILDUNG\n",
      "D19_BIO_OEKO\n",
      "D19_BUCH_CD\n",
      "D19_DIGIT_SERV\n",
      "D19_DROGERIEARTIKEL\n",
      "D19_ENERGIE\n",
      "D19_FREIZEIT\n",
      "D19_GARTEN\n",
      "D19_GESAMT_ANZ_12\n",
      "D19_GESAMT_ANZ_24\n",
      "D19_GESAMT_DATUM\n",
      "D19_GESAMT_OFFLINE_DATUM\n",
      "D19_GESAMT_ONLINE_DATUM\n",
      "D19_GESAMT_ONLINE_QUOTE_12\n",
      "D19_HANDWERK\n",
      "D19_HAUS_DEKO\n",
      "D19_KINDERARTIKEL\n",
      "D19_KONSUMTYP\n",
      "D19_KONSUMTYP_MAX\n",
      "D19_KOSMETIK\n",
      "D19_LEBENSMITTEL\n",
      "D19_LOTTO\n",
      "D19_NAHRUNGSERGAENZUNG\n",
      "D19_RATGEBER\n",
      "D19_REISEN\n",
      "D19_SAMMELARTIKEL\n",
      "D19_SCHUHE\n",
      "D19_SONSTIGE\n",
      "D19_SOZIALES\n",
      "D19_TECHNIK\n",
      "D19_TELKO_ANZ_12\n",
      "D19_TELKO_ANZ_24\n",
      "D19_TELKO_DATUM\n",
      "D19_TELKO_MOBILE\n",
      "D19_TELKO_OFFLINE_DATUM\n",
      "D19_TELKO_ONLINE_DATUM\n",
      "D19_TELKO_ONLINE_QUOTE_12\n",
      "D19_TELKO_REST\n",
      "D19_TIERARTIKEL\n",
      "D19_VERSAND_ANZ_12\n",
      "D19_VERSAND_ANZ_24\n",
      "D19_VERSAND_DATUM\n",
      "D19_VERSAND_OFFLINE_DATUM\n",
      "D19_VERSAND_ONLINE_DATUM\n",
      "D19_VERSAND_ONLINE_QUOTE_12\n",
      "D19_VERSAND_REST\n",
      "D19_VERSI_ANZ_12\n",
      "D19_VERSI_ANZ_24\n",
      "D19_VERSI_DATUM\n",
      "D19_VERSI_OFFLINE_DATUM\n",
      "D19_VERSI_ONLINE_DATUM\n",
      "D19_VERSI_ONLINE_QUOTE_12\n",
      "D19_VERSICHERUNGEN\n",
      "D19_VOLLSORTIMENT\n",
      "D19_WEIN_FEINKOST\n",
      "DSL_FLAG\n",
      "EINGEZOGENAM_HH_JAHR\n",
      "EWDICHTE\n",
      "FINANZ_ANLEGER\n",
      "FINANZ_HAUSBAUER\n",
      "FINANZ_MINIMALIST\n",
      "FINANZ_SPARER\n",
      "FINANZ_UNAUFFAELLIGER\n",
      "FINANZ_VORSORGER\n",
      "FINANZTYP\n",
      "FIRMENDICHTE\n",
      "GEBAEUDETYP\n",
      "GEBAEUDETYP_RASTER\n",
      "GEBURTSJAHR\n",
      "GEMEINDETYP\n",
      "GFK_URLAUBERTYP\n",
      "GREEN_AVANTGARDE\n",
      "HEALTH_TYP\n",
      "HH_DELTA_FLAG\n",
      "HH_EINKOMMEN_SCORE\n",
      "INNENSTADT\n",
      "KBA05_ALTER1\n",
      "KBA05_ALTER2\n",
      "KBA05_ALTER3\n",
      "KBA05_ALTER4\n",
      "KBA05_ANHANG\n",
      "KBA05_ANTG1\n",
      "KBA05_ANTG2\n",
      "KBA05_ANTG3\n",
      "KBA05_ANTG4\n",
      "KBA05_AUTOQUOT\n",
      "KBA05_CCM1\n",
      "KBA05_CCM2\n",
      "KBA05_CCM3\n",
      "KBA05_CCM4\n",
      "KBA05_DIESEL\n",
      "KBA05_FRAU\n",
      "KBA05_GBZ\n",
      "KBA05_HERST1\n",
      "KBA05_HERST2\n",
      "KBA05_HERST3\n",
      "KBA05_HERST4\n",
      "KBA05_HERST5\n",
      "KBA05_HERSTTEMP\n",
      "KBA05_KRSAQUOT\n",
      "KBA05_KRSHERST1\n",
      "KBA05_KRSHERST2\n",
      "KBA05_KRSHERST3\n",
      "KBA05_KRSKLEIN\n",
      "KBA05_KRSOBER\n",
      "KBA05_KRSVAN\n",
      "KBA05_KRSZUL\n",
      "KBA05_KW1\n",
      "KBA05_KW2\n",
      "KBA05_KW3\n",
      "KBA05_MAXAH\n",
      "KBA05_MAXBJ\n",
      "KBA05_MAXHERST\n",
      "KBA05_MAXSEG\n",
      "KBA05_MAXVORB\n",
      "KBA05_MOD1\n",
      "KBA05_MOD2\n",
      "KBA05_MOD3\n",
      "KBA05_MOD4\n",
      "KBA05_MOD8\n",
      "KBA05_MODTEMP\n",
      "KBA05_MOTOR\n",
      "KBA05_MOTRAD\n",
      "KBA05_SEG1\n",
      "KBA05_SEG10\n",
      "KBA05_SEG2\n",
      "KBA05_SEG3\n",
      "KBA05_SEG4\n",
      "KBA05_SEG5\n",
      "KBA05_SEG6\n",
      "KBA05_SEG7\n",
      "KBA05_SEG8\n",
      "KBA05_SEG9\n",
      "KBA05_VORB0\n",
      "KBA05_VORB1\n",
      "KBA05_VORB2\n",
      "KBA05_ZUL1\n",
      "KBA05_ZUL2\n",
      "KBA05_ZUL3\n",
      "KBA05_ZUL4\n",
      "KBA13_ALTERHALTER_30\n",
      "KBA13_ALTERHALTER_45\n",
      "KBA13_ALTERHALTER_60\n",
      "KBA13_ALTERHALTER_61\n",
      "KBA13_ANTG1\n",
      "KBA13_ANTG2\n",
      "KBA13_ANTG3\n",
      "KBA13_ANTG4\n",
      "KBA13_ANZAHL_PKW\n",
      "KBA13_AUDI\n",
      "KBA13_AUTOQUOTE\n",
      "KBA13_BAUMAX\n",
      "KBA13_BJ_1999\n",
      "KBA13_BJ_2000\n",
      "KBA13_BJ_2004\n",
      "KBA13_BJ_2006\n",
      "KBA13_BJ_2008\n",
      "KBA13_BJ_2009\n",
      "KBA13_BMW\n",
      "KBA13_CCM_0_1400\n",
      "KBA13_CCM_1000\n",
      "KBA13_CCM_1200\n",
      "KBA13_CCM_1400\n",
      "KBA13_CCM_1401_2500\n",
      "KBA13_CCM_1500\n",
      "KBA13_CCM_1600\n",
      "KBA13_CCM_1800\n",
      "KBA13_CCM_2000\n",
      "KBA13_CCM_2500\n",
      "KBA13_CCM_2501\n",
      "KBA13_CCM_3000\n",
      "KBA13_CCM_3001\n",
      "KBA13_FAB_ASIEN\n",
      "KBA13_FAB_SONSTIGE\n",
      "KBA13_FIAT\n",
      "KBA13_FORD\n",
      "KBA13_GBZ\n",
      "KBA13_HALTER_20\n",
      "KBA13_HALTER_25\n",
      "KBA13_HALTER_30\n",
      "KBA13_HALTER_35\n",
      "KBA13_HALTER_40\n",
      "KBA13_HALTER_45\n",
      "KBA13_HALTER_50\n",
      "KBA13_HALTER_55\n",
      "KBA13_HALTER_60\n",
      "KBA13_HALTER_65\n",
      "KBA13_HALTER_66\n",
      "KBA13_HERST_ASIEN\n",
      "KBA13_HERST_AUDI_VW\n",
      "KBA13_HERST_BMW_BENZ\n",
      "KBA13_HERST_EUROPA\n",
      "KBA13_HERST_FORD_OPEL\n",
      "KBA13_HERST_SONST\n",
      "KBA13_HHZ\n",
      "KBA13_KMH_0_140\n",
      "KBA13_KMH_110\n",
      "KBA13_KMH_140\n",
      "KBA13_KMH_140_210\n",
      "KBA13_KMH_180\n",
      "KBA13_KMH_210\n",
      "KBA13_KMH_211\n",
      "KBA13_KMH_250\n",
      "KBA13_KMH_251\n",
      "KBA13_KRSAQUOT\n",
      "KBA13_KRSHERST_AUDI_VW\n",
      "KBA13_KRSHERST_BMW_BENZ\n",
      "KBA13_KRSHERST_FORD_OPEL\n",
      "KBA13_KRSSEG_KLEIN\n",
      "KBA13_KRSSEG_OBER\n",
      "KBA13_KRSSEG_VAN\n",
      "KBA13_KRSZUL_NEU\n",
      "KBA13_KW_0_60\n",
      "KBA13_KW_110\n",
      "KBA13_KW_120\n",
      "KBA13_KW_121\n",
      "KBA13_KW_30\n",
      "KBA13_KW_40\n",
      "KBA13_KW_50\n",
      "KBA13_KW_60\n",
      "KBA13_KW_61_120\n",
      "KBA13_KW_70\n",
      "KBA13_KW_80\n",
      "KBA13_KW_90\n",
      "KBA13_MAZDA\n",
      "KBA13_MERCEDES\n",
      "KBA13_MOTOR\n",
      "KBA13_NISSAN\n",
      "KBA13_OPEL\n",
      "KBA13_PEUGEOT\n",
      "KBA13_RENAULT\n",
      "KBA13_SEG_GELAENDEWAGEN\n",
      "KBA13_SEG_GROSSRAUMVANS\n",
      "KBA13_SEG_KLEINST\n",
      "KBA13_SEG_KLEINWAGEN\n",
      "KBA13_SEG_KOMPAKTKLASSE\n",
      "KBA13_SEG_MINIVANS\n",
      "KBA13_SEG_MINIWAGEN\n",
      "KBA13_SEG_MITTELKLASSE\n",
      "KBA13_SEG_OBEREMITTELKLASSE\n",
      "KBA13_SEG_OBERKLASSE\n",
      "KBA13_SEG_SONSTIGE\n",
      "KBA13_SEG_SPORTWAGEN\n",
      "KBA13_SEG_UTILITIES\n",
      "KBA13_SEG_VAN\n",
      "KBA13_SEG_WOHNMOBILE\n",
      "KBA13_SITZE_4\n",
      "KBA13_SITZE_5\n",
      "KBA13_SITZE_6\n",
      "KBA13_TOYOTA\n",
      "KBA13_VORB_0\n",
      "KBA13_VORB_1\n",
      "KBA13_VORB_1_2\n",
      "KBA13_VORB_2\n",
      "KBA13_VORB_3\n",
      "KBA13_VW\n",
      "KKK\n",
      "KOMBIALTER\n",
      "KONSUMNAEHE\n",
      "KONSUMZELLE\n",
      "LP_FAMILIE_FEIN\n",
      "LP_LEBENSPHASE_GROB\n",
      "LP_STATUS_FEIN\n",
      "MIN_GEBAEUDEJAHR\n",
      "MOBI_RASTER\n",
      "MOBI_REGIO\n",
      "NATIONALITAET_KZ\n",
      "ONLINE_AFFINITAET\n",
      "ORTSGR_KLS9\n",
      "PLZ8_ANTG1\n",
      "PLZ8_ANTG2\n",
      "PLZ8_ANTG3\n",
      "PLZ8_ANTG4\n",
      "PLZ8_BAUMAX\n",
      "PLZ8_GBZ\n",
      "PLZ8_HHZ\n",
      "REGIOTYP\n",
      "RELAT_AB\n",
      "RETOURTYP_BK_S\n",
      "RT_KEIN_ANREIZ\n",
      "RT_SCHNAEPPCHEN\n",
      "RT_UEBERGROESSE\n",
      "SEMIO_DOM\n",
      "SEMIO_ERL\n",
      "SEMIO_FAM\n",
      "SEMIO_KAEM\n",
      "SEMIO_KRIT\n",
      "SEMIO_KULT\n",
      "SEMIO_LUST\n",
      "SEMIO_MAT\n",
      "SEMIO_PFLICHT\n",
      "SEMIO_RAT\n",
      "SEMIO_REL\n",
      "SEMIO_SOZ\n",
      "SEMIO_TRADV\n",
      "SEMIO_VERT\n",
      "SHOPPER_TYP\n",
      "SOHO_KZ\n",
      "STRUKTURTYP\n",
      "UMFELD_ALT\n",
      "UMFELD_JUNG\n",
      "UNGLEICHENN_FLAG\n",
      "VERDICHTUNGSRAUM\n",
      "VERS_TYP\n",
      "VHA\n",
      "VHN\n",
      "VK_DHT4A\n",
      "VK_DISTANZ\n",
      "VK_ZG11\n",
      "W_KEIT_KIND_HH\n",
      "WOHNDAUER_2008\n",
      "WOHNLAGE\n",
      "ZABEOTYP\n",
      "ANREDE_KZ\n",
      "ALTERSKATEGORIE_GROB\n",
      "LNR\n",
      "AKT_DAT_KL\n",
      "ALTERSKATEGORIE_FEIN\n",
      "ANZ_HAUSHALTE_AKTIV\n",
      "ANZ_HH_TITEL\n",
      "ANZ_KINDER\n",
      "ANZ_PERSONEN\n",
      "ANZ_STATISTISCHE_HAUSHALTE\n",
      "ANZ_TITEL\n",
      "ARBEIT\n",
      "BALLRAUM\n",
      "CAMEO_DEUG_2015\n",
      "CJT_GESAMTTYP\n",
      "CJT_KATALOGNUTZER\n",
      "CJT_TYP_1\n",
      "CJT_TYP_2\n",
      "CJT_TYP_3\n",
      "CJT_TYP_4\n",
      "CJT_TYP_5\n",
      "CJT_TYP_6\n",
      "D19_BANKEN_ANZ_12\n",
      "D19_BANKEN_ANZ_24\n",
      "D19_BANKEN_DATUM\n",
      "D19_BANKEN_DIREKT\n",
      "D19_BANKEN_GROSS\n",
      "D19_BANKEN_LOKAL\n",
      "D19_BANKEN_OFFLINE_DATUM\n",
      "D19_BANKEN_ONLINE_DATUM\n",
      "D19_BANKEN_ONLINE_QUOTE_12\n",
      "D19_BANKEN_REST\n",
      "D19_BEKLEIDUNG_GEH\n",
      "D19_BEKLEIDUNG_REST\n",
      "D19_BILDUNG\n",
      "D19_BIO_OEKO\n",
      "D19_BUCH_CD\n",
      "D19_DIGIT_SERV\n",
      "D19_DROGERIEARTIKEL\n",
      "D19_ENERGIE\n",
      "D19_FREIZEIT\n",
      "D19_GARTEN\n",
      "D19_GESAMT_ANZ_12\n",
      "D19_GESAMT_ANZ_24\n",
      "D19_GESAMT_DATUM\n",
      "D19_GESAMT_OFFLINE_DATUM\n",
      "D19_GESAMT_ONLINE_DATUM\n",
      "D19_GESAMT_ONLINE_QUOTE_12\n",
      "D19_HANDWERK\n",
      "D19_HAUS_DEKO\n",
      "D19_KINDERARTIKEL\n",
      "D19_KONSUMTYP\n",
      "D19_KONSUMTYP_MAX\n",
      "D19_KOSMETIK\n",
      "D19_LEBENSMITTEL\n",
      "D19_LOTTO\n",
      "D19_NAHRUNGSERGAENZUNG\n",
      "D19_RATGEBER\n",
      "D19_REISEN\n",
      "D19_SAMMELARTIKEL\n",
      "D19_SCHUHE\n",
      "D19_SONSTIGE\n",
      "D19_SOZIALES\n",
      "D19_TECHNIK\n",
      "D19_TELKO_ANZ_12\n",
      "D19_TELKO_ANZ_24\n",
      "D19_TELKO_DATUM\n",
      "D19_TELKO_MOBILE\n",
      "D19_TELKO_OFFLINE_DATUM\n",
      "D19_TELKO_ONLINE_DATUM\n",
      "D19_TELKO_ONLINE_QUOTE_12\n",
      "D19_TELKO_REST\n",
      "D19_TIERARTIKEL\n",
      "D19_VERSAND_ANZ_12\n",
      "D19_VERSAND_ANZ_24\n",
      "D19_VERSAND_DATUM\n",
      "D19_VERSAND_OFFLINE_DATUM\n",
      "D19_VERSAND_ONLINE_DATUM\n",
      "D19_VERSAND_ONLINE_QUOTE_12\n",
      "D19_VERSAND_REST\n",
      "D19_VERSI_ANZ_12\n",
      "D19_VERSI_ANZ_24\n",
      "D19_VERSI_DATUM\n",
      "D19_VERSI_OFFLINE_DATUM\n",
      "D19_VERSI_ONLINE_DATUM\n",
      "D19_VERSI_ONLINE_QUOTE_12\n",
      "D19_VERSICHERUNGEN\n",
      "D19_VOLLSORTIMENT\n",
      "D19_WEIN_FEINKOST\n",
      "DSL_FLAG\n",
      "EINGEZOGENAM_HH_JAHR\n",
      "EWDICHTE\n",
      "FINANZ_ANLEGER\n",
      "FINANZ_HAUSBAUER\n",
      "FINANZ_MINIMALIST\n",
      "FINANZ_SPARER\n",
      "FINANZ_UNAUFFAELLIGER\n",
      "FINANZ_VORSORGER\n",
      "FINANZTYP\n",
      "FIRMENDICHTE\n",
      "GEBAEUDETYP\n",
      "GEBAEUDETYP_RASTER\n",
      "GEBURTSJAHR\n",
      "GEMEINDETYP\n",
      "GFK_URLAUBERTYP\n",
      "GREEN_AVANTGARDE\n",
      "HEALTH_TYP\n",
      "HH_DELTA_FLAG\n",
      "HH_EINKOMMEN_SCORE\n",
      "INNENSTADT\n",
      "KBA05_ALTER1\n",
      "KBA05_ALTER2\n",
      "KBA05_ALTER3\n",
      "KBA05_ALTER4\n",
      "KBA05_ANHANG\n",
      "KBA05_ANTG1\n",
      "KBA05_ANTG2\n",
      "KBA05_ANTG3\n",
      "KBA05_ANTG4\n",
      "KBA05_AUTOQUOT\n",
      "KBA05_CCM1\n",
      "KBA05_CCM2\n",
      "KBA05_CCM3\n",
      "KBA05_CCM4\n",
      "KBA05_DIESEL\n",
      "KBA05_FRAU\n",
      "KBA05_GBZ\n",
      "KBA05_HERST1\n",
      "KBA05_HERST2\n",
      "KBA05_HERST3\n",
      "KBA05_HERST4\n",
      "KBA05_HERST5\n",
      "KBA05_HERSTTEMP\n",
      "KBA05_KRSAQUOT\n",
      "KBA05_KRSHERST1\n",
      "KBA05_KRSHERST2\n",
      "KBA05_KRSHERST3\n",
      "KBA05_KRSKLEIN\n",
      "KBA05_KRSOBER\n",
      "KBA05_KRSVAN\n",
      "KBA05_KRSZUL\n",
      "KBA05_KW1\n",
      "KBA05_KW2\n",
      "KBA05_KW3\n",
      "KBA05_MAXAH\n",
      "KBA05_MAXBJ\n",
      "KBA05_MAXHERST\n",
      "KBA05_MAXSEG\n",
      "KBA05_MAXVORB\n",
      "KBA05_MOD1\n",
      "KBA05_MOD2\n",
      "KBA05_MOD3\n",
      "KBA05_MOD4\n",
      "KBA05_MOD8\n",
      "KBA05_MODTEMP\n",
      "KBA05_MOTOR\n",
      "KBA05_MOTRAD\n",
      "KBA05_SEG1\n",
      "KBA05_SEG10\n",
      "KBA05_SEG2\n",
      "KBA05_SEG3\n",
      "KBA05_SEG4\n",
      "KBA05_SEG5\n",
      "KBA05_SEG6\n",
      "KBA05_SEG7\n",
      "KBA05_SEG8\n",
      "KBA05_SEG9\n",
      "KBA05_VORB0\n",
      "KBA05_VORB1\n",
      "KBA05_VORB2\n",
      "KBA05_ZUL1\n",
      "KBA05_ZUL2\n",
      "KBA05_ZUL3\n",
      "KBA05_ZUL4\n",
      "KBA13_ALTERHALTER_30\n",
      "KBA13_ALTERHALTER_45\n",
      "KBA13_ALTERHALTER_60\n",
      "KBA13_ALTERHALTER_61\n",
      "KBA13_ANTG1\n",
      "KBA13_ANTG2\n",
      "KBA13_ANTG3\n",
      "KBA13_ANTG4\n",
      "KBA13_ANZAHL_PKW\n",
      "KBA13_AUDI\n",
      "KBA13_AUTOQUOTE\n",
      "KBA13_BAUMAX\n",
      "KBA13_BJ_1999\n",
      "KBA13_BJ_2000\n",
      "KBA13_BJ_2004\n",
      "KBA13_BJ_2006\n",
      "KBA13_BJ_2008\n",
      "KBA13_BJ_2009\n",
      "KBA13_BMW\n",
      "KBA13_CCM_0_1400\n",
      "KBA13_CCM_1000\n",
      "KBA13_CCM_1200\n",
      "KBA13_CCM_1400\n",
      "KBA13_CCM_1401_2500\n",
      "KBA13_CCM_1500\n",
      "KBA13_CCM_1600\n",
      "KBA13_CCM_1800\n",
      "KBA13_CCM_2000\n",
      "KBA13_CCM_2500\n",
      "KBA13_CCM_2501\n",
      "KBA13_CCM_3000\n",
      "KBA13_CCM_3001\n",
      "KBA13_FAB_ASIEN\n",
      "KBA13_FAB_SONSTIGE\n",
      "KBA13_FIAT\n",
      "KBA13_FORD\n",
      "KBA13_GBZ\n",
      "KBA13_HALTER_20\n",
      "KBA13_HALTER_25\n",
      "KBA13_HALTER_30\n",
      "KBA13_HALTER_35\n",
      "KBA13_HALTER_40\n",
      "KBA13_HALTER_45\n",
      "KBA13_HALTER_50\n",
      "KBA13_HALTER_55\n",
      "KBA13_HALTER_60\n",
      "KBA13_HALTER_65\n",
      "KBA13_HALTER_66\n",
      "KBA13_HERST_ASIEN\n",
      "KBA13_HERST_AUDI_VW\n",
      "KBA13_HERST_BMW_BENZ\n",
      "KBA13_HERST_EUROPA\n",
      "KBA13_HERST_FORD_OPEL\n",
      "KBA13_HERST_SONST\n",
      "KBA13_HHZ\n",
      "KBA13_KMH_0_140\n",
      "KBA13_KMH_110\n",
      "KBA13_KMH_140\n",
      "KBA13_KMH_140_210\n",
      "KBA13_KMH_180\n",
      "KBA13_KMH_210\n",
      "KBA13_KMH_211\n",
      "KBA13_KMH_250\n",
      "KBA13_KMH_251\n",
      "KBA13_KRSAQUOT\n",
      "KBA13_KRSHERST_AUDI_VW\n",
      "KBA13_KRSHERST_BMW_BENZ\n",
      "KBA13_KRSHERST_FORD_OPEL\n",
      "KBA13_KRSSEG_KLEIN\n",
      "KBA13_KRSSEG_OBER\n",
      "KBA13_KRSSEG_VAN\n",
      "KBA13_KRSZUL_NEU\n",
      "KBA13_KW_0_60\n",
      "KBA13_KW_110\n",
      "KBA13_KW_120\n",
      "KBA13_KW_121\n",
      "KBA13_KW_30\n",
      "KBA13_KW_40\n",
      "KBA13_KW_50\n",
      "KBA13_KW_60\n",
      "KBA13_KW_61_120\n",
      "KBA13_KW_70\n",
      "KBA13_KW_80\n",
      "KBA13_KW_90\n",
      "KBA13_MAZDA\n",
      "KBA13_MERCEDES\n",
      "KBA13_MOTOR\n",
      "KBA13_NISSAN\n",
      "KBA13_OPEL\n",
      "KBA13_PEUGEOT\n",
      "KBA13_RENAULT\n",
      "KBA13_SEG_GELAENDEWAGEN\n",
      "KBA13_SEG_GROSSRAUMVANS\n",
      "KBA13_SEG_KLEINST\n",
      "KBA13_SEG_KLEINWAGEN\n",
      "KBA13_SEG_KOMPAKTKLASSE\n",
      "KBA13_SEG_MINIVANS\n",
      "KBA13_SEG_MINIWAGEN\n",
      "KBA13_SEG_MITTELKLASSE\n",
      "KBA13_SEG_OBEREMITTELKLASSE\n",
      "KBA13_SEG_OBERKLASSE\n",
      "KBA13_SEG_SONSTIGE\n",
      "KBA13_SEG_SPORTWAGEN\n",
      "KBA13_SEG_UTILITIES\n",
      "KBA13_SEG_VAN\n",
      "KBA13_SEG_WOHNMOBILE\n",
      "KBA13_SITZE_4\n",
      "KBA13_SITZE_5\n",
      "KBA13_SITZE_6\n",
      "KBA13_TOYOTA\n",
      "KBA13_VORB_0\n",
      "KBA13_VORB_1\n",
      "KBA13_VORB_1_2\n",
      "KBA13_VORB_2\n",
      "KBA13_VORB_3\n",
      "KBA13_VW\n",
      "KKK\n",
      "KOMBIALTER\n",
      "KONSUMNAEHE\n",
      "KONSUMZELLE\n",
      "LP_FAMILIE_FEIN\n",
      "LP_LEBENSPHASE_GROB\n",
      "LP_STATUS_FEIN\n",
      "MIN_GEBAEUDEJAHR\n",
      "MOBI_RASTER\n",
      "MOBI_REGIO\n",
      "NATIONALITAET_KZ\n",
      "ONLINE_AFFINITAET\n",
      "ORTSGR_KLS9\n",
      "PLZ8_ANTG1\n",
      "PLZ8_ANTG2\n",
      "PLZ8_ANTG3\n",
      "PLZ8_ANTG4\n",
      "PLZ8_BAUMAX\n",
      "PLZ8_GBZ\n",
      "PLZ8_HHZ\n",
      "REGIOTYP\n",
      "RELAT_AB\n",
      "RETOURTYP_BK_S\n",
      "RT_KEIN_ANREIZ\n",
      "RT_SCHNAEPPCHEN\n",
      "RT_UEBERGROESSE\n",
      "SEMIO_DOM\n",
      "SEMIO_ERL\n",
      "SEMIO_FAM\n",
      "SEMIO_KAEM\n",
      "SEMIO_KRIT\n",
      "SEMIO_KULT\n",
      "SEMIO_LUST\n",
      "SEMIO_MAT\n",
      "SEMIO_PFLICHT\n",
      "SEMIO_RAT\n",
      "SEMIO_REL\n",
      "SEMIO_SOZ\n",
      "SEMIO_TRADV\n",
      "SEMIO_VERT\n",
      "SHOPPER_TYP\n",
      "SOHO_KZ\n",
      "STRUKTURTYP\n",
      "UMFELD_ALT\n",
      "UMFELD_JUNG\n",
      "UNGLEICHENN_FLAG\n",
      "VERDICHTUNGSRAUM\n",
      "VERS_TYP\n",
      "VHA\n",
      "VHN\n",
      "VK_DHT4A\n",
      "VK_DISTANZ\n",
      "VK_ZG11\n",
      "W_KEIT_KIND_HH\n",
      "WOHNDAUER_2008\n",
      "WOHNLAGE\n",
      "ZABEOTYP\n",
      "ANREDE_KZ\n",
      "ALTERSKATEGORIE_GROB\n"
     ]
    }
   ],
   "source": [
    "# Apply function\n",
    "customers_clean = data_cleaning(customers_new)\n",
    "azdias_clean = data_cleaning(azdias_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "3a9b2d3c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of datasets after dropping columns:\n",
      "Customers: (134246, 359)\n",
      "Azdias: (737288, 359)\n"
     ]
    }
   ],
   "source": [
    "# Check the shape of data after performing cleaning operation\n",
    "print(\"Shape of datasets after cleaning:\")\n",
    "print(f\"Customers: {customers_clean.shape}\")\n",
    "print(f\"Azdias: {azdias_clean.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "c7288fc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save the cleaned datasets\n",
    "azdias_clean.to_csv('E:\\\\udct\\\\Arvato Project\\\\azdias_clean.csv', index=False)\n",
    "customers_clean.to_csv('E:\\\\udct\\\\Arvato Project\\\\customers_clean.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b10b25a4",
   "metadata": {},
   "source": [
    "#### After cleaning the data, we can scale it to make sure all the features have the same range. Let's use StandardScaler from sklearn to scale the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "5a857da3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale the data\n",
    "scalar = StandardScaler()\n",
    "scalar.fit(azdias_clean)\n",
    "\n",
    "azdias_clean = pd.DataFrame(scalar.transform(azdias_clean), columns = azdias_clean.columns)\n",
    "customers_clean = pd.DataFrame(scalar.transform(customers_clean), columns = customers_clean.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "577aadc8",
   "metadata": {},
   "source": [
    "## Part 1: Customer Segmentation Report\n",
    "\n",
    "The main bulk of your analysis will come in this part of the project. Here, you should 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, you 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",
   "id": "25a994ad",
   "metadata": {},
   "source": [
    "#### Principal Component Analysis (PCA)\n",
    "PCA is used to reduce the dimension of the data with minimum loss of information. As we have around 360 columns in our dataset, it will be easier to perform unsupervised analysis on reduced dimensions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "9698c181",
   "metadata": {},
   "outputs": [],
   "source": [
    "# PCA\n",
    "pca = PCA(random_state=123)\n",
    "azdias_pca = pca.fit_transform(azdias_clean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "5aa70808",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "123"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of components required to maintain 80% variance:\n",
    "n_components = min(np.where(np.cumsum(pca.explained_variance_ratio_)>0.8)[0]+1)\n",
    "n_components"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "db13b21a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1e490c35f70>"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot a graph of principanl component w.r.t variance explained\n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_axes([0,0,1,1],True)\n",
    "ax2 = ax1.twinx()\n",
    "ax1.plot(pca.explained_variance_ratio_, label='Variance', color = 'orange')\n",
    "ax2.plot(np.cumsum(pca.explained_variance_ratio_), label='Cumulative Variance',color = 'Green');\n",
    "ax1.set_xlabel(\"Principal Component\", fontsize = 12)\n",
    "ax1.set_ylabel(\"Variance Explained (%)\", fontsize = 12)\n",
    "ax1.set_title('n_components needed for > 80% explained variance: {}'.format(n_components), fontsize = 14);\n",
    "ax1.axvline(n_components, linestyle='dashed', color='black')\n",
    "ax2.axhline(np.cumsum(pca.explained_variance_ratio_)[n_components], linestyle='dashed', color='black')\n",
    "fig.legend(loc=(0.6,0.2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb3f42dd",
   "metadata": {},
   "source": [
    "#### The graph shows that around 123 components explain 80% of variance in the dataset. Let's see the feature importances for each component to understand what each PCA component represents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "id": "9a2ccb76",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to print important features\n",
    "def print_important_features(features, components, i):\n",
    "    '''\n",
    "    This function prints important features of PCA\n",
    "    \n",
    "    Input: \n",
    "    feature - The columns in the dataframe \n",
    "    components - The PCA components\n",
    "    \n",
    "    Output:      \n",
    "    None  \n",
    "    '''\n",
    "    feature_weights = dict(zip(features, components[i]))\n",
    "    feature_weights_sorted = sorted(feature_weights.items(), key=lambda kv: kv[1])\n",
    "    print('Lowest:')\n",
    "    for feature, weight in feature_weights_sorted[:5]:\n",
    "        print('\\t{:20} \\t{:.3f}'.format(feature, weight))\n",
    "        \n",
    "    print('Highest:')\n",
    "    for feature, weight in feature_weights_sorted[-5:]:\n",
    "        print('\\t{:20} \\t{:.3f}'.format(feature, weight))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "74f906e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.18%\n",
      "PCA 1\n",
      "Lowest:\n",
      "\tD19_GESAMT_ONLINE_DATUM \t-0.132\n",
      "\tD19_VERSAND_ONLINE_DATUM \t-0.127\n",
      "\tD19_GESAMT_DATUM     \t-0.126\n",
      "\tD19_VERSAND_DATUM    \t-0.124\n",
      "\tFINANZ_VORSORGER     \t-0.121\n",
      "Highest:\n",
      "\tD19_VERSAND_ANZ_24   \t0.124\n",
      "\tPRAEGENDE_JUGENDJAHRE_decade \t0.126\n",
      "\tD19_GESAMT_ANZ_12    \t0.129\n",
      "\tD19_GESAMT_ANZ_24    \t0.137\n",
      "\tONLINE_AFFINITAET    \t0.144\n"
     ]
    }
   ],
   "source": [
    "# First principal component\n",
    "i = 1\n",
    "print('{:5.2%}'.format(pca.explained_variance_ratio_[i]))\n",
    "print(\"PCA 1\")\n",
    "\n",
    "print_important_features(azdias_clean.columns.values, pca.components_, i) "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fcbca27",
   "metadata": {},
   "source": [
    "#### Component 1 is determined by online affinity, a dominating movement in the person's youth, and the person's financial behavior."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "3f497dc7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4.96%\n",
      "PCA 2\n",
      "Lowest:\n",
      "\tKBA13_SITZE_5        \t-0.136\n",
      "\tKBA13_SEG_KLEINWAGEN \t-0.115\n",
      "\tKBA13_KMH_140_210    \t-0.114\n",
      "\tKBA13_HALTER_20      \t-0.109\n",
      "\tKBA13_HALTER_25      \t-0.105\n",
      "Highest:\n",
      "\tKBA13_SITZE_4        \t0.134\n",
      "\tKBA13_BMW            \t0.138\n",
      "\tKBA13_MERCEDES       \t0.139\n",
      "\tKBA13_SEG_OBEREMITTELKLASSE \t0.143\n",
      "\tKBA13_HERST_BMW_BENZ \t0.163\n"
     ]
    }
   ],
   "source": [
    "# Second principal component\n",
    "i = 2\n",
    "print('{:5.2%}'.format(pca.explained_variance_ratio_[i]))\n",
    "print(\"PCA 2\")\n",
    "\n",
    "print_important_features(azdias_clean.columns.values, pca.components_, i)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1de0aa73",
   "metadata": {},
   "source": [
    "#### Component 2 is determined by number of cars, and numbers of shares of small and upper class cars. \n",
    "Higher component relates to higher probability of people have shares of upper class cars(i.e BMW, MERCEDES). So, people who hold more and belongs to upper class and middle class are likey to become mail-order company's customer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "a9077303",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.88%\n",
      "PCA 3\n",
      "Lowest:\n",
      "\tPRAEGENDE_JUGENDJAHRE_decade \t-0.149\n",
      "\tFINANZ_SPARER        \t-0.148\n",
      "\tFINANZ_ANLEGER       \t-0.147\n",
      "\tCJT_TYP_1            \t-0.145\n",
      "\tFINANZ_UNAUFFAELLIGER \t-0.144\n",
      "Highest:\n",
      "\tCJT_TYP_4            \t0.110\n",
      "\tCJT_TYP_3            \t0.113\n",
      "\tCJT_TYP_5            \t0.126\n",
      "\tALTERSKATEGORIE_GROB \t0.130\n",
      "\tFINANZ_VORSORGER     \t0.135\n"
     ]
    }
   ],
   "source": [
    "# Third principal component\n",
    "i = 3\n",
    "print('{:5.2%}'.format(pca.explained_variance_ratio_[i]))\n",
    "print(\"PCA 3\")\n",
    "\n",
    "print_important_features(azdias_clean.columns.values, pca.components_, i)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f07222e0",
   "metadata": {},
   "source": [
    "#### Component 3 is determined by financial topology, person's age, and money savers.\n",
    "Higher component relates to higher probability of people being older in age, money savers and crazy shopper are likely to become mail-order company's customer."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "368d6bab",
   "metadata": {},
   "source": [
    "#### K-Means\n",
    "Let's use the K-Means Clustering algorithm to cluster the general population into different segments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "b2ec4262",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   Centroids         Score\n",
      "0         30  1.551900e+07\n",
      "1         25  1.676203e+07\n",
      "2         22  1.769625e+07\n",
      "3         20  1.842470e+07\n",
      "4         15  2.091128e+07\n",
      "5         11  2.391628e+07\n",
      "6         10  2.492669e+07\n",
      "7          5  3.445664e+07\n",
      "8          2  4.985598e+07\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Sum of Squared Errors')"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Function to calculate K-Means score for a centroid\n",
    "def elbow_method(data, center):\n",
    "    '''\n",
    "    This function calculates the K-Means score for a centroid\n",
    "    \n",
    "    Input:\n",
    "    data - Dataframe with reduced dimentions\n",
    "    center - The centroids\n",
    "    \n",
    "    Output:\n",
    "    score - The score of centroid\n",
    "    '''\n",
    "    kmeans = KMeans(n_clusters = center, n_init=10, max_iter=300, random_state=42)\n",
    "    model = kmeans.fit(data)\n",
    "    score = np.abs(model.score(data))\n",
    "    \n",
    "    return score\n",
    "\n",
    "# Run possible number of cluster centroid through K-means score function\n",
    "score_pca = []\n",
    "centroids = [2,5,10,11,15,20,22,25,30]\n",
    "\n",
    "for x in centroids:\n",
    "    score_pca.append(elbow_method(azdias_pca[:,:5],x))\n",
    "    \n",
    "# Print score report\n",
    "print(pd.DataFrame(sorted(zip(centroids, score_pca), key=lambda x: x[1]), columns=['Centroids','Score']))\n",
    "\n",
    "# Plot a graph\n",
    "plt.plot(centroids,score_pca, linestyle='-', linewidth=3, marker='o')\n",
    "plt.xlabel('K-centroids', fontsize = 12)\n",
    "plt.ylabel('Sum of Squared Errors', fontsize = 12)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d92b249",
   "metadata": {},
   "source": [
    "#### The plot shows that the error rate is decreasing less gradually after 10 clusters. As a result, we choose 10 as our final cluster number."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "42e719dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Perform pipeline for a selected cluster\n",
    "n_clusters = 10\n",
    "cluster_p = Pipeline([\n",
    "                        ('pca', PCA(n_components)),\n",
    "                        ('kmeans', KMeans(n_clusters=n_clusters, random_state=42) )\n",
    "                    ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "e24bfa79",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 2, 1, 7, 1, 2, 2, 5, 5, 3])"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predict the cluster for dataset\n",
    "cluster_p.fit(azdias_clean)\n",
    "azdias_predict = cluster_p.predict(azdias_clean)\n",
    "azdias_predict[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "8eba012f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the percentage and count covered by the cluster of the azdias data\n",
    "azdias_cluster = pd.Series(azdias_predict).value_counts().reset_index()\n",
    "azdias_cluster.columns   = ['Cluster', 'Count']\n",
    "azdias_cluster['Dataset'] = 'Azdias'\n",
    "azdias_cluster['Percentage']  = azdias_cluster['Count'] / azdias_clean.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "721d643b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Cluster</th>\n",
       "      <th>Count</th>\n",
       "      <th>Dataset</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>102742</td>\n",
       "      <td>Azdias</td>\n",
       "      <td>0.139351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>94685</td>\n",
       "      <td>Azdias</td>\n",
       "      <td>0.128423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>94385</td>\n",
       "      <td>Azdias</td>\n",
       "      <td>0.128016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>93217</td>\n",
       "      <td>Azdias</td>\n",
       "      <td>0.126432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>76843</td>\n",
       "      <td>Azdias</td>\n",
       "      <td>0.104224</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Cluster   Count Dataset  Percentage\n",
       "0        2  102742  Azdias    0.139351\n",
       "1        6   94685  Azdias    0.128423\n",
       "2        1   94385  Azdias    0.128016\n",
       "3        3   93217  Azdias    0.126432\n",
       "4        8   76843  Azdias    0.104224"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "azdias_cluster.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "df5163d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([5, 3, 1, 7, 5, 6, 5, 9, 5, 6])"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predict the cluster for dataset\n",
    "customers_predict = cluster_p.predict(customers_clean)\n",
    "customers_predict[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "0208987b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the percentage and count covered by the cluster of the customers data\n",
    "customers_cluster = pd.Series(customers_predict).value_counts().reset_index()\n",
    "customers_cluster.columns   = ['Cluster', 'Count']\n",
    "customers_cluster['Dataset'] = 'Customers'\n",
    "customers_cluster['Percentage']  = customers_cluster['Count'] / customers_clean.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "id": "3ef2c5ac",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Cluster</th>\n",
       "      <th>Count</th>\n",
       "      <th>Dataset</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>43913</td>\n",
       "      <td>Customers</td>\n",
       "      <td>0.327108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>29414</td>\n",
       "      <td>Customers</td>\n",
       "      <td>0.219105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>26719</td>\n",
       "      <td>Customers</td>\n",
       "      <td>0.199030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>15529</td>\n",
       "      <td>Customers</td>\n",
       "      <td>0.115676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9</td>\n",
       "      <td>6273</td>\n",
       "      <td>Customers</td>\n",
       "      <td>0.046728</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Cluster  Count    Dataset  Percentage\n",
       "0        5  43913  Customers    0.327108\n",
       "1        6  29414  Customers    0.219105\n",
       "2        1  26719  Customers    0.199030\n",
       "3        3  15529  Customers    0.115676\n",
       "4        9   6273  Customers    0.046728"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customers_cluster.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "134da0bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Azdias</th>\n",
       "      <th>Customers</th>\n",
       "      <th>Difference</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cluster</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.062343</td>\n",
       "      <td>0.024276</td>\n",
       "      <td>-0.038067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.128016</td>\n",
       "      <td>0.199030</td>\n",
       "      <td>0.071014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.139351</td>\n",
       "      <td>0.007948</td>\n",
       "      <td>-0.131403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.126432</td>\n",
       "      <td>0.115676</td>\n",
       "      <td>-0.010757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.070031</td>\n",
       "      <td>0.020127</td>\n",
       "      <td>-0.049904</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Azdias  Customers  Difference\n",
       "Cluster                                 \n",
       "0        0.062343   0.024276   -0.038067\n",
       "1        0.128016   0.199030    0.071014\n",
       "2        0.139351   0.007948   -0.131403\n",
       "3        0.126432   0.115676   -0.010757\n",
       "4        0.070031   0.020127   -0.049904"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find the difference in the clusters for both datasets\n",
    "Cluster_Diff_Perc = pd.concat([azdias_cluster.groupby(['Cluster']).agg({'Percentage': ['sum']}), \n",
    "                      customers_cluster.groupby(['Cluster']).agg({'Percentage': ['sum']}) ], axis = 1)\n",
    "\n",
    "Cluster_Diff_Perc.columns = ['Azdias', 'Customers']\n",
    "Cluster_Diff_Perc['Difference'] = Cluster_Diff_Perc['Customers'] - Cluster_Diff_Perc['Azdias']\n",
    "Cluster_Diff_Perc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "a97a6cef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Comparision of the percentage of customer data to the azdias.\n",
    "Dataset_Diff = pd.concat([azdias_cluster, customers_cluster], axis = 0, sort = True) \n",
    "fig = plt.figure()\n",
    "sns.barplot(x = 'Cluster', y = 'Percentage', hue = 'Dataset', data = Dataset_Diff)\n",
    "plt.title(\"Azdias Vs Customers Distribution Plot\", fontsize = 14)\n",
    "plt.xlabel(\"Clusters\", fontsize = 12)\n",
    "plt.ylabel(\"Percentage\", fontsize = 12)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45cf22c1",
   "metadata": {},
   "source": [
    "#### Based on cluster distribution histogram on both datasets, it is clear to see that Cluster 5 is outstanding for customers, which means people in this group are more likely to be part of the mail-order company’s main customer base than other groups.\n",
    "Also, we noticed following characteristics of such people:\n",
    "1. Financially Capable or good\n",
    "2. Holds higher number of shares\n",
    "3. Higher online/transaction activities or crazy shopper\n",
    "4. Older in age\n",
    "5. Money savers"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6151d8b1",
   "metadata": {},
   "source": [
    "## Part 2: Supervised Learning Model\n",
    "\n",
    "Now that you'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, you'll need to create predictions on the \"TEST\" partition, where the \"RESPONSE\" column has been withheld."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "id": "a4f18872",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3444: DtypeWarning: Columns (18,19) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  exec(code_obj, self.user_global_ns, self.user_ns)\n"
     ]
    }
   ],
   "source": [
    "# Load the data\n",
    "mailout_train = pd.read_csv('E:\\\\udct\\\\Arvato Project\\\\Udacity_MAILOUT_052018_TRAIN.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "id": "5738eb51",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>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": 298,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mailout_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "id": "9f2d4552",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(42962, 367)\n"
     ]
    }
   ],
   "source": [
    "# Check the shape of data\n",
    "print(mailout_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "id": "372bc7ab",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rows in customers set that are not in mailout: ['PRODUCT_GROUP', 'CUSTOMER_GROUP', 'ONLINE_PURCHASE']\n",
      "Rows in mailout set that are not in azdias: ['RESPONSE']\n"
     ]
    }
   ],
   "source": [
    "# Find out the uniqueness in the data\n",
    "unique_cols_a = list(set(customers.columns.values) - set(mailout_train.columns.values))\n",
    "unique_cols_b = list(set(mailout_train.columns.values) - set(azdias.columns.values))\n",
    "\n",
    "print(\"Rows in customers set that are not in mailout: \" + str(unique_cols_a))\n",
    "print(\"Rows in mailout set that are not in azdias: \" + str(unique_cols_b))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "889f3e88",
   "metadata": {},
   "source": [
    "#### Data Pre-processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "id": "c87516d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to clean the data to use it for supervised model\n",
    "def data_cleaning_supervised_model(df):\n",
    "    '''\n",
    "    This function prepares the MAILOUT_TRAIN dataset for training under the supervised model. \n",
    "    \n",
    "    Input: \n",
    "    df -  Dataframe to perform cleaning operation\n",
    "    \n",
    "    Output\n",
    "    df - The transformed dataframe\n",
    "    '''\n",
    "    print('Format mixed-types \\n')      \n",
    "    df = fix_cameo(df)\n",
    "    \n",
    "    print('Replace unknowns \\n')   \n",
    "    replace_unknown(df, dict_unknown)\n",
    "    additional_missing_values = {'KBA05_MODTEMP': [6.0], 'LP_FAMILIE_FEIN': [0.0], 'LP_FAMILIE_GROB': [0.0],\n",
    "                                 'LP_LEBENSPHASE_FEIN': [0.0], 'LP_LEBENSPHASE_GROB': [0.0], 'ORTSGR_KLS9': [0.0], \n",
    "                                 'GEBURTSJAHR': [0]}\n",
    "    \n",
    "    replace_unknown(df, additional_missing_values)\n",
    "    \n",
    "    data_cleaning(df)\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "id": "1e997657",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Format mixed-types \n",
      "\n",
      "Replace unknowns \n",
      "\n",
      "LNR\n",
      "AGER_TYP\n",
      "AKT_DAT_KL\n",
      "ALTER_HH\n",
      "ALTER_KIND1\n",
      "ALTER_KIND2\n",
      "ALTER_KIND3\n",
      "ALTER_KIND4\n",
      "ALTERSKATEGORIE_FEIN\n",
      "ANZ_HAUSHALTE_AKTIV\n",
      "ANZ_HH_TITEL\n",
      "ANZ_KINDER\n",
      "ANZ_PERSONEN\n",
      "ANZ_STATISTISCHE_HAUSHALTE\n",
      "ANZ_TITEL\n",
      "ARBEIT\n",
      "BALLRAUM\n",
      "CAMEO_DEUG_2015\n",
      "CJT_GESAMTTYP\n",
      "CJT_KATALOGNUTZER\n",
      "CJT_TYP_1\n",
      "CJT_TYP_2\n",
      "CJT_TYP_3\n",
      "CJT_TYP_4\n",
      "CJT_TYP_5\n",
      "CJT_TYP_6\n",
      "D19_BANKEN_ANZ_12\n",
      "D19_BANKEN_ANZ_24\n",
      "D19_BANKEN_DATUM\n",
      "D19_BANKEN_DIREKT\n",
      "D19_BANKEN_GROSS\n",
      "D19_BANKEN_LOKAL\n",
      "D19_BANKEN_OFFLINE_DATUM\n",
      "D19_BANKEN_ONLINE_DATUM\n",
      "D19_BANKEN_ONLINE_QUOTE_12\n",
      "D19_BANKEN_REST\n",
      "D19_BEKLEIDUNG_GEH\n",
      "D19_BEKLEIDUNG_REST\n",
      "D19_BILDUNG\n",
      "D19_BIO_OEKO\n",
      "D19_BUCH_CD\n",
      "D19_DIGIT_SERV\n",
      "D19_DROGERIEARTIKEL\n",
      "D19_ENERGIE\n",
      "D19_FREIZEIT\n",
      "D19_GARTEN\n",
      "D19_GESAMT_ANZ_12\n",
      "D19_GESAMT_ANZ_24\n",
      "D19_GESAMT_DATUM\n",
      "D19_GESAMT_OFFLINE_DATUM\n",
      "D19_GESAMT_ONLINE_DATUM\n",
      "D19_GESAMT_ONLINE_QUOTE_12\n",
      "D19_HANDWERK\n",
      "D19_HAUS_DEKO\n",
      "D19_KINDERARTIKEL\n",
      "D19_KONSUMTYP\n",
      "D19_KONSUMTYP_MAX\n",
      "D19_KOSMETIK\n",
      "D19_LEBENSMITTEL\n",
      "D19_LOTTO\n",
      "D19_NAHRUNGSERGAENZUNG\n",
      "D19_RATGEBER\n",
      "D19_REISEN\n",
      "D19_SAMMELARTIKEL\n",
      "D19_SCHUHE\n",
      "D19_SONSTIGE\n",
      "D19_SOZIALES\n",
      "D19_TECHNIK\n",
      "D19_TELKO_ANZ_12\n",
      "D19_TELKO_ANZ_24\n",
      "D19_TELKO_DATUM\n",
      "D19_TELKO_MOBILE\n",
      "D19_TELKO_OFFLINE_DATUM\n",
      "D19_TELKO_ONLINE_DATUM\n",
      "D19_TELKO_ONLINE_QUOTE_12\n",
      "D19_TELKO_REST\n",
      "D19_TIERARTIKEL\n",
      "D19_VERSAND_ANZ_12\n",
      "D19_VERSAND_ANZ_24\n",
      "D19_VERSAND_DATUM\n",
      "D19_VERSAND_OFFLINE_DATUM\n",
      "D19_VERSAND_ONLINE_DATUM\n",
      "D19_VERSAND_ONLINE_QUOTE_12\n",
      "D19_VERSAND_REST\n",
      "D19_VERSI_ANZ_12\n",
      "D19_VERSI_ANZ_24\n",
      "D19_VERSI_DATUM\n",
      "D19_VERSI_OFFLINE_DATUM\n",
      "D19_VERSI_ONLINE_DATUM\n",
      "D19_VERSI_ONLINE_QUOTE_12\n",
      "D19_VERSICHERUNGEN\n",
      "D19_VOLLSORTIMENT\n",
      "D19_WEIN_FEINKOST\n",
      "DSL_FLAG\n",
      "EINGEZOGENAM_HH_JAHR\n",
      "EWDICHTE\n",
      "EXTSEL992\n",
      "FINANZ_ANLEGER\n",
      "FINANZ_HAUSBAUER\n",
      "FINANZ_MINIMALIST\n",
      "FINANZ_SPARER\n",
      "FINANZ_UNAUFFAELLIGER\n",
      "FINANZ_VORSORGER\n",
      "FINANZTYP\n",
      "FIRMENDICHTE\n",
      "GEBAEUDETYP\n",
      "GEBAEUDETYP_RASTER\n",
      "GEBURTSJAHR\n",
      "GEMEINDETYP\n",
      "GFK_URLAUBERTYP\n",
      "GREEN_AVANTGARDE\n",
      "HEALTH_TYP\n",
      "HH_DELTA_FLAG\n",
      "HH_EINKOMMEN_SCORE\n",
      "INNENSTADT\n",
      "KBA05_ALTER1\n",
      "KBA05_ALTER2\n",
      "KBA05_ALTER3\n",
      "KBA05_ALTER4\n",
      "KBA05_ANHANG\n",
      "KBA05_ANTG1\n",
      "KBA05_ANTG2\n",
      "KBA05_ANTG3\n",
      "KBA05_ANTG4\n",
      "KBA05_AUTOQUOT\n",
      "KBA05_BAUMAX\n",
      "KBA05_CCM1\n",
      "KBA05_CCM2\n",
      "KBA05_CCM3\n",
      "KBA05_CCM4\n",
      "KBA05_DIESEL\n",
      "KBA05_FRAU\n",
      "KBA05_GBZ\n",
      "KBA05_HERST1\n",
      "KBA05_HERST2\n",
      "KBA05_HERST3\n",
      "KBA05_HERST4\n",
      "KBA05_HERST5\n",
      "KBA05_HERSTTEMP\n",
      "KBA05_KRSAQUOT\n",
      "KBA05_KRSHERST1\n",
      "KBA05_KRSHERST2\n",
      "KBA05_KRSHERST3\n",
      "KBA05_KRSKLEIN\n",
      "KBA05_KRSOBER\n",
      "KBA05_KRSVAN\n",
      "KBA05_KRSZUL\n",
      "KBA05_KW1\n",
      "KBA05_KW2\n",
      "KBA05_KW3\n",
      "KBA05_MAXAH\n",
      "KBA05_MAXBJ\n",
      "KBA05_MAXHERST\n",
      "KBA05_MAXSEG\n",
      "KBA05_MAXVORB\n",
      "KBA05_MOD1\n",
      "KBA05_MOD2\n",
      "KBA05_MOD3\n",
      "KBA05_MOD4\n",
      "KBA05_MOD8\n",
      "KBA05_MODTEMP\n",
      "KBA05_MOTOR\n",
      "KBA05_MOTRAD\n",
      "KBA05_SEG1\n",
      "KBA05_SEG10\n",
      "KBA05_SEG2\n",
      "KBA05_SEG3\n",
      "KBA05_SEG4\n",
      "KBA05_SEG5\n",
      "KBA05_SEG6\n",
      "KBA05_SEG7\n",
      "KBA05_SEG8\n",
      "KBA05_SEG9\n",
      "KBA05_VORB0\n",
      "KBA05_VORB1\n",
      "KBA05_VORB2\n",
      "KBA05_ZUL1\n",
      "KBA05_ZUL2\n",
      "KBA05_ZUL3\n",
      "KBA05_ZUL4\n",
      "KBA13_ALTERHALTER_30\n",
      "KBA13_ALTERHALTER_45\n",
      "KBA13_ALTERHALTER_60\n",
      "KBA13_ALTERHALTER_61\n",
      "KBA13_ANTG1\n",
      "KBA13_ANTG2\n",
      "KBA13_ANTG3\n",
      "KBA13_ANTG4\n",
      "KBA13_ANZAHL_PKW\n",
      "KBA13_AUDI\n",
      "KBA13_AUTOQUOTE\n",
      "KBA13_BAUMAX\n",
      "KBA13_BJ_1999\n",
      "KBA13_BJ_2000\n",
      "KBA13_BJ_2004\n",
      "KBA13_BJ_2006\n",
      "KBA13_BJ_2008\n",
      "KBA13_BJ_2009\n",
      "KBA13_BMW\n",
      "KBA13_CCM_0_1400\n",
      "KBA13_CCM_1000\n",
      "KBA13_CCM_1200\n",
      "KBA13_CCM_1400\n",
      "KBA13_CCM_1401_2500\n",
      "KBA13_CCM_1500\n",
      "KBA13_CCM_1600\n",
      "KBA13_CCM_1800\n",
      "KBA13_CCM_2000\n",
      "KBA13_CCM_2500\n",
      "KBA13_CCM_2501\n",
      "KBA13_CCM_3000\n",
      "KBA13_CCM_3001\n",
      "KBA13_FAB_ASIEN\n",
      "KBA13_FAB_SONSTIGE\n",
      "KBA13_FIAT\n",
      "KBA13_FORD\n",
      "KBA13_GBZ\n",
      "KBA13_HALTER_20\n",
      "KBA13_HALTER_25\n",
      "KBA13_HALTER_30\n",
      "KBA13_HALTER_35\n",
      "KBA13_HALTER_40\n",
      "KBA13_HALTER_45\n",
      "KBA13_HALTER_50\n",
      "KBA13_HALTER_55\n",
      "KBA13_HALTER_60\n",
      "KBA13_HALTER_65\n",
      "KBA13_HALTER_66\n",
      "KBA13_HERST_ASIEN\n",
      "KBA13_HERST_AUDI_VW\n",
      "KBA13_HERST_BMW_BENZ\n",
      "KBA13_HERST_EUROPA\n",
      "KBA13_HERST_FORD_OPEL\n",
      "KBA13_HERST_SONST\n",
      "KBA13_HHZ\n",
      "KBA13_KMH_0_140\n",
      "KBA13_KMH_110\n",
      "KBA13_KMH_140\n",
      "KBA13_KMH_140_210\n",
      "KBA13_KMH_180\n",
      "KBA13_KMH_210\n",
      "KBA13_KMH_211\n",
      "KBA13_KMH_250\n",
      "KBA13_KMH_251\n",
      "KBA13_KRSAQUOT\n",
      "KBA13_KRSHERST_AUDI_VW\n",
      "KBA13_KRSHERST_BMW_BENZ\n",
      "KBA13_KRSHERST_FORD_OPEL\n",
      "KBA13_KRSSEG_KLEIN\n",
      "KBA13_KRSSEG_OBER\n",
      "KBA13_KRSSEG_VAN\n",
      "KBA13_KRSZUL_NEU\n",
      "KBA13_KW_0_60\n",
      "KBA13_KW_110\n",
      "KBA13_KW_120\n",
      "KBA13_KW_121\n",
      "KBA13_KW_30\n",
      "KBA13_KW_40\n",
      "KBA13_KW_50\n",
      "KBA13_KW_60\n",
      "KBA13_KW_61_120\n",
      "KBA13_KW_70\n",
      "KBA13_KW_80\n",
      "KBA13_KW_90\n",
      "KBA13_MAZDA\n",
      "KBA13_MERCEDES\n",
      "KBA13_MOTOR\n",
      "KBA13_NISSAN\n",
      "KBA13_OPEL\n",
      "KBA13_PEUGEOT\n",
      "KBA13_RENAULT\n",
      "KBA13_SEG_GELAENDEWAGEN\n",
      "KBA13_SEG_GROSSRAUMVANS\n",
      "KBA13_SEG_KLEINST\n",
      "KBA13_SEG_KLEINWAGEN\n",
      "KBA13_SEG_KOMPAKTKLASSE\n",
      "KBA13_SEG_MINIVANS\n",
      "KBA13_SEG_MINIWAGEN\n",
      "KBA13_SEG_MITTELKLASSE\n",
      "KBA13_SEG_OBEREMITTELKLASSE\n",
      "KBA13_SEG_OBERKLASSE\n",
      "KBA13_SEG_SONSTIGE\n",
      "KBA13_SEG_SPORTWAGEN\n",
      "KBA13_SEG_UTILITIES\n",
      "KBA13_SEG_VAN\n",
      "KBA13_SEG_WOHNMOBILE\n",
      "KBA13_SITZE_4\n",
      "KBA13_SITZE_5\n",
      "KBA13_SITZE_6\n",
      "KBA13_TOYOTA\n",
      "KBA13_VORB_0\n",
      "KBA13_VORB_1\n",
      "KBA13_VORB_1_2\n",
      "KBA13_VORB_2\n",
      "KBA13_VORB_3\n",
      "KBA13_VW\n",
      "KK_KUNDENTYP\n",
      "KKK\n",
      "KOMBIALTER\n",
      "KONSUMNAEHE\n",
      "KONSUMZELLE\n",
      "LP_FAMILIE_FEIN\n",
      "LP_LEBENSPHASE_GROB\n",
      "LP_STATUS_FEIN\n",
      "MIN_GEBAEUDEJAHR\n",
      "MOBI_RASTER\n",
      "MOBI_REGIO\n",
      "NATIONALITAET_KZ\n",
      "ONLINE_AFFINITAET\n",
      "ORTSGR_KLS9\n",
      "PLZ8_ANTG1\n",
      "PLZ8_ANTG2\n",
      "PLZ8_ANTG3\n",
      "PLZ8_ANTG4\n",
      "PLZ8_BAUMAX\n",
      "PLZ8_GBZ\n",
      "PLZ8_HHZ\n",
      "REGIOTYP\n",
      "RELAT_AB\n",
      "RETOURTYP_BK_S\n",
      "RT_KEIN_ANREIZ\n",
      "RT_SCHNAEPPCHEN\n",
      "RT_UEBERGROESSE\n",
      "SEMIO_DOM\n",
      "SEMIO_ERL\n",
      "SEMIO_FAM\n",
      "SEMIO_KAEM\n",
      "SEMIO_KRIT\n",
      "SEMIO_KULT\n",
      "SEMIO_LUST\n",
      "SEMIO_MAT\n",
      "SEMIO_PFLICHT\n",
      "SEMIO_RAT\n",
      "SEMIO_REL\n",
      "SEMIO_SOZ\n",
      "SEMIO_TRADV\n",
      "SEMIO_VERT\n",
      "SHOPPER_TYP\n",
      "SOHO_KZ\n",
      "STRUKTURTYP\n",
      "TITEL_KZ\n",
      "UMFELD_ALT\n",
      "UMFELD_JUNG\n",
      "UNGLEICHENN_FLAG\n",
      "VERDICHTUNGSRAUM\n",
      "VERS_TYP\n",
      "VHA\n",
      "VHN\n",
      "VK_DHT4A\n",
      "VK_DISTANZ\n",
      "VK_ZG11\n",
      "W_KEIT_KIND_HH\n",
      "WOHNDAUER_2008\n",
      "WOHNLAGE\n",
      "ZABEOTYP\n",
      "RESPONSE\n",
      "ANREDE_KZ\n",
      "ALTERSKATEGORIE_GROB\n"
     ]
    }
   ],
   "source": [
    "# Apply the function\n",
    "mailout_train_clean = data_cleaning_supervised_model(mailout_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "id": "3f5022ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop the columns which are not required\n",
    "drop_cols = ['ALTER_KIND4','ALTER_KIND3','TITEL_KZ','ALTER_KIND2','ALTER_KIND1',\n",
    "                 'KK_KUNDENTYP','KBA05_BAUMAX','AGER_TYP','EXTSEL992','ALTER_HH']\n",
    "\n",
    "mailout_train_clean = mailout_train_clean.drop(drop_cols, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "id": "9b2e6116",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42962, 360)"
      ]
     },
     "execution_count": 304,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the shape of cleaned data\n",
    "mailout_train_clean.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "id": "4f64580a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    42430\n",
      "1      532\n",
      "Name: RESPONSE, dtype: int64\n",
      "0    0.987617\n",
      "1    0.012383\n",
      "Name: RESPONSE, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Inspect the column response\n",
    "print(mailout_train_clean.RESPONSE.value_counts())\n",
    "print(mailout_train_clean.RESPONSE.value_counts() / np.sum(mailout_train_clean.RESPONSE.value_counts()))\n",
    "sns.countplot(x = \"RESPONSE\", data=mailout_train);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f50962f5",
   "metadata": {},
   "source": [
    "The data is highly imbalanced. We have more thn 40k - 0 responses and only 532 - 1 responses. So, even if the model predicts zero for all the data points we will end up with having accuracy 98.7% which does not make sense in this case.\n",
    "\n",
    "To address this imbalance while evaluating the model we need to chose a metric which will take this class imbalance into accounts. The usual metrics used for imbalanced classification are Precision and Recall or or Area under Receiver Operating Curve (AUROC)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37da3023",
   "metadata": {},
   "source": [
    "#### Let's train a model on unscaled data to check performance of various models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "id": "797663b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = mailout_train_clean.iloc[:, mailout_train_clean.columns != 'RESPONSE']\n",
    "y_train = mailout_train_clean['RESPONSE']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "id": "4e9caab7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale the data\n",
    "scaler_model = StandardScaler()\n",
    "X_train = scaler_model.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "id": "36d63c2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the dataset into Train and Test set\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "id": "d8a9401f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to fit the method\n",
    "def train_model(model, X_train, y_train, X_val, y_val):\n",
    "    '''\n",
    "    This function is the customization of the fit method.\n",
    "    \n",
    "    Input: \n",
    "    model - The model from the list of the classifiers/models\n",
    "    X_train - The training data\n",
    "    y_train - The training labels\n",
    "    X_val - The validation of data\n",
    "    y_val - The validation of labels\n",
    "    \n",
    "    Output: \n",
    "    ROC-AUC score \n",
    "    Training time\n",
    "    '''\n",
    "    t = time.time()\n",
    "    model = model.fit(X_train, y_train)\n",
    "    y_pred = model.predict_proba(X_val)[:,1]\n",
    "    \n",
    "    roc_score = roc_auc_score(y_val, y_pred)\n",
    "\n",
    "    train_time = time.time() - t\n",
    "    \n",
    "    return roc_score, train_time "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "id": "ba1563ed",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Models to check for performance\n",
    "models = [(\"LogisticRegression\", LogisticRegression(random_state=42)),\n",
    "         (\"DecisionTreeClassifier\", DecisionTreeClassifier(random_state=42)),\n",
    "         (\"RandomForestClassifier\", RandomForestClassifier(random_state=42)),\n",
    "         (\"GradientBoostingClassifier\", GradientBoostingClassifier(random_state=42)),\n",
    "         (\"AdaBoostClassifier\", AdaBoostClassifier(random_state=42)),\n",
    "         (\"XGBClassifier\",xgb.XGBClassifier(random_state=42))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "id": "c4248a4d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to iterate over classifier and fir method\n",
    "def test_model(classifiers, X_train, y_train, X_val, y_val):\n",
    "    '''\n",
    "    This function iterates over declared classifiers and calls the custom fit method.\n",
    "    \n",
    "    Input: \n",
    "    classifiers - The list of the classifiers/models\n",
    "    X_train - The training data\n",
    "    y_train - The training labels\n",
    "    X_val - The validation data\n",
    "    y_val - The validation labels\n",
    "    \n",
    "    Output: \n",
    "    df - The resulting dataframe for comparison\n",
    "    '''\n",
    "    result={\n",
    "    'Classifier':[],\n",
    "    'Score':[],\n",
    "    'Train Time':[]\n",
    "    }\n",
    "    \n",
    "    for name, classifier in classifiers:\n",
    "        score, t = train_model(classifier, X_train, y_train, X_val, y_val)\n",
    "        result['Classifier'].append(name)\n",
    "        result['Score'].append(score)\n",
    "        result['Train Time'].append(t)\n",
    "    \n",
    "    results_df = pd.DataFrame.from_dict(result, orient='index').transpose()\n",
    "    \n",
    "    return results_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "id": "fbcdf9f6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n",
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\xgboost\\sklearn.py:1224: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
      "  warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[13:55:59] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.5.1/src/learner.cc:1115: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\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>Classifier</th>\n",
       "      <th>Score</th>\n",
       "      <th>Train Time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>0.663947</td>\n",
       "      <td>1.52789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>0.520014</td>\n",
       "      <td>3.741404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>13.891999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GradientBoostingClassifier</td>\n",
       "      <td>0.745092</td>\n",
       "      <td>112.122969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AdaBoostClassifier</td>\n",
       "      <td>0.723016</td>\n",
       "      <td>25.895756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>XGBClassifier</td>\n",
       "      <td>0.654841</td>\n",
       "      <td>22.680465</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Classifier     Score  Train Time\n",
       "0          LogisticRegression  0.663947     1.52789\n",
       "1      DecisionTreeClassifier  0.520014    3.741404\n",
       "2      RandomForestClassifier  0.594675   13.891999\n",
       "3  GradientBoostingClassifier  0.745092  112.122969\n",
       "4          AdaBoostClassifier  0.723016   25.895756\n",
       "5               XGBClassifier  0.654841   22.680465"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Apply the function\n",
    "test_model(models, X_train, y_train, X_val, y_val)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2851be36",
   "metadata": {},
   "source": [
    "#### The above table shows that the GradientBoostingClassifier has the highest score, but it is the slowest to train. The AdaBoostClasifier have the next highest score and also took less time to train. So, let's select these algorithms for hyper parameter tuning."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58c200ef",
   "metadata": {},
   "source": [
    "### Grid Search\n",
    "\n",
    "#### GradientBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 369,
   "id": "e52aea0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Grid serach for GradientBoostClassifier model\n",
    "gbc = GradientBoostingClassifier(random_state=42)\n",
    "\n",
    "parameters = {'learning_rate': [0.1, 0.2], \n",
    "              'n_estimators': [100],\n",
    "              'max_depth': [3, 5], \n",
    "              'min_samples_split': [2, 4]\n",
    "             } "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "id": "d652e317",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 8 candidates, totalling 40 fits\n",
      "[CV 1/5; 1/8] START learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 1/5; 1/8] END learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.801 total time= 1.5min\n",
      "[CV 2/5; 1/8] START learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 2/5; 1/8] END learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.710 total time= 1.5min\n",
      "[CV 3/5; 1/8] START learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 3/5; 1/8] END learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.748 total time= 1.5min\n",
      "[CV 4/5; 1/8] START learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 4/5; 1/8] END learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.755 total time= 1.5min\n",
      "[CV 5/5; 1/8] START learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 5/5; 1/8] END learning_rate=0.1, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.736 total time= 1.5min\n",
      "[CV 1/5; 2/8] START learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 1/5; 2/8] END learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.804 total time= 1.5min\n",
      "[CV 2/5; 2/8] START learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 2/5; 2/8] END learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.708 total time= 1.5min\n",
      "[CV 3/5; 2/8] START learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 3/5; 2/8] END learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.746 total time= 1.4min\n",
      "[CV 4/5; 2/8] START learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 4/5; 2/8] END learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.757 total time= 1.5min\n",
      "[CV 5/5; 2/8] START learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 5/5; 2/8] END learning_rate=0.1, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.751 total time= 1.5min\n",
      "[CV 1/5; 3/8] START learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 1/5; 3/8] END learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.777 total time= 2.3min\n",
      "[CV 2/5; 3/8] START learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 2/5; 3/8] END learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.684 total time= 2.3min\n",
      "[CV 3/5; 3/8] START learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 3/5; 3/8] END learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.742 total time= 2.3min\n",
      "[CV 4/5; 3/8] START learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 4/5; 3/8] END learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.733 total time= 2.3min\n",
      "[CV 5/5; 3/8] START learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 5/5; 3/8] END learning_rate=0.1, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.730 total time= 2.3min\n",
      "[CV 1/5; 4/8] START learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 1/5; 4/8] END learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.788 total time= 2.3min\n",
      "[CV 2/5; 4/8] START learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 2/5; 4/8] END learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.691 total time= 2.3min\n",
      "[CV 3/5; 4/8] START learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 3/5; 4/8] END learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.738 total time= 2.3min\n",
      "[CV 4/5; 4/8] START learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 4/5; 4/8] END learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.733 total time= 2.3min\n",
      "[CV 5/5; 4/8] START learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 5/5; 4/8] END learning_rate=0.1, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.726 total time= 2.3min\n",
      "[CV 1/5; 5/8] START learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 1/5; 5/8] END learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.770 total time= 1.4min\n",
      "[CV 2/5; 5/8] START learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 2/5; 5/8] END learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.680 total time= 1.4min\n",
      "[CV 3/5; 5/8] START learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 3/5; 5/8] END learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.740 total time= 1.4min\n",
      "[CV 4/5; 5/8] START learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 4/5; 5/8] END learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.732 total time= 1.5min\n",
      "[CV 5/5; 5/8] START learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100\n",
      "[CV 5/5; 5/8] END learning_rate=0.2, max_depth=3, min_samples_split=2, n_estimators=100;, score=0.708 total time= 1.4min\n",
      "[CV 1/5; 6/8] START learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 1/5; 6/8] END learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.766 total time= 1.5min\n",
      "[CV 2/5; 6/8] START learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 2/5; 6/8] END learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.674 total time= 1.4min\n",
      "[CV 3/5; 6/8] START learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 3/5; 6/8] END learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.751 total time= 1.4min\n",
      "[CV 4/5; 6/8] START learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 4/5; 6/8] END learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.735 total time= 1.5min\n",
      "[CV 5/5; 6/8] START learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100\n",
      "[CV 5/5; 6/8] END learning_rate=0.2, max_depth=3, min_samples_split=4, n_estimators=100;, score=0.707 total time= 1.5min\n",
      "[CV 1/5; 7/8] START learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 1/5; 7/8] END learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.753 total time= 2.3min\n",
      "[CV 2/5; 7/8] START learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 2/5; 7/8] END learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.666 total time= 2.3min\n",
      "[CV 3/5; 7/8] START learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 3/5; 7/8] END learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.744 total time= 2.3min\n",
      "[CV 4/5; 7/8] START learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 4/5; 7/8] END learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.723 total time= 2.3min\n",
      "[CV 5/5; 7/8] START learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100\n",
      "[CV 5/5; 7/8] END learning_rate=0.2, max_depth=5, min_samples_split=2, n_estimators=100;, score=0.734 total time= 2.3min\n",
      "[CV 1/5; 8/8] START learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 1/5; 8/8] END learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.768 total time= 2.3min\n",
      "[CV 2/5; 8/8] START learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 2/5; 8/8] END learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.686 total time= 2.3min\n",
      "[CV 3/5; 8/8] START learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 3/5; 8/8] END learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.733 total time= 2.3min\n",
      "[CV 4/5; 8/8] START learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 4/5; 8/8] END learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.711 total time= 2.3min\n",
      "[CV 5/5; 8/8] START learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100\n",
      "[CV 5/5; 8/8] END learning_rate=0.2, max_depth=5, min_samples_split=4, n_estimators=100;, score=0.740 total time= 2.3min\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(estimator=GradientBoostingClassifier(random_state=42),\n",
       "             param_grid={'learning_rate': [0.1, 0.2], 'max_depth': [3, 5],\n",
       "                         'min_samples_split': [2, 4], 'n_estimators': [100]},\n",
       "             scoring='roc_auc', verbose=10)"
      ]
     },
     "execution_count": 317,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit the grid search object to the training data and find the optimal parameters\n",
    "gbc_grid = GridSearchCV(gbc, parameters, scoring = 'roc_auc', verbose=10)\n",
    "\n",
    "gbc_grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "id": "f69ac0e6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.7531651682793756\n",
      "Best Params:  {'learning_rate': 0.1, 'max_depth': 3, 'min_samples_split': 4, 'n_estimators': 100}\n",
      "ROC score: 0.9103\n",
      "Final ROC score: 0.9074\n"
     ]
    }
   ],
   "source": [
    "# Get the estimator and predict\n",
    "print(\"Best Score: \", gbc_grid.best_score_)\n",
    "print(\"Best Params: \", gbc_grid.best_params_)\n",
    "best_clf = gbc_grid.best_estimator_\n",
    "best_predictions = best_clf.predict_proba(X_train)[:, 1]\n",
    "\n",
    "# Make predictions using unoptimized and the best model\n",
    "predictions = (gbc.fit(X_train, y_train)).predict_proba(X_train)[:, 1]\n",
    "\n",
    "print(\"ROC score: {:.4f}\".format(roc_auc_score(y_train, predictions)))\n",
    "print(\"Final ROC score: {:.4f}\".format(roc_auc_score(y_train, best_predictions)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e267e2d4",
   "metadata": {},
   "source": [
    "#### AdaBoostClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "id": "660b7027",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['algorithm', 'base_estimator', 'learning_rate', 'n_estimators', 'random_state'])"
      ]
     },
     "execution_count": 324,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = AdaBoostClassifier()\n",
    "model.get_params().keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "id": "63600137",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Grid serach for AdaBoostClassifier model\n",
    "abc = AdaBoostClassifier(random_state=42)\n",
    "\n",
    "parameters = {'n_estimators':[50],\n",
    "              'random_state': [42], \n",
    "             }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "id": "0ddf9a8d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 1 candidates, totalling 5 fits\n",
      "[CV 1/5] END ..n_estimators=50, random_state=42;, score=0.761 total time=  20.3s\n",
      "[CV 2/5] END ..n_estimators=50, random_state=42;, score=0.672 total time=  20.0s\n",
      "[CV 3/5] END ..n_estimators=50, random_state=42;, score=0.688 total time=  20.1s\n",
      "[CV 4/5] END ..n_estimators=50, random_state=42;, score=0.709 total time=  20.0s\n",
      "[CV 5/5] END ..n_estimators=50, random_state=42;, score=0.715 total time=  20.1s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(estimator=AdaBoostClassifier(random_state=42),\n",
       "             param_grid={'n_estimators': [50], 'random_state': [42]},\n",
       "             scoring='roc_auc', verbose=5)"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fit the grid search object to the training data and find the optimal parameters\n",
    "abc_grid = GridSearchCV(abc, parameters, scoring = 'roc_auc', verbose=5)\n",
    "\n",
    "abc_grid.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "id": "fcda8530",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Score:  0.7088097731981077\n",
      "Best Params:  {'n_estimators': 50, 'random_state': 42}\n",
      "ROC score: 0.9103\n",
      "Final ROC score: 0.8408\n"
     ]
    }
   ],
   "source": [
    "# Get the estimator and predict\n",
    "print(\"Best Score: \", abc_grid.best_score_)\n",
    "print(\"Best Params: \", abc_grid.best_params_)\n",
    "best_clf = abc_grid.best_estimator_\n",
    "best_predictions = best_clf.predict_proba(X_train)[:, 1]\n",
    "\n",
    "# Make predictions using unoptimized and the best model\n",
    "predictions = (abc.fit(X_train, y_train)).predict_proba(X_train)[:, 1]\n",
    "\n",
    "print(\"ROC score: {:.4f}\".format(roc_auc_score(y_train, predictions)))\n",
    "print(\"Final ROC score: {:.4f}\".format(roc_auc_score(y_train, best_predictions)))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a329fd33",
   "metadata": {},
   "source": [
    "#### After fine tuning the classifiers, using ROC as a scoring metric, we've observed that the performance of the GradientBoost have been improved. Therefore GradientBoost will be used for the final test data set for the Kaggle competition."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2011bd46",
   "metadata": {},
   "source": [
    "## Part 3: Kaggle Competition\n",
    "\n",
    "Now that you'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. If you click on the link [here](http://www.kaggle.com/t/21e6d45d4c574c7fa2d868f0e8c83140), you'll be taken to the competition page where, if you have a Kaggle account, you can enter.\n",
    "\n",
    "Your entry to the competition should be a CSV file with two columns. The first column should be a copy of \"LNR\", which acts as an ID number for each individual in the \"TEST\" partition. The second column, \"RESPONSE\", should be some measure of how likely each individual became a customer – this might not be a straightforward probability. As you should have found in Part 2, there is a large output class imbalance, where most individuals did not respond to the mailout. Thus, predicting individual classes and using accuracy does not seem to be an appropriate performance evaluation method. Instead, the competition will be using AUC to evaluate performance. The exact values of the \"RESPONSE\" column do not matter as much: only that the higher values try to capture as many of the actual customers as possible, early in the ROC curve sweep."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "id": "fbed73ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3444: DtypeWarning: Columns (18,19) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  exec(code_obj, self.user_global_ns, self.user_ns)\n"
     ]
    }
   ],
   "source": [
    "# Load the data\n",
    "mailout_test = pd.read_csv('E:\\\\udct\\\\Arvato Project\\\\Udacity_MAILOUT_052018_TEST.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "id": "36e73b12",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42833, 366)"
      ]
     },
     "execution_count": 352,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the shape of the data\n",
    "mailout_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 353,
   "id": "66e53d17",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Format mixed-types \n",
      "\n",
      "Replace unknowns \n",
      "\n",
      "LNR\n",
      "AGER_TYP\n",
      "AKT_DAT_KL\n",
      "ALTER_HH\n",
      "ALTER_KIND1\n",
      "ALTER_KIND2\n",
      "ALTER_KIND3\n",
      "ALTER_KIND4\n",
      "ALTERSKATEGORIE_FEIN\n",
      "ANZ_HAUSHALTE_AKTIV\n",
      "ANZ_HH_TITEL\n",
      "ANZ_KINDER\n",
      "ANZ_PERSONEN\n",
      "ANZ_STATISTISCHE_HAUSHALTE\n",
      "ANZ_TITEL\n",
      "ARBEIT\n",
      "BALLRAUM\n",
      "CAMEO_DEUG_2015\n",
      "CJT_GESAMTTYP\n",
      "CJT_KATALOGNUTZER\n",
      "CJT_TYP_1\n",
      "CJT_TYP_2\n",
      "CJT_TYP_3\n",
      "CJT_TYP_4\n",
      "CJT_TYP_5\n",
      "CJT_TYP_6\n",
      "D19_BANKEN_ANZ_12\n",
      "D19_BANKEN_ANZ_24\n",
      "D19_BANKEN_DATUM\n",
      "D19_BANKEN_DIREKT\n",
      "D19_BANKEN_GROSS\n",
      "D19_BANKEN_LOKAL\n",
      "D19_BANKEN_OFFLINE_DATUM\n",
      "D19_BANKEN_ONLINE_DATUM\n",
      "D19_BANKEN_ONLINE_QUOTE_12\n",
      "D19_BANKEN_REST\n",
      "D19_BEKLEIDUNG_GEH\n",
      "D19_BEKLEIDUNG_REST\n",
      "D19_BILDUNG\n",
      "D19_BIO_OEKO\n",
      "D19_BUCH_CD\n",
      "D19_DIGIT_SERV\n",
      "D19_DROGERIEARTIKEL\n",
      "D19_ENERGIE\n",
      "D19_FREIZEIT\n",
      "D19_GARTEN\n",
      "D19_GESAMT_ANZ_12\n",
      "D19_GESAMT_ANZ_24\n",
      "D19_GESAMT_DATUM\n",
      "D19_GESAMT_OFFLINE_DATUM\n",
      "D19_GESAMT_ONLINE_DATUM\n",
      "D19_GESAMT_ONLINE_QUOTE_12\n",
      "D19_HANDWERK\n",
      "D19_HAUS_DEKO\n",
      "D19_KINDERARTIKEL\n",
      "D19_KONSUMTYP\n",
      "D19_KONSUMTYP_MAX\n",
      "D19_KOSMETIK\n",
      "D19_LEBENSMITTEL\n",
      "D19_LOTTO\n",
      "D19_NAHRUNGSERGAENZUNG\n",
      "D19_RATGEBER\n",
      "D19_REISEN\n",
      "D19_SAMMELARTIKEL\n",
      "D19_SCHUHE\n",
      "D19_SONSTIGE\n",
      "D19_SOZIALES\n",
      "D19_TECHNIK\n",
      "D19_TELKO_ANZ_12\n",
      "D19_TELKO_ANZ_24\n",
      "D19_TELKO_DATUM\n",
      "D19_TELKO_MOBILE\n",
      "D19_TELKO_OFFLINE_DATUM\n",
      "D19_TELKO_ONLINE_DATUM\n",
      "D19_TELKO_ONLINE_QUOTE_12\n",
      "D19_TELKO_REST\n",
      "D19_TIERARTIKEL\n",
      "D19_VERSAND_ANZ_12\n",
      "D19_VERSAND_ANZ_24\n",
      "D19_VERSAND_DATUM\n",
      "D19_VERSAND_OFFLINE_DATUM\n",
      "D19_VERSAND_ONLINE_DATUM\n",
      "D19_VERSAND_ONLINE_QUOTE_12\n",
      "D19_VERSAND_REST\n",
      "D19_VERSI_ANZ_12\n",
      "D19_VERSI_ANZ_24\n",
      "D19_VERSI_DATUM\n",
      "D19_VERSI_OFFLINE_DATUM\n",
      "D19_VERSI_ONLINE_DATUM\n",
      "D19_VERSI_ONLINE_QUOTE_12\n",
      "D19_VERSICHERUNGEN\n",
      "D19_VOLLSORTIMENT\n",
      "D19_WEIN_FEINKOST\n",
      "DSL_FLAG\n",
      "EINGEZOGENAM_HH_JAHR\n",
      "EWDICHTE\n",
      "EXTSEL992\n",
      "FINANZ_ANLEGER\n",
      "FINANZ_HAUSBAUER\n",
      "FINANZ_MINIMALIST\n",
      "FINANZ_SPARER\n",
      "FINANZ_UNAUFFAELLIGER\n",
      "FINANZ_VORSORGER\n",
      "FINANZTYP\n",
      "FIRMENDICHTE\n",
      "GEBAEUDETYP\n",
      "GEBAEUDETYP_RASTER\n",
      "GEBURTSJAHR\n",
      "GEMEINDETYP\n",
      "GFK_URLAUBERTYP\n",
      "GREEN_AVANTGARDE\n",
      "HEALTH_TYP\n",
      "HH_DELTA_FLAG\n",
      "HH_EINKOMMEN_SCORE\n",
      "INNENSTADT\n",
      "KBA05_ALTER1\n",
      "KBA05_ALTER2\n",
      "KBA05_ALTER3\n",
      "KBA05_ALTER4\n",
      "KBA05_ANHANG\n",
      "KBA05_ANTG1\n",
      "KBA05_ANTG2\n",
      "KBA05_ANTG3\n",
      "KBA05_ANTG4\n",
      "KBA05_AUTOQUOT\n",
      "KBA05_BAUMAX\n",
      "KBA05_CCM1\n",
      "KBA05_CCM2\n",
      "KBA05_CCM3\n",
      "KBA05_CCM4\n",
      "KBA05_DIESEL\n",
      "KBA05_FRAU\n",
      "KBA05_GBZ\n",
      "KBA05_HERST1\n",
      "KBA05_HERST2\n",
      "KBA05_HERST3\n",
      "KBA05_HERST4\n",
      "KBA05_HERST5\n",
      "KBA05_HERSTTEMP\n",
      "KBA05_KRSAQUOT\n",
      "KBA05_KRSHERST1\n",
      "KBA05_KRSHERST2\n",
      "KBA05_KRSHERST3\n",
      "KBA05_KRSKLEIN\n",
      "KBA05_KRSOBER\n",
      "KBA05_KRSVAN\n",
      "KBA05_KRSZUL\n",
      "KBA05_KW1\n",
      "KBA05_KW2\n",
      "KBA05_KW3\n",
      "KBA05_MAXAH\n",
      "KBA05_MAXBJ\n",
      "KBA05_MAXHERST\n",
      "KBA05_MAXSEG\n",
      "KBA05_MAXVORB\n",
      "KBA05_MOD1\n",
      "KBA05_MOD2\n",
      "KBA05_MOD3\n",
      "KBA05_MOD4\n",
      "KBA05_MOD8\n",
      "KBA05_MODTEMP\n",
      "KBA05_MOTOR\n",
      "KBA05_MOTRAD\n",
      "KBA05_SEG1\n",
      "KBA05_SEG10\n",
      "KBA05_SEG2\n",
      "KBA05_SEG3\n",
      "KBA05_SEG4\n",
      "KBA05_SEG5\n",
      "KBA05_SEG6\n",
      "KBA05_SEG7\n",
      "KBA05_SEG8\n",
      "KBA05_SEG9\n",
      "KBA05_VORB0\n",
      "KBA05_VORB1\n",
      "KBA05_VORB2\n",
      "KBA05_ZUL1\n",
      "KBA05_ZUL2\n",
      "KBA05_ZUL3\n",
      "KBA05_ZUL4\n",
      "KBA13_ALTERHALTER_30\n",
      "KBA13_ALTERHALTER_45\n",
      "KBA13_ALTERHALTER_60\n",
      "KBA13_ALTERHALTER_61\n",
      "KBA13_ANTG1\n",
      "KBA13_ANTG2\n",
      "KBA13_ANTG3\n",
      "KBA13_ANTG4\n",
      "KBA13_ANZAHL_PKW\n",
      "KBA13_AUDI\n",
      "KBA13_AUTOQUOTE\n",
      "KBA13_BAUMAX\n",
      "KBA13_BJ_1999\n",
      "KBA13_BJ_2000\n",
      "KBA13_BJ_2004\n",
      "KBA13_BJ_2006\n",
      "KBA13_BJ_2008\n",
      "KBA13_BJ_2009\n",
      "KBA13_BMW\n",
      "KBA13_CCM_0_1400\n",
      "KBA13_CCM_1000\n",
      "KBA13_CCM_1200\n",
      "KBA13_CCM_1400\n",
      "KBA13_CCM_1401_2500\n",
      "KBA13_CCM_1500\n",
      "KBA13_CCM_1600\n",
      "KBA13_CCM_1800\n",
      "KBA13_CCM_2000\n",
      "KBA13_CCM_2500\n",
      "KBA13_CCM_2501\n",
      "KBA13_CCM_3000\n",
      "KBA13_CCM_3001\n",
      "KBA13_FAB_ASIEN\n",
      "KBA13_FAB_SONSTIGE\n",
      "KBA13_FIAT\n",
      "KBA13_FORD\n",
      "KBA13_GBZ\n",
      "KBA13_HALTER_20\n",
      "KBA13_HALTER_25\n",
      "KBA13_HALTER_30\n",
      "KBA13_HALTER_35\n",
      "KBA13_HALTER_40\n",
      "KBA13_HALTER_45\n",
      "KBA13_HALTER_50\n",
      "KBA13_HALTER_55\n",
      "KBA13_HALTER_60\n",
      "KBA13_HALTER_65\n",
      "KBA13_HALTER_66\n",
      "KBA13_HERST_ASIEN\n",
      "KBA13_HERST_AUDI_VW\n",
      "KBA13_HERST_BMW_BENZ\n",
      "KBA13_HERST_EUROPA\n",
      "KBA13_HERST_FORD_OPEL\n",
      "KBA13_HERST_SONST\n",
      "KBA13_HHZ\n",
      "KBA13_KMH_0_140\n",
      "KBA13_KMH_110\n",
      "KBA13_KMH_140\n",
      "KBA13_KMH_140_210\n",
      "KBA13_KMH_180\n",
      "KBA13_KMH_210\n",
      "KBA13_KMH_211\n",
      "KBA13_KMH_250\n",
      "KBA13_KMH_251\n",
      "KBA13_KRSAQUOT\n",
      "KBA13_KRSHERST_AUDI_VW\n",
      "KBA13_KRSHERST_BMW_BENZ\n",
      "KBA13_KRSHERST_FORD_OPEL\n",
      "KBA13_KRSSEG_KLEIN\n",
      "KBA13_KRSSEG_OBER\n",
      "KBA13_KRSSEG_VAN\n",
      "KBA13_KRSZUL_NEU\n",
      "KBA13_KW_0_60\n",
      "KBA13_KW_110\n",
      "KBA13_KW_120\n",
      "KBA13_KW_121\n",
      "KBA13_KW_30\n",
      "KBA13_KW_40\n",
      "KBA13_KW_50\n",
      "KBA13_KW_60\n",
      "KBA13_KW_61_120\n",
      "KBA13_KW_70\n",
      "KBA13_KW_80\n",
      "KBA13_KW_90\n",
      "KBA13_MAZDA\n",
      "KBA13_MERCEDES\n",
      "KBA13_MOTOR\n",
      "KBA13_NISSAN\n",
      "KBA13_OPEL\n",
      "KBA13_PEUGEOT\n",
      "KBA13_RENAULT\n",
      "KBA13_SEG_GELAENDEWAGEN\n",
      "KBA13_SEG_GROSSRAUMVANS\n",
      "KBA13_SEG_KLEINST\n",
      "KBA13_SEG_KLEINWAGEN\n",
      "KBA13_SEG_KOMPAKTKLASSE\n",
      "KBA13_SEG_MINIVANS\n",
      "KBA13_SEG_MINIWAGEN\n",
      "KBA13_SEG_MITTELKLASSE\n",
      "KBA13_SEG_OBEREMITTELKLASSE\n",
      "KBA13_SEG_OBERKLASSE\n",
      "KBA13_SEG_SONSTIGE\n",
      "KBA13_SEG_SPORTWAGEN\n",
      "KBA13_SEG_UTILITIES\n",
      "KBA13_SEG_VAN\n",
      "KBA13_SEG_WOHNMOBILE\n",
      "KBA13_SITZE_4\n",
      "KBA13_SITZE_5\n",
      "KBA13_SITZE_6\n",
      "KBA13_TOYOTA\n",
      "KBA13_VORB_0\n",
      "KBA13_VORB_1\n",
      "KBA13_VORB_1_2\n",
      "KBA13_VORB_2\n",
      "KBA13_VORB_3\n",
      "KBA13_VW\n",
      "KK_KUNDENTYP\n",
      "KKK\n",
      "KOMBIALTER\n",
      "KONSUMNAEHE\n",
      "KONSUMZELLE\n",
      "LP_FAMILIE_FEIN\n",
      "LP_LEBENSPHASE_GROB\n",
      "LP_STATUS_FEIN\n",
      "MIN_GEBAEUDEJAHR\n",
      "MOBI_RASTER\n",
      "MOBI_REGIO\n",
      "NATIONALITAET_KZ\n",
      "ONLINE_AFFINITAET\n",
      "ORTSGR_KLS9\n",
      "PLZ8_ANTG1\n",
      "PLZ8_ANTG2\n",
      "PLZ8_ANTG3\n",
      "PLZ8_ANTG4\n",
      "PLZ8_BAUMAX\n",
      "PLZ8_GBZ\n",
      "PLZ8_HHZ\n",
      "REGIOTYP\n",
      "RELAT_AB\n",
      "RETOURTYP_BK_S\n",
      "RT_KEIN_ANREIZ\n",
      "RT_SCHNAEPPCHEN\n",
      "RT_UEBERGROESSE\n",
      "SEMIO_DOM\n",
      "SEMIO_ERL\n",
      "SEMIO_FAM\n",
      "SEMIO_KAEM\n",
      "SEMIO_KRIT\n",
      "SEMIO_KULT\n",
      "SEMIO_LUST\n",
      "SEMIO_MAT\n",
      "SEMIO_PFLICHT\n",
      "SEMIO_RAT\n",
      "SEMIO_REL\n",
      "SEMIO_SOZ\n",
      "SEMIO_TRADV\n",
      "SEMIO_VERT\n",
      "SHOPPER_TYP\n",
      "SOHO_KZ\n",
      "STRUKTURTYP\n",
      "TITEL_KZ\n",
      "UMFELD_ALT\n",
      "UMFELD_JUNG\n",
      "UNGLEICHENN_FLAG\n",
      "VERDICHTUNGSRAUM\n",
      "VERS_TYP\n",
      "VHA\n",
      "VHN\n",
      "VK_DHT4A\n",
      "VK_DISTANZ\n",
      "VK_ZG11\n",
      "W_KEIT_KIND_HH\n",
      "WOHNDAUER_2008\n",
      "WOHNLAGE\n",
      "ZABEOTYP\n",
      "ANREDE_KZ\n",
      "ALTERSKATEGORIE_GROB\n"
     ]
    }
   ],
   "source": [
    "# Perform data cleaning\n",
    "mailout_test_clean = data_cleaning_supervised_model(mailout_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "id": "233abca8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop the columns which are not required\n",
    "drop_cols = ['ALTER_KIND4','ALTER_KIND3','TITEL_KZ','ALTER_KIND2','ALTER_KIND1',\n",
    "                 'KK_KUNDENTYP','KBA05_BAUMAX','AGER_TYP','EXTSEL992','ALTER_HH']\n",
    "\n",
    "mailout_test_clean = mailout_test_clean.drop(drop_cols, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "id": "05996aa8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42833, 359)"
      ]
     },
     "execution_count": 361,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check the shape of the data\n",
    "mailout_test_clean.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 364,
   "id": "71a3d242",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale the data\n",
    "mailout_test_clean = scaler_model.fit_transform(mailout_test_clean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 365,
   "id": "f017396e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.97767981, 0.02232019],\n",
       "       [0.98700496, 0.01299504],\n",
       "       [0.99810962, 0.00189038],\n",
       "       ...,\n",
       "       [0.98372151, 0.01627849],\n",
       "       [0.97778084, 0.02221916],\n",
       "       [0.97482435, 0.02517565]])"
      ]
     },
     "execution_count": 365,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predict the model\n",
    "predict_proba = gbc_best.predict_proba(mailout_test_clean)\n",
    "predict_proba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 368,
   "id": "58c5a947",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>LNR</th>\n",
       "      <th>RESPONSE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1754</td>\n",
       "      <td>0.022320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1770</td>\n",
       "      <td>0.012995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1465</td>\n",
       "      <td>0.001890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1470</td>\n",
       "      <td>0.001841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1478</td>\n",
       "      <td>0.001200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1782</td>\n",
       "      <td>0.002900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1485</td>\n",
       "      <td>0.004282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1519</td>\n",
       "      <td>0.016915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1835</td>\n",
       "      <td>0.009339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1522</td>\n",
       "      <td>0.001887</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    LNR  RESPONSE\n",
       "0  1754  0.022320\n",
       "1  1770  0.012995\n",
       "2  1465  0.001890\n",
       "3  1470  0.001841\n",
       "4  1478  0.001200\n",
       "5  1782  0.002900\n",
       "6  1485  0.004282\n",
       "7  1519  0.016915\n",
       "8  1835  0.009339\n",
       "9  1522  0.001887"
      ]
     },
     "execution_count": 368,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate data for the competition\n",
    "result = pd.DataFrame({'LNR':mailout_test.LNR, 'RESPONSE':predict_proba[:,1]})\n",
    "result.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 370,
   "id": "9ab22a01",
   "metadata": {},
   "outputs": [],
   "source": [
    "result.to_csv('E:\\\\udct\\\\Arvato Project\\\\Arvato_Capstone_Submission.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}