Score_logistic_2014_version02.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# The goal of project is to predict sample's 2014 credit according 2013 information, and then compare with default credit. Finally calculate the scorecard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from sklearn import preprocessing\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df= pd.read_csv(\"/Users/yaqinhei/Downloads/CreditCardDefaultForecast.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>Account_Number</th>\n",
       "      <th>Default</th>\n",
       "      <th>Limit</th>\n",
       "      <th>Annual_Income</th>\n",
       "      <th>Month_On_Book</th>\n",
       "      <th>Gender_Code</th>\n",
       "      <th>Age_By_2013</th>\n",
       "      <th>Education_Code</th>\n",
       "      <th>Industry_code</th>\n",
       "      <th>OS_201312</th>\n",
       "      <th>...</th>\n",
       "      <th>Payment_201309</th>\n",
       "      <th>OS_201308</th>\n",
       "      <th>DUE_201308</th>\n",
       "      <th>Payment_201308</th>\n",
       "      <th>OS_201307</th>\n",
       "      <th>DUE_201307</th>\n",
       "      <th>Payment_201307</th>\n",
       "      <th>OS_201306</th>\n",
       "      <th>DUE_201306</th>\n",
       "      <th>Payment_201306</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>820000011</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>M</td>\n",
       "      <td>39</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4769.19</td>\n",
       "      <td>...</td>\n",
       "      <td>700.00</td>\n",
       "      <td>4889.00</td>\n",
       "      <td>1</td>\n",
       "      <td>200.00</td>\n",
       "      <td>4804.63</td>\n",
       "      <td>1</td>\n",
       "      <td>240.00</td>\n",
       "      <td>4782.88</td>\n",
       "      <td>1</td>\n",
       "      <td>405.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>820000046</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>59</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>343.51</td>\n",
       "      <td>...</td>\n",
       "      <td>300.00</td>\n",
       "      <td>106.94</td>\n",
       "      <td>0</td>\n",
       "      <td>1820.00</td>\n",
       "      <td>3427.26</td>\n",
       "      <td>0</td>\n",
       "      <td>1300.00</td>\n",
       "      <td>-7.56</td>\n",
       "      <td>0</td>\n",
       "      <td>1000.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820000054</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>36</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>820000062</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>M</td>\n",
       "      <td>38</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>...</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>820000089</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>57</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Account_Number  Default  Limit  Annual_Income  Month_On_Book Gender_Code  \\\n",
       "0       820000011        0   5000          30000            147           M   \n",
       "1       820000046        0   5000          30000            147           F   \n",
       "2       820000054        0   5000          30000            147           F   \n",
       "3       820000062        0   5000          30000            147           M   \n",
       "4       820000089        0   5000          30000            147           F   \n",
       "\n",
       "   Age_By_2013  Education_Code  Industry_code  OS_201312  ...  Payment_201309  \\\n",
       "0           39             NaN            4.0    4769.19  ...          700.00   \n",
       "1           59             NaN            4.0     343.51  ...          300.00   \n",
       "2           36             NaN            4.0      -0.01  ...          321.45   \n",
       "3           38             NaN            4.0      -4.72  ...          860.00   \n",
       "4           57             NaN            4.0       0.00  ...            4.05   \n",
       "\n",
       "   OS_201308  DUE_201308  Payment_201308  OS_201307  DUE_201307  \\\n",
       "0    4889.00           1          200.00    4804.63           1   \n",
       "1     106.94           0         1820.00    3427.26           0   \n",
       "2      -0.01           0          321.45      -0.01           0   \n",
       "3      -4.72           0          860.00      -4.72           0   \n",
       "4       0.00           0            4.05       0.00           0   \n",
       "\n",
       "   Payment_201307  OS_201306  DUE_201306  Payment_201306  \n",
       "0          240.00    4782.88           1          405.00  \n",
       "1         1300.00      -7.56           0         1000.00  \n",
       "2          321.45      -0.01           0          321.45  \n",
       "3          860.00      -4.72           0          860.00  \n",
       "4            4.05       0.00           0            4.05  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1000 entries, 0 to 999\n",
      "Data columns (total 30 columns):\n",
      " #   Column          Non-Null Count  Dtype  \n",
      "---  ------          --------------  -----  \n",
      " 0   Account_Number  1000 non-null   int64  \n",
      " 1   Default         1000 non-null   int64  \n",
      " 2   Limit           1000 non-null   int64  \n",
      " 3   Annual_Income   1000 non-null   int64  \n",
      " 4   Month_On_Book   1000 non-null   int64  \n",
      " 5   Gender_Code     1000 non-null   object \n",
      " 6   Age_By_2013     1000 non-null   int64  \n",
      " 7   Education_Code  166 non-null    float64\n",
      " 8   Industry_code   989 non-null    float64\n",
      " 9   OS_201312       1000 non-null   float64\n",
      " 10  DUE_201312      1000 non-null   int64  \n",
      " 11  Payment_201312  1000 non-null   float64\n",
      " 12  OS_201311       1000 non-null   float64\n",
      " 13  DUE_201311      1000 non-null   int64  \n",
      " 14  Payment_201311  1000 non-null   float64\n",
      " 15  OS_201310       1000 non-null   float64\n",
      " 16  DUE_201310      1000 non-null   int64  \n",
      " 17  Payment_201310  1000 non-null   float64\n",
      " 18  OS_201309       1000 non-null   float64\n",
      " 19  DUE_201309      1000 non-null   int64  \n",
      " 20  Payment_201309  1000 non-null   float64\n",
      " 21  OS_201308       1000 non-null   float64\n",
      " 22  DUE_201308      1000 non-null   int64  \n",
      " 23  Payment_201308  1000 non-null   float64\n",
      " 24  OS_201307       1000 non-null   float64\n",
      " 25  DUE_201307      1000 non-null   int64  \n",
      " 26  Payment_201307  1000 non-null   float64\n",
      " 27  OS_201306       1000 non-null   float64\n",
      " 28  DUE_201306      1000 non-null   int64  \n",
      " 29  Payment_201306  1000 non-null   float64\n",
      "dtypes: float64(16), int64(13), object(1)\n",
      "memory usage: 234.5+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Account_Number</th>\n",
       "      <th>Default</th>\n",
       "      <th>Limit</th>\n",
       "      <th>Annual_Income</th>\n",
       "      <th>Month_On_Book</th>\n",
       "      <th>Age_By_2013</th>\n",
       "      <th>Education_Code</th>\n",
       "      <th>Industry_code</th>\n",
       "      <th>OS_201312</th>\n",
       "      <th>DUE_201312</th>\n",
       "      <th>...</th>\n",
       "      <th>Payment_201309</th>\n",
       "      <th>OS_201308</th>\n",
       "      <th>DUE_201308</th>\n",
       "      <th>Payment_201308</th>\n",
       "      <th>OS_201307</th>\n",
       "      <th>DUE_201307</th>\n",
       "      <th>Payment_201307</th>\n",
       "      <th>OS_201306</th>\n",
       "      <th>DUE_201306</th>\n",
       "      <th>Payment_201306</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.000000e+03</td>\n",
       "      <td>1000.00000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1.000000e+03</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>166.000000</td>\n",
       "      <td>989.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>8.200296e+08</td>\n",
       "      <td>0.10000</td>\n",
       "      <td>11227.000000</td>\n",
       "      <td>7.002719e+04</td>\n",
       "      <td>145.947000</td>\n",
       "      <td>46.626000</td>\n",
       "      <td>8.228916</td>\n",
       "      <td>17.237614</td>\n",
       "      <td>4061.497690</td>\n",
       "      <td>0.447000</td>\n",
       "      <td>...</td>\n",
       "      <td>1581.047770</td>\n",
       "      <td>4020.573920</td>\n",
       "      <td>0.379000</td>\n",
       "      <td>1740.649040</td>\n",
       "      <td>3884.317200</td>\n",
       "      <td>0.380000</td>\n",
       "      <td>1675.287450</td>\n",
       "      <td>4030.244420</td>\n",
       "      <td>0.416000</td>\n",
       "      <td>1619.730790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.088140e+04</td>\n",
       "      <td>0.30015</td>\n",
       "      <td>10482.621636</td>\n",
       "      <td>6.403540e+04</td>\n",
       "      <td>0.755488</td>\n",
       "      <td>6.462808</td>\n",
       "      <td>4.643210</td>\n",
       "      <td>8.726800</td>\n",
       "      <td>7839.986061</td>\n",
       "      <td>0.709715</td>\n",
       "      <td>...</td>\n",
       "      <td>4395.876909</td>\n",
       "      <td>7396.924587</td>\n",
       "      <td>0.579392</td>\n",
       "      <td>5408.747207</td>\n",
       "      <td>7006.564627</td>\n",
       "      <td>0.603294</td>\n",
       "      <td>4744.071334</td>\n",
       "      <td>7245.036458</td>\n",
       "      <td>0.585907</td>\n",
       "      <td>4775.807626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>8.200000e+08</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2.520000e+04</td>\n",
       "      <td>142.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-24000.620000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-10015.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-9280.380000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-8750.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>8.200068e+08</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>6000.000000</td>\n",
       "      <td>3.813600e+04</td>\n",
       "      <td>146.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>95.947500</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>75.772500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>8.200137e+08</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>8000.000000</td>\n",
       "      <td>5.219400e+04</td>\n",
       "      <td>146.000000</td>\n",
       "      <td>46.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>204.425000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>280.000000</td>\n",
       "      <td>220.330000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>282.500000</td>\n",
       "      <td>222.865000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>264.400000</td>\n",
       "      <td>151.340000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>250.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.200213e+08</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>12000.000000</td>\n",
       "      <td>7.800000e+04</td>\n",
       "      <td>146.000000</td>\n",
       "      <td>51.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>5873.190000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>5785.655000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>5757.282500</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>5848.725000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.203763e+08</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>122500.000000</td>\n",
       "      <td>1.061856e+06</td>\n",
       "      <td>147.000000</td>\n",
       "      <td>72.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>104349.890000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>52179.120000</td>\n",
       "      <td>83724.190000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>100000.000000</td>\n",
       "      <td>85890.650000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>52179.120000</td>\n",
       "      <td>82580.910000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>52179.120000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 29 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Account_Number     Default          Limit  Annual_Income  \\\n",
       "count    1.000000e+03  1000.00000    1000.000000   1.000000e+03   \n",
       "mean     8.200296e+08     0.10000   11227.000000   7.002719e+04   \n",
       "std      6.088140e+04     0.30015   10482.621636   6.403540e+04   \n",
       "min      8.200000e+08     0.00000    2000.000000   2.520000e+04   \n",
       "25%      8.200068e+08     0.00000    6000.000000   3.813600e+04   \n",
       "50%      8.200137e+08     0.00000    8000.000000   5.219400e+04   \n",
       "75%      8.200213e+08     0.00000   12000.000000   7.800000e+04   \n",
       "max      8.203763e+08     1.00000  122500.000000   1.061856e+06   \n",
       "\n",
       "       Month_On_Book  Age_By_2013  Education_Code  Industry_code  \\\n",
       "count    1000.000000  1000.000000      166.000000     989.000000   \n",
       "mean      145.947000    46.626000        8.228916      17.237614   \n",
       "std         0.755488     6.462808        4.643210       8.726800   \n",
       "min       142.000000    35.000000        0.000000       1.000000   \n",
       "25%       146.000000    42.000000        8.000000      11.000000   \n",
       "50%       146.000000    46.000000       11.000000      19.000000   \n",
       "75%       146.000000    51.000000       11.000000      20.000000   \n",
       "max       147.000000    72.000000       11.000000      35.000000   \n",
       "\n",
       "           OS_201312   DUE_201312  ...  Payment_201309     OS_201308  \\\n",
       "count    1000.000000  1000.000000  ...     1000.000000   1000.000000   \n",
       "mean     4061.497690     0.447000  ...     1581.047770   4020.573920   \n",
       "std      7839.986061     0.709715  ...     4395.876909   7396.924587   \n",
       "min    -24000.620000     0.000000  ...        0.000000 -10015.000000   \n",
       "25%         0.000000     0.000000  ...      100.000000      0.000000   \n",
       "50%       204.425000     0.000000  ...      280.000000    220.330000   \n",
       "75%      5873.190000     1.000000  ...     1000.000000   5785.655000   \n",
       "max    104349.890000     4.000000  ...    52179.120000  83724.190000   \n",
       "\n",
       "        DUE_201308  Payment_201308     OS_201307   DUE_201307  Payment_201307  \\\n",
       "count  1000.000000     1000.000000   1000.000000  1000.000000     1000.000000   \n",
       "mean      0.379000     1740.649040   3884.317200     0.380000     1675.287450   \n",
       "std       0.579392     5408.747207   7006.564627     0.603294     4744.071334   \n",
       "min       0.000000        0.000000  -9280.380000     0.000000        0.000000   \n",
       "25%       0.000000       95.947500      0.000000     0.000000      100.000000   \n",
       "50%       0.000000      282.500000    222.865000     0.000000      264.400000   \n",
       "75%       1.000000     1000.000000   5757.282500     1.000000     1000.000000   \n",
       "max       4.000000   100000.000000  85890.650000     4.000000    52179.120000   \n",
       "\n",
       "          OS_201306   DUE_201306  Payment_201306  \n",
       "count   1000.000000  1000.000000     1000.000000  \n",
       "mean    4030.244420     0.416000     1619.730790  \n",
       "std     7245.036458     0.585907     4775.807626  \n",
       "min    -8750.000000     0.000000        0.000000  \n",
       "25%        0.000000     0.000000       75.772500  \n",
       "50%      151.340000     0.000000      250.000000  \n",
       "75%     5848.725000     1.000000     1000.000000  \n",
       "max    82580.910000     3.000000    52179.120000  \n",
       "\n",
       "[8 rows x 29 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# First step obeserve data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>DUE_201307</th>\n",
       "      <th>DUE_201308</th>\n",
       "      <th>Month_On_Book</th>\n",
       "      <th>OS_201306</th>\n",
       "      <th>OS_201309</th>\n",
       "      <th>OS_201310</th>\n",
       "      <th>OS_201311</th>\n",
       "      <th>OS_201312</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Default</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>268</td>\n",
       "      <td>280</td>\n",
       "      <td>131530</td>\n",
       "      <td>3237843.43</td>\n",
       "      <td>3298491.28</td>\n",
       "      <td>3271822.84</td>\n",
       "      <td>3303163.91</td>\n",
       "      <td>3251029.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>112</td>\n",
       "      <td>99</td>\n",
       "      <td>14417</td>\n",
       "      <td>792400.99</td>\n",
       "      <td>759684.85</td>\n",
       "      <td>789229.66</td>\n",
       "      <td>813339.78</td>\n",
       "      <td>810468.56</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         DUE_201307  DUE_201308  Month_On_Book   OS_201306   OS_201309  \\\n",
       "Default                                                                  \n",
       "0               268         280         131530  3237843.43  3298491.28   \n",
       "1               112          99          14417   792400.99   759684.85   \n",
       "\n",
       "          OS_201310   OS_201311   OS_201312  \n",
       "Default                                      \n",
       "0        3271822.84  3303163.91  3251029.13  \n",
       "1         789229.66   813339.78   810468.56  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "table = df.pivot_table(['Month_On_Book','OS_201312','OS_201311','OS_201310','OS_201309','DUE_201308','DUE_201307','OS_201306',],\n",
    "                                   index=['Default'],\n",
    "                                   aggfunc=np.sum)\n",
    "table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(x ='Default', data = df, palette = 'hls')\n",
    "plt.show()\n",
    "plt.savefig('count_plot')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Second, calculate and add measures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "detail = pd.DataFrame(df, columns = ['Account_Number','Default','Annual_Income', 'Month_On_Book', 'Age_By_2013', 'Education_Code', 'Industry_code'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1000 entries, 0 to 999\n",
      "Data columns (total 7 columns):\n",
      " #   Column          Non-Null Count  Dtype  \n",
      "---  ------          --------------  -----  \n",
      " 0   Account_Number  1000 non-null   int64  \n",
      " 1   Default         1000 non-null   int64  \n",
      " 2   Annual_Income   1000 non-null   int64  \n",
      " 3   Month_On_Book   1000 non-null   int64  \n",
      " 4   Age_By_2013     1000 non-null   int64  \n",
      " 5   Education_Code  166 non-null    float64\n",
      " 6   Industry_code   989 non-null    float64\n",
      "dtypes: float64(2), int64(5)\n",
      "memory usage: 54.8 KB\n"
     ]
    }
   ],
   "source": [
    "detail.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bin variables follow the the most frequency level and reduce level to 2 or 3."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1 information details\n",
    "\n",
    "for col_name in detail.columns:\n",
    "    if detail[col_name].dtypes == 'object':\n",
    "        unique_cat = len(detail[col_name].unique())\n",
    "        print(\"Feature '{col_name}' has {unique_cat} unique categories\".format(col_name = col_name, unique_cat = unique_cat))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19.0    272\n",
      "30.0    133\n",
      "17.0    111\n",
      "5.0     103\n",
      "11.0     75\n",
      "4.0      54\n",
      "34.0     30\n",
      "12.0     28\n",
      "20.0     23\n",
      "2.0      20\n",
      "Name: Industry_code, dtype: int64\n"
     ]
    }
   ],
   "source": [
    " print(detail['Industry_code'].value_counts().sort_values(ascending = False).head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.0    115\n",
      "0.0      38\n",
      "10.0      6\n",
      "8.0       4\n",
      "3.0       3\n",
      "Name: Education_Code, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(detail['Education_Code'].value_counts().sort_values(ascending = False).head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "others     885\n",
      "hanshou    115\n",
      "Name: Education_Code, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "detail['Education_Code'] = ['hanshou' if detail == 11 else 'others' for detail in detail['Education_Code']]\n",
    "\n",
    "print(detail['Education_Code'].value_counts().sort_values(ascending = False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "other     871\n",
      "public    129\n",
      "Name: Industry_code, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "detail['Industry_code'] = ['public' if detail == 4 or detail == 11 else 'other' for detail in detail['Industry_code']]\n",
    "\n",
    "print(detail['Industry_code'].value_counts().sort_values(ascending = False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "detail['Annual_Income_Bins'] = pd.qcut(detail['Annual_Income'], 3, \n",
    "                                   labels=['Income_5k', 'Income_10k','Income_15k'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "detail['Age_By_2013_Bins'] = pd.qcut(detail['Age_By_2013'], 3, \n",
    "                                   labels=['age_to_46', 'age_to_51','age_above51'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "detail['Month_On_Book_Bins'] = pd.qcut(detail['Month_On_Book'], 2, \n",
    "                                   labels=['book_to_145','book_after_145'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "da = pd.DataFrame(detail, columns = ['Account_Number','Age_By_2013_Bins','Annual_Income_Bins', 'Month_On_Book_Bins'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1000 entries, 0 to 999\n",
      "Data columns (total 4 columns):\n",
      " #   Column              Non-Null Count  Dtype   \n",
      "---  ------              --------------  -----   \n",
      " 0   Account_Number      1000 non-null   int64   \n",
      " 1   Age_By_2013_Bins    1000 non-null   category\n",
      " 2   Annual_Income_Bins  1000 non-null   category\n",
      " 3   Month_On_Book_Bins  1000 non-null   category\n",
      "dtypes: category(3), int64(1)\n",
      "memory usage: 11.2 KB\n"
     ]
    }
   ],
   "source": [
    "da.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Consider the times of overdue Last 6 months and the max overdue for last 3 months."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# due days analysis (1000 rows)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "due = pd.DataFrame(df, columns = ['Account_Number','DUE_201306','DUE_201307','DUE_201308','DUE_201309','DUE_201310','DUE_201311','DUE_201312'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "col = due.loc[: , \"DUE_201306\":\"DUE_201312\"]\n",
    "due['due_count'] = (col>0).sum(axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "col = due.loc[: , \"DUE_201310\":\"DUE_201312\"]\n",
    "due['max_recent'] = col.max(axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>Account_Number</th>\n",
       "      <th>due_count</th>\n",
       "      <th>max_recent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>820000011</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>820000046</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820000054</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>820000062</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>820000089</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>820361391</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>820363998</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>820365133</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>820372130</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>820376305</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Account_Number  due_count  max_recent\n",
       "0         820000011          7           1\n",
       "1         820000046          0           0\n",
       "2         820000054          0           0\n",
       "3         820000062          0           0\n",
       "4         820000089          0           0\n",
       "..              ...        ...         ...\n",
       "995       820361391          6           2\n",
       "996       820363998          5           1\n",
       "997       820365133          7           1\n",
       "998       820372130          6           2\n",
       "999       820376305          7           1\n",
       "\n",
       "[1000 rows x 3 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "due1 = pd.DataFrame(due, columns= ['Account_Number','due_count', 'max_recent'])\n",
    "due1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Limit Utilization = Outstanding /Limit rate "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur = pd.DataFrame(df, columns = ['Account_Number','Gender_Code','OS_201306','OS_201307','OS_201308','OS_201309','OS_201310','OS_201311','OS_201312'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur['Limit'] = df.Limit "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur = creur[creur['Limit']>0] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "col = creur.loc[: , \"OS_201306\":\"OS_201312\"]\n",
    "creur['ur_mean'] = col.mean(axis=1)\n",
    "creur['ur_mean'] = creur['ur_mean']/creur.Limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur['ur_max'] = col.max(axis=1)\n",
    "creur['ur_max'] = creur['ur_max']/creur.Limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur['ur_min'] = col.min(axis=1)\n",
    "creur['ur_min'] = creur['ur_min']/creur.Limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "creur['ur_mean'] = np.where(creur.ur_mean <= 0, 0, creur.ur_mean)\n",
    "creur['ur_min'] = np.where(creur.ur_min <= 0, 0, creur.ur_min)\n",
    "creur['ur_max'] = np.where(creur.ur_max <= 0, 0, creur.ur_max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     Account_Number Gender_Code  OS_201306  OS_201307  OS_201308  OS_201309  \\\n",
      "0         820000011           M    4782.88    4804.63    4889.00    4640.36   \n",
      "1         820000046           F      -7.56    3427.26     106.94     704.81   \n",
      "2         820000054           F      -0.01      -0.01      -0.01      -0.01   \n",
      "3         820000062           M      -4.72      -4.72      -4.72      -4.72   \n",
      "4         820000089           F       0.00       0.00       0.00       0.00   \n",
      "..              ...         ...        ...        ...        ...        ...   \n",
      "995       820361391           M    5981.50    6049.13    5980.09    5967.94   \n",
      "996       820363998           M    4793.98    5747.07    5786.45    5695.93   \n",
      "997       820365133           F   49810.96    7103.77    8680.90   10781.20   \n",
      "998       820372130           F    5158.27    5025.60    5127.04    5361.23   \n",
      "999       820376305           M    6039.79    6170.03    6102.76    6226.08   \n",
      "\n",
      "     OS_201310  OS_201311  OS_201312  Limit   ur_mean    ur_max    ur_min  \n",
      "0      4611.06    4866.75    4769.19   5000  0.953253  0.977800  0.922212  \n",
      "1       -98.95    4071.65     343.51   5000  0.244219  0.814330  0.000000  \n",
      "2        -0.01      -0.01      -0.01   5000  0.000000  0.000000  0.000000  \n",
      "3        -4.72      -4.72      -4.72   5000  0.000000  0.000000  0.000000  \n",
      "4         0.00       0.00       0.00   5000  0.000000  0.000000  0.000000  \n",
      "..         ...        ...        ...    ...       ...       ...       ...  \n",
      "995    6006.35    5977.72    6185.04   6000  1.003518  1.030840  0.994657  \n",
      "996    5980.07    5919.96    5926.88   6000  0.948818  0.996678  0.798997  \n",
      "997   12641.05   14481.51   16443.61  50000  0.342694  0.996219  0.142075  \n",
      "998    5184.26    5369.01    4865.33   5400  0.954781  0.994261  0.900987  \n",
      "999    6156.62    6230.38    6331.94   6300  0.980898  1.005070  0.958697  \n",
      "\n",
      "[1000 rows x 13 columns]\n"
     ]
    }
   ],
   "source": [
    "print(creur)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "cre = pd. DataFrame(creur, columns = ['Account_Number','Gender_Code','ur_mean', 'ur_max', 'ur_min'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>Account_Number</th>\n",
       "      <th>ur_mean</th>\n",
       "      <th>ur_max</th>\n",
       "      <th>ur_min</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1.000000e+03</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>8.200296e+08</td>\n",
       "      <td>0.383786</td>\n",
       "      <td>0.482466</td>\n",
       "      <td>0.296130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.088140e+04</td>\n",
       "      <td>0.416238</td>\n",
       "      <td>0.455930</td>\n",
       "      <td>0.401284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>8.200000e+08</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>8.200068e+08</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>8.200137e+08</td>\n",
       "      <td>0.134350</td>\n",
       "      <td>0.481572</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>8.200213e+08</td>\n",
       "      <td>0.868446</td>\n",
       "      <td>0.977880</td>\n",
       "      <td>0.766957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.203763e+08</td>\n",
       "      <td>1.037999</td>\n",
       "      <td>1.082977</td>\n",
       "      <td>1.018140</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Account_Number      ur_mean       ur_max       ur_min\n",
       "count    1.000000e+03  1000.000000  1000.000000  1000.000000\n",
       "mean     8.200296e+08     0.383786     0.482466     0.296130\n",
       "std      6.088140e+04     0.416238     0.455930     0.401284\n",
       "min      8.200000e+08     0.000000     0.000000     0.000000\n",
       "25%      8.200068e+08     0.000000     0.000000     0.000000\n",
       "50%      8.200137e+08     0.134350     0.481572     0.000000\n",
       "75%      8.200213e+08     0.868446     0.977880     0.766957\n",
       "max      8.203763e+08     1.037999     1.082977     1.018140"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cre.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create repayment rate varibles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "part = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>Account_Number</th>\n",
       "      <th>Default</th>\n",
       "      <th>Limit</th>\n",
       "      <th>Annual_Income</th>\n",
       "      <th>Month_On_Book</th>\n",
       "      <th>Gender_Code</th>\n",
       "      <th>Age_By_2013</th>\n",
       "      <th>Education_Code</th>\n",
       "      <th>Industry_code</th>\n",
       "      <th>OS_201312</th>\n",
       "      <th>...</th>\n",
       "      <th>Payment_201309</th>\n",
       "      <th>OS_201308</th>\n",
       "      <th>DUE_201308</th>\n",
       "      <th>Payment_201308</th>\n",
       "      <th>OS_201307</th>\n",
       "      <th>DUE_201307</th>\n",
       "      <th>Payment_201307</th>\n",
       "      <th>OS_201306</th>\n",
       "      <th>DUE_201306</th>\n",
       "      <th>Payment_201306</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>820000011</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>M</td>\n",
       "      <td>39</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4769.19</td>\n",
       "      <td>...</td>\n",
       "      <td>700.00</td>\n",
       "      <td>4889.00</td>\n",
       "      <td>1</td>\n",
       "      <td>200.00</td>\n",
       "      <td>4804.63</td>\n",
       "      <td>1</td>\n",
       "      <td>240.00</td>\n",
       "      <td>4782.88</td>\n",
       "      <td>1</td>\n",
       "      <td>405.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>820000046</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>59</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>343.51</td>\n",
       "      <td>...</td>\n",
       "      <td>300.00</td>\n",
       "      <td>106.94</td>\n",
       "      <td>0</td>\n",
       "      <td>1820.00</td>\n",
       "      <td>3427.26</td>\n",
       "      <td>0</td>\n",
       "      <td>1300.00</td>\n",
       "      <td>-7.56</td>\n",
       "      <td>0</td>\n",
       "      <td>1000.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820000054</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>36</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>321.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>820000062</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>M</td>\n",
       "      <td>38</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>...</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "      <td>-4.72</td>\n",
       "      <td>0</td>\n",
       "      <td>860.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>820000089</td>\n",
       "      <td>0</td>\n",
       "      <td>5000</td>\n",
       "      <td>30000</td>\n",
       "      <td>147</td>\n",
       "      <td>F</td>\n",
       "      <td>57</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>4.05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Account_Number  Default  Limit  Annual_Income  Month_On_Book Gender_Code  \\\n",
       "0       820000011        0   5000          30000            147           M   \n",
       "1       820000046        0   5000          30000            147           F   \n",
       "2       820000054        0   5000          30000            147           F   \n",
       "3       820000062        0   5000          30000            147           M   \n",
       "4       820000089        0   5000          30000            147           F   \n",
       "\n",
       "   Age_By_2013  Education_Code  Industry_code  OS_201312  ...  Payment_201309  \\\n",
       "0           39             NaN            4.0    4769.19  ...          700.00   \n",
       "1           59             NaN            4.0     343.51  ...          300.00   \n",
       "2           36             NaN            4.0      -0.01  ...          321.45   \n",
       "3           38             NaN            4.0      -4.72  ...          860.00   \n",
       "4           57             NaN            4.0       0.00  ...            4.05   \n",
       "\n",
       "   OS_201308  DUE_201308  Payment_201308  OS_201307  DUE_201307  \\\n",
       "0    4889.00           1          200.00    4804.63           1   \n",
       "1     106.94           0         1820.00    3427.26           0   \n",
       "2      -0.01           0          321.45      -0.01           0   \n",
       "3      -4.72           0          860.00      -4.72           0   \n",
       "4       0.00           0            4.05       0.00           0   \n",
       "\n",
       "   Payment_201307  OS_201306  DUE_201306  Payment_201306  \n",
       "0          240.00    4782.88           1          405.00  \n",
       "1         1300.00      -7.56           0         1000.00  \n",
       "2          321.45      -0.01           0          321.45  \n",
       "3          860.00      -4.72           0          860.00  \n",
       "4            4.05       0.00           0            4.05  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "part= pd.DataFrame()\n",
    "part['Account_Number'] = df.Account_Number"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "part['OS_201306'] = np.where(df.OS_201306 <= 0, df.Payment_201306, df.OS_201306 )\n",
    "part['OS_201307'] = np.where(df.OS_201307 <= 0, df.Payment_201307, df.OS_201307 )\n",
    "part['OS_201308'] = np.where(df.OS_201308 <= 0, df.Payment_201308, df.OS_201308 )\n",
    "part['OS_201309'] = np.where(df.OS_201309 <= 0, df.Payment_201309, df.OS_201309 )\n",
    "part['OS_201310'] = np.where(df.OS_201310 <= 0, df.Payment_201310, df.OS_201310 )\n",
    "part['OS_201311'] = np.where(df.OS_201311 <= 0, df.Payment_201311, df.OS_201311 )\n",
    "part['OS_201312'] = np.where(df.OS_201312 <= 0, df.Payment_201312, df.OS_201312 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "part['payradio06'] = df.Payment_201306 / part.OS_201306\n",
    "part['payradio07'] = df.Payment_201307 / part.OS_201307\n",
    "part['payradio08'] = df.Payment_201308 / part.OS_201308\n",
    "part['payradio09'] = df.Payment_201309 / part.OS_201309\n",
    "part['payradio10'] = df.Payment_201310 / part.OS_201310\n",
    "part['payradio11'] = df.Payment_201311 / part.OS_201312\n",
    "part['payradio12'] = df.Payment_201312 / part.OS_201312"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     Account_Number  OS_201306  OS_201307  OS_201308  OS_201309  OS_201310  \\\n",
      "0         820000011    4782.88    4804.63    4889.00    4640.36    4611.06   \n",
      "1         820000046    1000.00    3427.26     106.94     704.81     600.00   \n",
      "2         820000054     321.45     321.45     321.45     321.45     321.45   \n",
      "3         820000062     860.00     860.00     860.00     860.00     860.00   \n",
      "4         820000089       4.05       4.05       4.05       4.05       4.05   \n",
      "..              ...        ...        ...        ...        ...        ...   \n",
      "995       820361391    5981.50    6049.13    5980.09    5967.94    6006.35   \n",
      "996       820363998    4793.98    5747.07    5786.45    5695.93    5980.07   \n",
      "997       820365133   49810.96    7103.77    8680.90   10781.20   12641.05   \n",
      "998       820372130    5158.27    5025.60    5127.04    5361.23    5184.26   \n",
      "999       820376305    6039.79    6170.03    6102.76    6226.08    6156.62   \n",
      "\n",
      "     OS_201311  OS_201312  payradio06  payradio07  payradio08  payradio09  \\\n",
      "0      4866.75    4769.19    0.084677    0.049952    0.040908    0.150850   \n",
      "1      4071.65     343.51    1.000000    0.379312   17.018889    0.425647   \n",
      "2       321.45     321.45    1.000000    1.000000    1.000000    1.000000   \n",
      "3       860.00     860.00    1.000000    1.000000    1.000000    1.000000   \n",
      "4         4.05       4.05    1.000000    1.000000    1.000000    1.000000   \n",
      "..         ...        ...         ...         ...         ...         ...   \n",
      "995    5977.72    6185.04    0.066873    0.033063    0.070233    0.016756   \n",
      "996    5919.96    5926.88    0.010430    0.078301    0.086409    0.087782   \n",
      "997   14481.51   16443.61    0.010038    0.028154    0.057598    0.018551   \n",
      "998    5369.01    4865.33    0.029080    0.062613    0.061375    0.058694   \n",
      "999    6230.38    6331.94    0.026491    0.025932    0.026218    0.025698   \n",
      "\n",
      "     payradio10  payradio11  payradio12  \n",
      "0      0.093254    0.041936    0.062904  \n",
      "1      1.000000    0.582225    7.277809  \n",
      "2      1.000000    1.000000    1.000000  \n",
      "3      1.000000    1.000000    1.000000  \n",
      "4      1.000000    1.000000    1.000000  \n",
      "..          ...         ...         ...  \n",
      "995    0.033298    0.064672    0.064672  \n",
      "996    0.083611    0.025308    0.042181  \n",
      "997    0.021322    0.019460    0.015812  \n",
      "998    0.054010    0.057550    0.041107  \n",
      "999    0.025988    0.025269    0.025269  \n",
      "\n",
      "[1000 rows x 15 columns]\n"
     ]
    }
   ],
   "source": [
    "print(part)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>Account_Number</th>\n",
       "      <th>payradio_max</th>\n",
       "      <th>payradio_min</th>\n",
       "      <th>payradio_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>820000011</td>\n",
       "      <td>0.150850</td>\n",
       "      <td>0.040908</td>\n",
       "      <td>0.074926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>820000046</td>\n",
       "      <td>17.018889</td>\n",
       "      <td>0.379312</td>\n",
       "      <td>3.954840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820000054</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>820000062</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>820000089</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>820361391</td>\n",
       "      <td>0.070233</td>\n",
       "      <td>0.016756</td>\n",
       "      <td>0.049938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>820363998</td>\n",
       "      <td>0.087782</td>\n",
       "      <td>0.010430</td>\n",
       "      <td>0.059146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>820365133</td>\n",
       "      <td>0.057598</td>\n",
       "      <td>0.010038</td>\n",
       "      <td>0.024419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>820372130</td>\n",
       "      <td>0.062613</td>\n",
       "      <td>0.029080</td>\n",
       "      <td>0.052061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>820376305</td>\n",
       "      <td>0.026491</td>\n",
       "      <td>0.025269</td>\n",
       "      <td>0.025838</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Account_Number  payradio_max  payradio_min  payradio_mean\n",
       "0         820000011      0.150850      0.040908       0.074926\n",
       "1         820000046     17.018889      0.379312       3.954840\n",
       "2         820000054      1.000000      1.000000       1.000000\n",
       "3         820000062      1.000000      1.000000       1.000000\n",
       "4         820000089      1.000000      1.000000       1.000000\n",
       "..              ...           ...           ...            ...\n",
       "995       820361391      0.070233      0.016756       0.049938\n",
       "996       820363998      0.087782      0.010430       0.059146\n",
       "997       820365133      0.057598      0.010038       0.024419\n",
       "998       820372130      0.062613      0.029080       0.052061\n",
       "999       820376305      0.026491      0.025269       0.025838\n",
       "\n",
       "[1000 rows x 4 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "col = part.loc[: , \"payradio06\":\"payradio12\"]\n",
    "part['payradio_max'] = col.max(axis=1)\n",
    "part['payradio_min'] = col.min(axis=1)\n",
    "part['payradio_mean'] = col.mean(axis=1)\n",
    "part['Account_Number'] = df.Account_Number\n",
    "part = pd.DataFrame(part, columns = ['Account_Number','payradio_max','payradio_min', 'payradio_mean'])\n",
    "part"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Merge dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1000 entries, 0 to 999\n",
      "Data columns (total 6 columns):\n",
      " #   Column              Non-Null Count  Dtype   \n",
      "---  ------              --------------  -----   \n",
      " 0   Account_Number      1000 non-null   int64   \n",
      " 1   Age_By_2013_Bins    1000 non-null   category\n",
      " 2   Annual_Income_Bins  1000 non-null   category\n",
      " 3   Month_On_Book_Bins  1000 non-null   category\n",
      " 4   due_count           1000 non-null   int64   \n",
      " 5   max_recent          1000 non-null   int64   \n",
      "dtypes: category(3), int64(3)\n",
      "memory usage: 34.5 KB\n"
     ]
    }
   ],
   "source": [
    "cr = pd.merge(da, due1, on='Account_Number', how='outer')\n",
    "cr.info()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     Account_Number Age_By_2013_Bins Annual_Income_Bins Month_On_Book_Bins  \\\n",
      "0         820000011        age_to_46          Income_5k     book_after_145   \n",
      "1         820000046      age_above51          Income_5k     book_after_145   \n",
      "2         820000054        age_to_46          Income_5k     book_after_145   \n",
      "3         820000062        age_to_46          Income_5k     book_after_145   \n",
      "4         820000089      age_above51          Income_5k     book_after_145   \n",
      "..              ...              ...                ...                ...   \n",
      "995       820361391        age_to_46         Income_10k        book_to_145   \n",
      "996       820363998      age_above51         Income_10k        book_to_145   \n",
      "997       820365133        age_to_51         Income_15k        book_to_145   \n",
      "998       820372130      age_above51          Income_5k        book_to_145   \n",
      "999       820376305      age_above51          Income_5k        book_to_145   \n",
      "\n",
      "     due_count  max_recent Gender_Code   ur_mean    ur_max    ur_min  \n",
      "0            7           1           M  0.953253  0.977800  0.922212  \n",
      "1            0           0           F  0.244219  0.814330  0.000000  \n",
      "2            0           0           F  0.000000  0.000000  0.000000  \n",
      "3            0           0           M  0.000000  0.000000  0.000000  \n",
      "4            0           0           F  0.000000  0.000000  0.000000  \n",
      "..         ...         ...         ...       ...       ...       ...  \n",
      "995          6           2           M  1.003518  1.030840  0.994657  \n",
      "996          5           1           M  0.948818  0.996678  0.798997  \n",
      "997          7           1           F  0.342694  0.996219  0.142075  \n",
      "998          6           2           F  0.954781  0.994261  0.900987  \n",
      "999          7           1           M  0.980898  1.005070  0.958697  \n",
      "\n",
      "[1000 rows x 10 columns]\n"
     ]
    }
   ],
   "source": [
    "cred = pd.merge(cr, cre, on='Account_Number', how='outer')\n",
    "print(cred)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     Account_Number Age_By_2013_Bins Annual_Income_Bins Month_On_Book_Bins  \\\n",
      "0         820000011        age_to_46          Income_5k     book_after_145   \n",
      "1         820000046      age_above51          Income_5k     book_after_145   \n",
      "2         820000054        age_to_46          Income_5k     book_after_145   \n",
      "3         820000062        age_to_46          Income_5k     book_after_145   \n",
      "4         820000089      age_above51          Income_5k     book_after_145   \n",
      "..              ...              ...                ...                ...   \n",
      "995       820361391        age_to_46         Income_10k        book_to_145   \n",
      "996       820363998      age_above51         Income_10k        book_to_145   \n",
      "997       820365133        age_to_51         Income_15k        book_to_145   \n",
      "998       820372130      age_above51          Income_5k        book_to_145   \n",
      "999       820376305      age_above51          Income_5k        book_to_145   \n",
      "\n",
      "     due_count  max_recent Gender_Code   ur_mean    ur_max    ur_min  \\\n",
      "0            7           1           M  0.953253  0.977800  0.922212   \n",
      "1            0           0           F  0.244219  0.814330  0.000000   \n",
      "2            0           0           F  0.000000  0.000000  0.000000   \n",
      "3            0           0           M  0.000000  0.000000  0.000000   \n",
      "4            0           0           F  0.000000  0.000000  0.000000   \n",
      "..         ...         ...         ...       ...       ...       ...   \n",
      "995          6           2           M  1.003518  1.030840  0.994657   \n",
      "996          5           1           M  0.948818  0.996678  0.798997   \n",
      "997          7           1           F  0.342694  0.996219  0.142075   \n",
      "998          6           2           F  0.954781  0.994261  0.900987   \n",
      "999          7           1           M  0.980898  1.005070  0.958697   \n",
      "\n",
      "     payradio_max  payradio_min  payradio_mean  \n",
      "0        0.150850      0.040908       0.074926  \n",
      "1       17.018889      0.379312       3.954840  \n",
      "2        1.000000      1.000000       1.000000  \n",
      "3        1.000000      1.000000       1.000000  \n",
      "4        1.000000      1.000000       1.000000  \n",
      "..            ...           ...            ...  \n",
      "995      0.070233      0.016756       0.049938  \n",
      "996      0.087782      0.010430       0.059146  \n",
      "997      0.057598      0.010038       0.024419  \n",
      "998      0.062613      0.029080       0.052061  \n",
      "999      0.026491      0.025269       0.025838  \n",
      "\n",
      "[1000 rows x 13 columns]\n"
     ]
    }
   ],
   "source": [
    "c = pd.merge(cred, part, on='Account_Number', how='outer')\n",
    "print(c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Until now, we get dataset 'c' for 1000 rows"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = pd.get_dummies(c, columns=['Age_By_2013_Bins'])\n",
    "c = pd.get_dummies(c, columns=['Annual_Income_Bins'])\n",
    "c = pd.get_dummies(c, columns=['Month_On_Book_Bins'])\n",
    "c = pd.get_dummies(c, columns=['Gender_Code'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = c\n",
    "y = detail.Default"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1000 entries, 0 to 999\n",
      "Data columns (total 19 columns):\n",
      " #   Column                             Non-Null Count  Dtype  \n",
      "---  ------                             --------------  -----  \n",
      " 0   Account_Number                     1000 non-null   int64  \n",
      " 1   due_count                          1000 non-null   int64  \n",
      " 2   max_recent                         1000 non-null   int64  \n",
      " 3   ur_mean                            1000 non-null   float64\n",
      " 4   ur_max                             1000 non-null   float64\n",
      " 5   ur_min                             1000 non-null   float64\n",
      " 6   payradio_max                       988 non-null    float64\n",
      " 7   payradio_min                       988 non-null    float64\n",
      " 8   payradio_mean                      988 non-null    float64\n",
      " 9   Age_By_2013_Bins_age_to_46         1000 non-null   uint8  \n",
      " 10  Age_By_2013_Bins_age_to_51         1000 non-null   uint8  \n",
      " 11  Age_By_2013_Bins_age_above51       1000 non-null   uint8  \n",
      " 12  Annual_Income_Bins_Income_5k       1000 non-null   uint8  \n",
      " 13  Annual_Income_Bins_Income_10k      1000 non-null   uint8  \n",
      " 14  Annual_Income_Bins_Income_15k      1000 non-null   uint8  \n",
      " 15  Month_On_Book_Bins_book_to_145     1000 non-null   uint8  \n",
      " 16  Month_On_Book_Bins_book_after_145  1000 non-null   uint8  \n",
      " 17  Gender_Code_F                      1000 non-null   uint8  \n",
      " 18  Gender_Code_M                      1000 non-null   uint8  \n",
      "dtypes: float64(6), int64(3), uint8(10)\n",
      "memory usage: 87.9 KB\n"
     ]
    }
   ],
   "source": [
    "c.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.any(np.isnan(c))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.all(np.isfinite(c))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = c.replace(np.nan, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sklearn\n",
    "from sklearn import linear_model as lm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Establish the corresponding model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1) GLM model for first dataset 'c'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "####### Linear Regression Model ########\n",
      "Intercept:  -3135.966923814476\n",
      "Coefficient:  [ 3.82425202e-06 -2.74179131e-03  3.95180303e-02  6.27449644e-02\n",
      "  8.05027071e-03  3.81828785e-02 -3.86052717e-05  2.52687033e-02\n",
      "  6.12037251e-05  7.47878469e-03  9.43356359e-03 -1.69123483e-02\n",
      "  4.54528086e-03  1.94841758e-03 -6.49369845e-03 -1.25920359e-02\n",
      "  1.25920359e-02 -8.37147792e-03  8.37147792e-03]\n",
      "\n",
      "####### Logistic regression Model ########\n",
      "                 Generalized Linear Model Regression Results                  \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   No. Observations:                 1000\n",
      "Model:                            GLM   Df Residuals:                      985\n",
      "Model Family:                Binomial   Df Model:                           14\n",
      "Link Function:                  logit   Scale:                          1.0000\n",
      "Method:                          IRLS   Log-Likelihood:                    nan\n",
      "Date:                Mon, 01 Jun 2020   Deviance:                          nan\n",
      "Time:                        16:44:03   Pearson chi2:                 5.85e+17\n",
      "No. Iterations:                   100   Covariance Type:             nonrobust\n",
      "=====================================================================================================\n",
      "                                        coef    std err          z      P>|z|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------------\n",
      "Account_Number                        0.0004   9.41e-05      3.888      0.000       0.000       0.001\n",
      "due_count                            -0.6430      0.467     -1.376      0.169      -1.559       0.273\n",
      "max_recent                            2.7310      1.190      2.294      0.022       0.398       5.064\n",
      "ur_mean                               3.6567      7.411      0.493      0.622     -10.869      18.182\n",
      "ur_max                                1.9319      4.475      0.432      0.666      -6.838      10.702\n",
      "ur_min                               -1.6685      3.516     -0.475      0.635      -8.560       5.223\n",
      "payradio_max                         -0.2698      0.863     -0.313      0.755      -1.962       1.422\n",
      "payradio_min                         -9.6080     14.846     -0.647      0.518     -38.706      19.490\n",
      "payradio_mean                         0.8924      3.035      0.294      0.769      -5.056       6.840\n",
      "Age_By_2013_Bins_age_to_46        -3.157e+04   8119.649     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Age_By_2013_Bins_age_to_51        -3.157e+04   8119.599     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Age_By_2013_Bins_age_above51      -3.157e+04   8120.415     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Annual_Income_Bins_Income_5k      -3.157e+04   8119.804     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Annual_Income_Bins_Income_10k     -3.157e+04   8119.781     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Annual_Income_Bins_Income_15k     -3.157e+04   8120.079     -3.888      0.000   -4.75e+04   -1.57e+04\n",
      "Month_On_Book_Bins_book_to_145    -9.472e+04   2.44e+04     -3.888      0.000   -1.42e+05    -4.7e+04\n",
      "Month_On_Book_Bins_book_after_145     5.1369      1.321      3.888      0.000       2.547       7.726\n",
      "Gender_Code_F                     -4.736e+04   1.22e+04     -3.888      0.000   -7.12e+04   -2.35e+04\n",
      "Gender_Code_M                     -4.735e+04   1.22e+04     -3.888      0.000   -7.12e+04   -2.35e+04\n",
      "const                             -9.471e+04   2.44e+04     -3.888      0.000   -1.42e+05    -4.7e+04\n",
      "=====================================================================================================\n"
     ]
    }
   ],
   "source": [
    "#  Generalized Linear Model \n",
    "\n",
    "print('####### Linear Regression Model ########')\n",
    "# Create linear regression object\n",
    "lr = lm.LinearRegression()\n",
    "\n",
    "x= c.iloc[:, :] # independent variable\n",
    "y= list(y)         # dependent variable \n",
    "\n",
    "# Train the model using the training sets\n",
    "lr.fit(x, y)\n",
    "\n",
    "print (\"Intercept: \", lr.intercept_)\n",
    "print (\"Coefficient: \", lr.coef_)\n",
    "\n",
    "print('\\n####### Logistic regression Model ########')\n",
    "import statsmodels.api as sm\n",
    "\n",
    "# To be able to run GLM, we'll have to add the intercept constant to x variable\n",
    "x = sm.add_constant(x, prepend=False)\n",
    "\n",
    "# Instantiate a gaussian family model with the default link function.\n",
    "model = sm.GLM(y, x, family = sm.families.Binomial())\n",
    "model = model.fit()\n",
    "print (model.summary())\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# GLM Model explain:\n",
    "# p-value shows max_recent , age, annual income, month in bank, gender are statistically significant.\n",
    "# log(odds) ratio for max(credit in recent 3 months) =  2.4247\n",
    "# log(odds) ratio for months in books> 145) = 4.3949\n",
    "# Notice Log-likelihood is nan: it should be some log0 situation, we need to set the minmum predict value like 1e-10 should fix the problem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2 Multiple Logistic Regression "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Nomalized Dataset\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "sc.fit(c)\n",
    "X = sc.transform(c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "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>Account_Number</th>\n",
       "      <th>Default</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>820000011</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>820000046</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820000054</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>820000062</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>820000089</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>820361391</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>820363998</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>820365133</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>820372130</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>820376305</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Account_Number  Default\n",
       "0         820000011        0\n",
       "1         820000046        0\n",
       "2         820000054        0\n",
       "3         820000062        0\n",
       "4         820000089        0\n",
       "..              ...      ...\n",
       "995       820361391        1\n",
       "996       820363998        1\n",
       "997       820365133        1\n",
       "998       820372130        1\n",
       "999       820376305        1\n",
       "\n",
       "[1000 rows x 2 columns]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_g = pd.DataFrame(df, columns = ['Account_Number','Default'])\n",
    "y_g"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "# split data into train and test\n",
    "from sklearn.model_selection import train_test_split \n",
    "X_train, X_test, y_train, y_test = train_test_split(c, y_g.Default, test_size=0.3, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "105    0\n",
      "68     0\n",
      "479    0\n",
      "399    0\n",
      "434    0\n",
      "      ..\n",
      "835    0\n",
      "192    0\n",
      "629    0\n",
      "559    0\n",
      "684    0\n",
      "Name: Default, Length: 700, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=10, class_weight=None, dual=False, fit_intercept=True,\n",
       "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
       "                   multi_class='warn', n_jobs=None, penalty='l1',\n",
       "                   random_state=0, solver='warn', tol=0.0001, verbose=0,\n",
       "                   warm_start=False)"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "# l1 regularization gives better results\n",
    "lr = LogisticRegression(penalty='l1', C=10, random_state=0)\n",
    "lr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train - Accuracy : 0.9271428571428572\n",
      "Train - Confusion matrix : [[631   7]\n",
      " [ 44  18]]\n",
      "Train - classification report :               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.99      0.96       638\n",
      "           1       0.72      0.29      0.41        62\n",
      "\n",
      "    accuracy                           0.93       700\n",
      "   macro avg       0.83      0.64      0.69       700\n",
      "weighted avg       0.92      0.93      0.91       700\n",
      "\n",
      "Test - Accuracy : 0.89\n",
      "Test - Confusion matrix : [[260   2]\n",
      " [ 31   7]]\n",
      "Test - classification report :               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.89      0.99      0.94       262\n",
      "           1       0.78      0.18      0.30        38\n",
      "\n",
      "    accuracy                           0.89       300\n",
      "   macro avg       0.84      0.59      0.62       300\n",
      "weighted avg       0.88      0.89      0.86       300\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "# generate evaluation metrics\n",
    "print (\"Train - Accuracy :\", metrics.accuracy_score(y_train, lr.predict(X_train)))\n",
    "print (\"Train - Confusion matrix :\",metrics.confusion_matrix(y_train, lr.predict(X_train)))\n",
    "print (\"Train - classification report :\", metrics.classification_report(y_train, lr.predict(X_train)))\n",
    "\n",
    "print (\"Test - Accuracy :\", metrics.accuracy_score(y_test, lr.predict(X_test)))\n",
    "print (\"Test - Confusion matrix :\",metrics.confusion_matrix(y_test, lr.predict(X_test)))\n",
    "print (\"Test - classification report :\", metrics.classification_report(y_test, lr.predict(X_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Since we have inbalanced data, F1-score is better than accuracy.\n",
    "# F1-score for trainning dataset is good at \"good\" credit, but not for \"bad\" credit predict. The same with training dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "cm = confusion_matrix(y, model.predict(x).round())\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8, 8))\n",
    "ax.imshow(cm)\n",
    "ax.grid(False)\n",
    "ax.xaxis.set(ticks=(0, 1), ticklabels=('Predicted 0s', 'Predicted 1s'))\n",
    "ax.yaxis.set(ticks=(0, 1), ticklabels=('Actual 0s', 'Actual 1s'))\n",
    "ax.set_ylim(1.5, -0.5)\n",
    "for i in range(2):\n",
    "    for j in range(2):\n",
    "        ax.text(j, i, cm[i, j], ha='center', va='center', color='red')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LogisticRegression(C=10, class_weight=None, dual=False, fit_intercept=True,\n",
      "                   intercept_scaling=1, l1_ratio=None, max_iter=100,\n",
      "                   multi_class='warn', n_jobs=None, penalty='l1',\n",
      "                   random_state=0, solver='warn', tol=0.0001, verbose=0,\n",
      "                   warm_start=False)\n"
     ]
    }
   ],
   "source": [
    "print(lr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.89\n"
     ]
    }
   ],
   "source": [
    "# Use score method to get accuracy of model\n",
    "score = lr.score(X_test, y_test)\n",
    "print(score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# We get predict dataframe for training dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>Default</th>\n",
       "      <th>predict</th>\n",
       "      <th>accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>479</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>434</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>835</th>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>629</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>559</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>684</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>700 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Default  predict  accuracy\n",
       "105        0      0.0       0.0\n",
       "68         0      0.0       0.0\n",
       "479        0      0.0       0.0\n",
       "399        0      0.0       0.0\n",
       "434        0      0.0       0.0\n",
       "..       ...      ...       ...\n",
       "835        0      NaN       NaN\n",
       "192        0      0.0       0.0\n",
       "629        0      0.0       0.0\n",
       "559        0      0.0       0.0\n",
       "684        0      0.0       0.0\n",
       "\n",
       "[700 rows x 3 columns]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train_pre = pd.DataFrame({'Default': y_train})\n",
    "data_train_pre['predict'] = pd.DataFrame({'predict': lr.predict(X_train)}) \n",
    "data_train_pre['accuracy'] = data_train_pre['predict'] - data_train_pre['Default']\n",
    "data_train_pre"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "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>Default</th>\n",
       "      <th>predict</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>700.000000</td>\n",
       "      <td>497.000000</td>\n",
       "      <td>497.000000</td>\n",
       "      <td>497.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.088571</td>\n",
       "      <td>0.032193</td>\n",
       "      <td>0.032193</td>\n",
       "      <td>727.050488</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.284327</td>\n",
       "      <td>0.176691</td>\n",
       "      <td>0.176691</td>\n",
       "      <td>101.964273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1285.550540</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Default     predict    accuracy        score\n",
       "count  700.000000  497.000000  497.000000   497.000000\n",
       "mean     0.088571    0.032193    0.032193   727.050488\n",
       "std      0.284327    0.176691    0.176691   101.964273\n",
       "min      0.000000    0.000000    0.000000   708.472524\n",
       "25%      0.000000    0.000000    0.000000   708.472524\n",
       "50%      0.000000    0.000000    0.000000   708.472524\n",
       "75%      0.000000    0.000000    0.000000   708.472524\n",
       "max      1.000000    1.000000    1.000000  1285.550540"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# set up scoring logic\n",
    "\n",
    "target_score = 4000\n",
    "target_odds = 300\n",
    "pts_double_odds = 400\n",
    "\n",
    "factor = pts_double_odds / np.log(2)\n",
    "offset = target_score - factor * np.log(target_odds)\n",
    "\n",
    "# Score\n",
    "\n",
    "data_train_pre['score'] = offset + factor * data_train_pre['predict'].values\n",
    "data_train_pre.describe()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# We get predict dataframe for testing dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(lr.predict(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0\n",
      " 0 0 0 0]\n"
     ]
    }
   ],
   "source": [
    "print(lr.predict(X_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "d = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "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>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>300.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>725.784864</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>98.606617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>708.472524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1285.550540</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             score\n",
       "count   300.000000\n",
       "mean    725.784864\n",
       "std      98.606617\n",
       "min     708.472524\n",
       "25%     708.472524\n",
       "50%     708.472524\n",
       "75%     708.472524\n",
       "max    1285.550540"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# set up scoring logic\n",
    "\n",
    "target_score = 4000\n",
    "target_odds = 300\n",
    "pts_double_odds = 400\n",
    "\n",
    "factor = pts_double_odds / np.log(2)\n",
    "offset = target_score - factor * np.log(target_odds)\n",
    "\n",
    "# Score\n",
    "\n",
    "d['score'] = offset + factor * lr.predict(X_test)\n",
    "d.describe()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Until now, we got all the predict scores for samples in 2014 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# let's see for 'max_recent' variables for detail"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1000"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x1 = c.max_recent[:, np.newaxis]\n",
    "len(x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1000"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y1 = detail.Default\n",
    "len(y1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R Squared :  0.1137692479673077\n",
      "Mean Absolute Error:  0.15952153536588473\n",
      "Root Mean Squared Error:  0.2824194888511456\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": [
    "\n",
    "\n",
    "# Create linear regression object\n",
    "lr = lm.LinearRegression()\n",
    "\n",
    "# Train the model using the training sets\n",
    "lr.fit(x1, y1)\n",
    "\n",
    "# plotting fitted line\n",
    "plt.scatter(x1, y1,  color='black')\n",
    "plt.plot(x1, lr.predict(x1), color='blue', linewidth=3)\n",
    "plt.title('max_recent vs Credit')\n",
    "plt.ylabel('Credit')\n",
    "plt.xlabel('max_recent')\n",
    "\n",
    "# add predict value to the data frame\n",
    "detail['Default_pred'] = lr.predict(x1)\n",
    "\n",
    "# Using built-in function\n",
    "print (\"R Squared : \", r2_score(detail.Default, detail.Default_pred))\n",
    "print (\"Mean Absolute Error: \", mean_absolute_error(detail.Default, detail.Default_pred))\n",
    "print (\"Root Mean Squared Error: \", np.sqrt(mean_squared_error(detail.Default, detail.Default_pred)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = sorted(detail.Default)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.10050048 0.03839163 0.03839163 0.03839163 0.03839163 0.4666763\n",
      " 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048\n",
      " 0.10050048 0.23819801 0.03839163 0.03839163 0.23819801 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.23819801 0.10050048 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048 0.03839163\n",
      " 0.23819801 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.03839163 0.23819801\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163\n",
      " 0.03839163 0.03839163 0.10050048 0.23819801 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.10050048 0.23819801\n",
      " 0.23819801 0.10050048 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.23819801 0.03839163 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.03839163 0.10050048\n",
      " 0.23819801 0.03839163 0.10050048 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048 0.03839163\n",
      " 0.23819801 0.03839163 0.10050048 0.23819801 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.23819801 0.03839163 0.10050048 0.23819801\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.03839163 0.03839163 0.23819801\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163 0.23819801\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048\n",
      " 0.23819801 0.03839163 0.23819801 0.10050048 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.23819801 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163 0.03839163\n",
      " 0.03839163 0.4666763  0.10050048 0.10050048 0.23819801 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.23819801 0.03839163 0.03839163\n",
      " 0.03839163 0.10050048 0.23819801 0.03839163 0.03839163 0.03839163\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.23819801 0.23819801 0.10050048 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.4666763  0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.23819801 0.23819801 0.03839163\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.23819801 0.10050048 0.03839163 0.03839163 0.23819801\n",
      " 0.23819801 0.03839163 0.10050048 0.23819801 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.23819801 0.03839163 0.10050048\n",
      " 0.23819801 0.23819801 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.23819801 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163 0.23819801\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.03839163 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.23819801 0.10050048\n",
      " 0.23819801 0.03839163 0.10050048 0.23819801 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.03839163 0.4666763  0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.23819801 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.03839163 0.10050048 0.23819801 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.23819801 0.4666763  0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048 0.4666763\n",
      " 0.03839163 0.03839163 0.03839163 0.23819801 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.23819801 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.4666763  0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.10050048 0.23819801 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.4666763  0.23819801 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.23819801 0.10050048 0.4666763\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.23819801 0.10050048 0.10050048 0.10050048 0.10050048 0.4666763\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048 0.23819801\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.03839163 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.10050048 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.03839163 0.03839163\n",
      " 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163 0.23819801\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.23819801 0.10050048 0.10050048 0.71004425 0.03839163 0.23819801\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.23819801 0.23819801\n",
      " 0.03839163 0.03839163 0.23819801 0.10050048 0.10050048 0.10050048\n",
      " 0.4666763  0.10050048 0.23819801 0.03839163 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.03839163 0.03839163 0.23819801 0.23819801\n",
      " 0.10050048 0.23819801 0.10050048 0.03839163 0.03839163 0.23819801\n",
      " 0.10050048 0.10050048 0.10050048 0.03839163 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.23819801 0.23819801 0.10050048 0.23819801 0.4666763\n",
      " 0.03839163 0.03839163 0.23819801 0.10050048 0.10050048 0.23819801\n",
      " 0.23819801 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.10050048 0.03839163 0.10050048 0.23819801\n",
      " 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048 0.03839163\n",
      " 0.03839163 0.23819801 0.10050048 0.23819801 0.03839163 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.4666763  0.23819801 0.10050048\n",
      " 0.10050048 0.10050048 0.4666763  0.03839163 0.10050048 0.03839163\n",
      " 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.03839163 0.4666763  0.10050048 0.03839163 0.10050048\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.10050048 0.23819801 0.23819801 0.10050048 0.03839163\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.03839163 0.10050048 0.4666763  0.03839163 0.03839163 0.23819801\n",
      " 0.10050048 0.23819801 0.10050048 0.10050048 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.03839163 0.03839163 0.10050048 0.03839163\n",
      " 0.10050048 0.10050048 0.03839163 0.23819801 0.10050048 0.03839163\n",
      " 0.03839163 0.03839163 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.03839163 0.03839163\n",
      " 0.10050048 0.03839163 0.10050048 0.23819801 0.10050048 0.03839163\n",
      " 0.03839163 0.10050048 0.23819801 0.4666763  0.10050048 0.10050048\n",
      " 0.03839163 0.03839163 0.03839163 0.03839163 0.10050048 0.23819801\n",
      " 0.03839163 0.03839163 0.23819801 0.23819801 0.4666763  0.03839163\n",
      " 0.03839163 0.10050048 0.23819801 0.03839163 0.23819801 0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.10050048 0.23819801 0.23819801\n",
      " 0.10050048 0.23819801 0.23819801 0.10050048 0.10050048 0.10050048\n",
      " 0.10050048 0.10050048 0.23819801 0.10050048 0.10050048 0.23819801\n",
      " 0.4666763  0.23819801 0.71004425 0.71004425 0.23819801 0.10050048\n",
      " 0.10050048 0.23819801 0.10050048 0.71004425 0.10050048 0.4666763\n",
      " 0.23819801 0.4666763  0.10050048 0.10050048 0.03839163 0.23819801\n",
      " 0.10050048 0.71004425 0.23819801 0.10050048 0.10050048 0.71004425\n",
      " 0.10050048 0.10050048 0.23819801 0.23819801 0.10050048 0.4666763\n",
      " 0.71004425 0.10050048 0.10050048 0.23819801 0.23819801 0.23819801\n",
      " 0.10050048 0.03839163 0.03839163 0.71004425 0.4666763  0.10050048\n",
      " 0.10050048 0.10050048 0.03839163 0.10050048 0.10050048 0.10050048\n",
      " 0.23819801 0.10050048 0.10050048 0.23819801 0.10050048 0.23819801\n",
      " 0.10050048 0.10050048 0.10050048 0.23819801 0.4666763  0.10050048\n",
      " 0.10050048 0.10050048 0.10050048 0.23819801 0.03839163 0.23819801\n",
      " 0.10050048 0.10050048 0.23819801 0.10050048]\n",
      "[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0 0 0 0 0 0 0 1 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0 1 0\n",
      " 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n",
      " 0]\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": [
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "# manually add intercept\n",
    "#detail['intercept'] = 1\n",
    "#independent_variables = [x1.max_recent, 'intercept']\n",
    "\n",
    "#x1 = detail[independent_variables]       # independent variable\n",
    "#y = detail['Default']                      # dependent variable \n",
    "\n",
    "# instantiate a logistic regression model, and fit with X and y\n",
    "model = LogisticRegression()\n",
    "model = model.fit(x1, y)\n",
    "\n",
    "# check the accuracy on the training set\n",
    "model.score(x1, y)\n",
    "\n",
    "# predict_proba will return array containing proability of y = 0 and y = 1\n",
    "print (model.predict_proba(x1)[:,1])\n",
    "\n",
    "# predict will give convert the probability(y=1) values > .5 to 1 else 0  \n",
    "print (model.predict(x1))\n",
    "\n",
    "# plotting fitted line\n",
    "plt.scatter(detail.Annual_Income, y,  color='black')\n",
    "plt.yticks([0.0, 0.5, 1.0])\n",
    "plt.plot(detail.Annual_Income, model.predict_proba(x1)[:,1], color='blue', linewidth=3)\n",
    "plt.title('max_recent vs Default')\n",
    "plt.ylabel('Credit')\n",
    "plt.xlabel('max_recent')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy : 0.906\n",
      "AUC : 0.7448777777777777\n",
      "Confusion matrix : [[899   1]\n",
      " [ 93   7]]\n",
      "classification report :               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.91      1.00      0.95       900\n",
      "           1       0.88      0.07      0.13       100\n",
      "\n",
      "    accuracy                           0.91      1000\n",
      "   macro avg       0.89      0.53      0.54      1000\n",
      "weighted avg       0.90      0.91      0.87      1000\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "\n",
    "# generate evaluation metrics\n",
    "print (\"Accuracy :\", metrics.accuracy_score(y, model.predict(x1)))\n",
    "print (\"AUC :\", metrics.roc_auc_score(y, model.predict_proba(x1)[:,1]))\n",
    "\n",
    "print (\"Confusion matrix :\",metrics.confusion_matrix(y, model.predict(x1)))\n",
    "print (\"classification report :\", metrics.classification_report(y, model.predict(x1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ROC AUC: 0.74\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": [
    "# Determine the false positive and true positive rates\n",
    "fpr, tpr, _ = metrics.roc_curve(y, model.predict_proba(x1)[:,1])\n",
    " \n",
    "# Calculate the AUC\n",
    "roc_auc = metrics.auc(fpr, tpr)\n",
    "print ('ROC AUC: %0.2f' % roc_auc)\n",
    " \n",
    "# Plot of a ROC curve for a specific class\n",
    "plt.figure()\n",
    "plt.plot(fpr, tpr, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC Curve')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ROC curve = .74, it shows thethold around (0.4, 0.8) might be better for predict when we consider max bad credit in recent 3 months."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# File line "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C = 1 (default), Accuracy : 0.906\n",
      "C = 10, Accuracy : 0.906\n",
      "C = 100, Accuracy : 0.906\n",
      "C = 1000, Accuracy : 0.906\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": [
    "# instantiate a logistic regression model with default c value, and fit with X and y\n",
    "model = LogisticRegression()\n",
    "model = model.fit(x1, y)\n",
    "\n",
    "# check the accuracy on the training set\n",
    "print (\"C = 1 (default), Accuracy :\", metrics.accuracy_score(y, model.predict(x1)))\n",
    "\n",
    "# instantiate a logistic regression model with c = 10, and fit with X and y\n",
    "model1 = LogisticRegression(C=10)\n",
    "model1 = model1.fit(x1, y)\n",
    "\n",
    "# check the accuracy on the training set\n",
    "print (\"C = 10, Accuracy :\", metrics.accuracy_score(y, model1.predict(x1)))\n",
    "\n",
    "# instantiate a logistic regression model with c = 100, and fit with X and y\n",
    "model2 = LogisticRegression(C=100)\n",
    "model2 = model2.fit(x1, y)\n",
    "\n",
    "# check the accuracy on the training set\n",
    "print (\"C = 100, Accuracy :\", metrics.accuracy_score(y, model2.predict(x1)))\n",
    "\n",
    "# instantiate a logistic regression model with c = 1000, and fit with X and y\n",
    "model3 = LogisticRegression(C=1000)\n",
    "model3 = model3.fit(x1, y)\n",
    "\n",
    "# check the accuracy on the training set\n",
    "print (\"C = 1000, Accuracy :\", metrics.accuracy_score(y, model3.predict(x1)))\n",
    "\n",
    "\n",
    "# plotting fitted line\n",
    "plt.scatter(detail.Month_On_Book, y,  color='black', label='Result')\n",
    "plt.yticks([0.0, 0.5, 1.0])\n",
    "plt.plot(c.max_recent, model.predict_proba(x1)[:,1], color='gray', linewidth=2, label='C=1.0')\n",
    "plt.plot(c.max_recent, model1.predict_proba(x1)[:,1], color='blue', linewidth=2,label='C=10')\n",
    "plt.plot(c.max_recent, model2.predict_proba(x1)[:,1], color='green', linewidth=2,label='C=100')\n",
    "plt.plot(c.max_recent, model3.predict_proba(x1)[:,1], color='red', linewidth=2,label='C=1000')\n",
    "plt.legend(loc='lower right') # legend location\n",
    "plt.title('max_recent  vs Default')\n",
    "plt.ylabel('Credit')\n",
    "plt.xlabel('max_recent')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# F1-score is better for 'good' credit but not for 'bad' credit person. \n",
    "\n",
    "# In summary, Multiple variables Logistic regression model predicts good for 'good' credit sample but not so good for 'bad' samples. \n",
    "\n",
    "# Private informations like age, annual income, most recent 3 months credit, and months booked in bank is statistically significant to predict 2014's credit. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}