doc/eda.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataTransformerRegistry.enable('json')"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "import altair as alt\n",
    "import selenium\n",
    "alt.data_transformers.enable('json')\n",
    "#alt.renderers.enable('notebook')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data description\n",
    "\n",
    "The dataset we are analyzing is from [UCI machine learning repository](https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients). It's about customers default payments in Taiwan, if the customer will fail to pay in October 2005(`DEFAULT_NEXT_MONTH`). There are 30,000 instances in total and 23 exploratory variables about customers' payment, bill histories and demographic factors. Details are as below according to the data description from UCI,\n",
    "\n",
    "- `LIMIT_BAL`: Amount of the given credit (NT dollar): it includes both the individual consumer credit and his/her family (supplementary) credit. \n",
    "- `SEX`: Gender(1 = male; 2 = female).\n",
    "- `EDUCATION`: Education (1 = graduate school; 2 = university; 3 = high school; 4 = others). \n",
    "- `MARRIAGE`: Marital status (1 = married; 2 = single; 3 = others).  \n",
    "- `AGE`: Age (year).  \n",
    "- `PAY_1`, `PAY_2`, ..., `PAY_6`: Past monthly repayment status in September 2005, August 2005, ..., April 2005 respectively. ( -1 = pay duly; 1 = payment delay for one month; 2 = payment delay for two months; . . .; 8 = payment delay for eight months; 9 = payment delay for nine months and above.)  \n",
    "- `BILL_AMT1`, `BILL_AMT2`, ..., `BILL_AMT6`: Amount of bill statement (NT dollar) in September 2005, August 2005, ..., April 2005 respectively.  \n",
    "- `PAY_AMT1`, `PAY_AMT2`, ..., `PAY_AMT6`: Amount of previous payment (NT dollar) in September 2005, August 2005, ..., April 2005 respectively.  \n",
    "\n",
    "We are interested to build a good model to predict if the customers will default in the next month which is in October 2005 accurately and thus find out what features strongly predict default payment from the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('../data/credit-default-data.csv', index_col=1, skiprows=1)\n",
    "df = df.drop(columns=['0'])\n",
    "df = df.rename(columns={'default payment next month':'DEFAULT_NEXT_MONTH',\n",
    "                        'PAY_0': 'PAY_1'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert df.shape[0] == 30000\n",
    "assert df.shape[1] == 24\n",
    "assert set(df.columns) == set(['LIMIT_BAL', 'SEX', 'EDUCATION', 'MARRIAGE', 'AGE', 'PAY_1', 'PAY_2',\n",
    "       'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6', 'BILL_AMT1', 'BILL_AMT2',\n",
    "       'BILL_AMT3', 'BILL_AMT4', 'BILL_AMT5', 'BILL_AMT6', 'PAY_AMT1',\n",
    "       'PAY_AMT2', 'PAY_AMT3', 'PAY_AMT4', 'PAY_AMT5', 'PAY_AMT6',\n",
    "       'DEFAULT_NEXT_MONTH'])"
   ]
  },
  {
   "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>LIMIT_BAL</th>\n",
       "      <th>SEX</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>MARRIAGE</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_1</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>PAY_4</th>\n",
       "      <th>PAY_5</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <th>DEFAULT_NEXT_MONTH</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <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>29996</th>\n",
       "      <td>220000</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>39</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>88004</td>\n",
       "      <td>31237</td>\n",
       "      <td>15980</td>\n",
       "      <td>8500</td>\n",
       "      <td>20000</td>\n",
       "      <td>5003</td>\n",
       "      <td>3047</td>\n",
       "      <td>5000</td>\n",
       "      <td>1000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29997</th>\n",
       "      <td>150000</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>8979</td>\n",
       "      <td>5190</td>\n",
       "      <td>0</td>\n",
       "      <td>1837</td>\n",
       "      <td>3526</td>\n",
       "      <td>8998</td>\n",
       "      <td>129</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29998</th>\n",
       "      <td>30000</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>37</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>20878</td>\n",
       "      <td>20582</td>\n",
       "      <td>19357</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22000</td>\n",
       "      <td>4200</td>\n",
       "      <td>2000</td>\n",
       "      <td>3100</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29999</th>\n",
       "      <td>80000</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>52774</td>\n",
       "      <td>11855</td>\n",
       "      <td>48944</td>\n",
       "      <td>85900</td>\n",
       "      <td>3409</td>\n",
       "      <td>1178</td>\n",
       "      <td>1926</td>\n",
       "      <td>52964</td>\n",
       "      <td>1804</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30000</th>\n",
       "      <td>50000</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>46</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>36535</td>\n",
       "      <td>32428</td>\n",
       "      <td>15313</td>\n",
       "      <td>2078</td>\n",
       "      <td>1800</td>\n",
       "      <td>1430</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       LIMIT_BAL  SEX  EDUCATION  MARRIAGE  AGE  PAY_1  PAY_2  PAY_3  PAY_4  \\\n",
       "ID                                                                            \n",
       "29996     220000    1          3         1   39      0      0      0      0   \n",
       "29997     150000    1          3         2   43     -1     -1     -1     -1   \n",
       "29998      30000    1          2         2   37      4      3      2     -1   \n",
       "29999      80000    1          3         1   41      1     -1      0      0   \n",
       "30000      50000    1          2         1   46      0      0      0      0   \n",
       "\n",
       "       PAY_5  ...  BILL_AMT4  BILL_AMT5  BILL_AMT6  PAY_AMT1  PAY_AMT2  \\\n",
       "ID            ...                                                        \n",
       "29996      0  ...      88004      31237      15980      8500     20000   \n",
       "29997      0  ...       8979       5190          0      1837      3526   \n",
       "29998      0  ...      20878      20582      19357         0         0   \n",
       "29999      0  ...      52774      11855      48944     85900      3409   \n",
       "30000      0  ...      36535      32428      15313      2078      1800   \n",
       "\n",
       "       PAY_AMT3  PAY_AMT4  PAY_AMT5  PAY_AMT6  DEFAULT_NEXT_MONTH  \n",
       "ID                                                                 \n",
       "29996      5003      3047      5000      1000                   0  \n",
       "29997      8998       129         0         0                   0  \n",
       "29998     22000      4200      2000      3100                   1  \n",
       "29999      1178      1926     52964      1804                   1  \n",
       "30000      1430      1000      1000      1000                   1  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop(columns=['DEFAULT_NEXT_MONTH'])\n",
    "y = df['DEFAULT_NEXT_MONTH']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of train examples:  22500\n",
      "Number of test examples:  7500\n"
     ]
    }
   ],
   "source": [
    "##split the data to training and test set and do EDA on the training set only\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=122)\n",
    "print('Number of train examples: ', len(y_train))\n",
    "print('Number of test examples: ', len(y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "numeric_features = ['LIMIT_BAL', 'AGE', 'BILL_AMT1', 'BILL_AMT2',\n",
    "       'BILL_AMT3', 'BILL_AMT4', 'BILL_AMT5', 'BILL_AMT6', 'PAY_AMT1',\n",
    "       'PAY_AMT2', 'PAY_AMT3', 'PAY_AMT4', 'PAY_AMT5', 'PAY_AMT6']\n",
    "\n",
    "categorical_features = ['SEX', 'EDUCATION', 'MARRIAGE', 'PAY_1', 'PAY_2',\n",
    "       'PAY_3', 'PAY_4', 'PAY_5', 'PAY_6']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>LIMIT_BAL</th>\n",
       "      <th>AGE</th>\n",
       "      <th>BILL_AMT1</th>\n",
       "      <th>BILL_AMT2</th>\n",
       "      <th>BILL_AMT3</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.00000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>2.250000e+04</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "      <td>22500.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>167229.763556</td>\n",
       "      <td>35.487022</td>\n",
       "      <td>50992.89800</td>\n",
       "      <td>48905.718978</td>\n",
       "      <td>46629.685644</td>\n",
       "      <td>42932.418844</td>\n",
       "      <td>39905.282444</td>\n",
       "      <td>38385.688222</td>\n",
       "      <td>5714.377733</td>\n",
       "      <td>5.848260e+03</td>\n",
       "      <td>5132.902667</td>\n",
       "      <td>4728.448311</td>\n",
       "      <td>4725.760978</td>\n",
       "      <td>5282.126533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>129384.485693</td>\n",
       "      <td>9.182223</td>\n",
       "      <td>73064.68632</td>\n",
       "      <td>70748.066294</td>\n",
       "      <td>68376.985307</td>\n",
       "      <td>63802.950987</td>\n",
       "      <td>60135.853082</td>\n",
       "      <td>58733.428102</td>\n",
       "      <td>17078.235838</td>\n",
       "      <td>2.191690e+04</td>\n",
       "      <td>16892.473653</td>\n",
       "      <td>15430.720628</td>\n",
       "      <td>15138.455175</td>\n",
       "      <td>18506.384982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>21.000000</td>\n",
       "      <td>-165580.00000</td>\n",
       "      <td>-69777.000000</td>\n",
       "      <td>-157264.000000</td>\n",
       "      <td>-170000.000000</td>\n",
       "      <td>-81334.000000</td>\n",
       "      <td>-339603.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</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>50000.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>3565.75000</td>\n",
       "      <td>2928.000000</td>\n",
       "      <td>2577.000000</td>\n",
       "      <td>2313.000000</td>\n",
       "      <td>1711.750000</td>\n",
       "      <td>1190.000000</td>\n",
       "      <td>990.000000</td>\n",
       "      <td>8.000000e+02</td>\n",
       "      <td>390.000000</td>\n",
       "      <td>285.750000</td>\n",
       "      <td>238.000000</td>\n",
       "      <td>119.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>140000.000000</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>22169.00000</td>\n",
       "      <td>20859.000000</td>\n",
       "      <td>19889.000000</td>\n",
       "      <td>18855.500000</td>\n",
       "      <td>17875.000000</td>\n",
       "      <td>16715.000000</td>\n",
       "      <td>2100.000000</td>\n",
       "      <td>2.001000e+03</td>\n",
       "      <td>1800.000000</td>\n",
       "      <td>1500.000000</td>\n",
       "      <td>1500.000000</td>\n",
       "      <td>1500.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>240000.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>66732.75000</td>\n",
       "      <td>63104.250000</td>\n",
       "      <td>59532.500000</td>\n",
       "      <td>53339.500000</td>\n",
       "      <td>49743.000000</td>\n",
       "      <td>48863.500000</td>\n",
       "      <td>5006.000000</td>\n",
       "      <td>5.000000e+03</td>\n",
       "      <td>4512.000000</td>\n",
       "      <td>4000.000000</td>\n",
       "      <td>4000.000000</td>\n",
       "      <td>4000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>800000.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>746814.00000</td>\n",
       "      <td>743970.000000</td>\n",
       "      <td>855086.000000</td>\n",
       "      <td>616836.000000</td>\n",
       "      <td>587067.000000</td>\n",
       "      <td>568638.000000</td>\n",
       "      <td>873552.000000</td>\n",
       "      <td>1.227082e+06</td>\n",
       "      <td>889043.000000</td>\n",
       "      <td>621000.000000</td>\n",
       "      <td>426529.000000</td>\n",
       "      <td>528666.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           LIMIT_BAL           AGE     BILL_AMT1      BILL_AMT2  \\\n",
       "count   22500.000000  22500.000000   22500.00000   22500.000000   \n",
       "mean   167229.763556     35.487022   50992.89800   48905.718978   \n",
       "std    129384.485693      9.182223   73064.68632   70748.066294   \n",
       "min     10000.000000     21.000000 -165580.00000  -69777.000000   \n",
       "25%     50000.000000     28.000000    3565.75000    2928.000000   \n",
       "50%    140000.000000     34.000000   22169.00000   20859.000000   \n",
       "75%    240000.000000     41.000000   66732.75000   63104.250000   \n",
       "max    800000.000000     75.000000  746814.00000  743970.000000   \n",
       "\n",
       "           BILL_AMT3      BILL_AMT4      BILL_AMT5      BILL_AMT6  \\\n",
       "count   22500.000000   22500.000000   22500.000000   22500.000000   \n",
       "mean    46629.685644   42932.418844   39905.282444   38385.688222   \n",
       "std     68376.985307   63802.950987   60135.853082   58733.428102   \n",
       "min   -157264.000000 -170000.000000  -81334.000000 -339603.000000   \n",
       "25%      2577.000000    2313.000000    1711.750000    1190.000000   \n",
       "50%     19889.000000   18855.500000   17875.000000   16715.000000   \n",
       "75%     59532.500000   53339.500000   49743.000000   48863.500000   \n",
       "max    855086.000000  616836.000000  587067.000000  568638.000000   \n",
       "\n",
       "            PAY_AMT1      PAY_AMT2       PAY_AMT3       PAY_AMT4  \\\n",
       "count   22500.000000  2.250000e+04   22500.000000   22500.000000   \n",
       "mean     5714.377733  5.848260e+03    5132.902667    4728.448311   \n",
       "std     17078.235838  2.191690e+04   16892.473653   15430.720628   \n",
       "min         0.000000  0.000000e+00       0.000000       0.000000   \n",
       "25%       990.000000  8.000000e+02     390.000000     285.750000   \n",
       "50%      2100.000000  2.001000e+03    1800.000000    1500.000000   \n",
       "75%      5006.000000  5.000000e+03    4512.000000    4000.000000   \n",
       "max    873552.000000  1.227082e+06  889043.000000  621000.000000   \n",
       "\n",
       "            PAY_AMT5       PAY_AMT6  \n",
       "count   22500.000000   22500.000000  \n",
       "mean     4725.760978    5282.126533  \n",
       "std     15138.455175   18506.384982  \n",
       "min         0.000000       0.000000  \n",
       "25%       238.000000     119.750000  \n",
       "50%      1500.000000    1500.000000  \n",
       "75%      4000.000000    4000.000000  \n",
       "max    426529.000000  528666.000000  "
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[numeric_features].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the above table, we can see mean, standard deviation, min, max etc. The bill amount, payment amount and credit limit ranges are roughly similar which are around 800,000. It's interesting that The medians for the bill statement amounts are around 20,000, but the medians for payment amounts are 2,000. Age ranges from 21 to 75 which is reasonable."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, there are so many negative numbers for bill amount. Let's see if that's some kind of errors. Below I pulled some rows that have negative values at `BILL_AMT6'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_neg = df[df.BILL_AMT6<-50000][numeric_features].copy()\n",
    "df_neg['DEFAULT_NEXT_MONTH'] = df[df.BILL_AMT6<-50000]['DEFAULT_NEXT_MONTH']\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the table, the customer paid big amount at that month for some reason, but the amount matches basically."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One issue with the dataset is that there are many entries under repayment status columns of values 0 These values should be considered N/A. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total NA columns 14542\n",
      "percentage of NA columns:48%\n"
     ]
    }
   ],
   "source": [
    "NA_columns = df[(df.PAY_AMT1 == 0) | (df.PAY_AMT2 == 0) | (df.PAY_AMT3 == 0) | (df.PAY_AMT4 == 0) | (df.PAY_AMT5 == 0) | (df.PAY_AMT6 == 0)].shape[0] \n",
    "total_columns = df.shape[0]\n",
    "print(f\"Total NA columns {NA_columns}\")\n",
    "print(f\"percentage of NA columns:{round(NA_columns/total_columns * 100)}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are several options we can choose to deal with these missing values. We could omit the rows with 0 values and just focus our analysis on the 52% of data. Since the 0 values represent a big portion of the data, and because the missing values may not be random we may also choose to use those N/A rows by classifying the missing data as N/A rather than 0."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Numeric Features  \n",
    "Let's see the histograms of numeric features distributions for easier comparisons. There are some outliers for the bill statement amounts and payment amounts, but mode ranges are consistent. Because we are going to normalize the number when fitting the predicting model, I don't think it would be an issue."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def num_bar(data, fea):\n",
    "    return alt.Chart(data).mark_bar(opacity=0.7).encode(\n",
    "        alt.X(fea, type='quantitative', bin=alt.Bin(maxbins=10)),\n",
    "        alt.Y('count()')\n",
    "    ).properties(title=f'Count of {fea}', height=150, width=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = alt.vconcat()\n",
    "col = alt.hconcat()\n",
    "ncol = 4  #how many to display in each row\n",
    "\n",
    "for i in range(len(numeric_features)):\n",
    "    col |= num_bar(X_train, numeric_features[i])\n",
    "    if (i+1)%ncol==0:\n",
    "        row &= col\n",
    "        col = alt.hconcat() \n",
    "num_bar_chart = row & col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_bar_chart.save('../results/num_bar_chart.png', webdriver='firefox')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-ffdf232926808bd25ec5ef105eb570ec.json"
       },
       "vconcat": [
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "LIMIT_BAL",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of LIMIT_BAL",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "AGE",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of AGE",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT1",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT1",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT2",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT2",
           "width": 200
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT3",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT3",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT4",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT4",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT5",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT5",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "BILL_AMT6",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of BILL_AMT6",
           "width": 200
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT1",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT1",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT2",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT2",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT3",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT3",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT4",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT4",
           "width": 200
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT5",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT5",
           "width": 200
          },
          {
           "encoding": {
            "x": {
             "bin": {
              "maxbins": 10
             },
             "field": "PAY_AMT6",
             "type": "quantitative"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 150,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_AMT6",
           "width": 200
          }
         ]
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_bar_chart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Categorical features  \n",
    "How many people are in each class of each categorical feature? Bar chart is easy to compare the difference.  \n",
    "For example, from the following plots, for Count of SEX(gender) we can see class 2 - Female is almost twice the number of the class 1 - Male.\n",
    "\n",
    "There are lots of 0s for `PAY_1`, `PAY_2`, ..., `PAY_6` which are past monthly status. 0 is not documented in the original data description. I guess it probably means \"unknown\". But because we are fitting the prediction model, we don't really need to know the meaning of 0. It's still some sort of the information about the status, so we'll leave as is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cat_bar(data, fea):\n",
    "    return alt.Chart(data).mark_bar(opacity=0.7).encode(\n",
    "        alt.X(fea, type='nominal'),\n",
    "        alt.Y('count()')\n",
    "    ).properties(title=f'Count of {fea}', height=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = alt.vconcat()\n",
    "col = alt.hconcat()\n",
    "ncol = 3  #how many to display in each row\n",
    "\n",
    "for i in range(len(categorical_features)):\n",
    "    col |= cat_bar(X_train, categorical_features[i])\n",
    "    if (i+1)%ncol==0:\n",
    "        row &= col\n",
    "        col = alt.hconcat()  \n",
    "cat_bar_chart = row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-ffdf232926808bd25ec5ef105eb570ec.json"
       },
       "vconcat": [
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "SEX",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of SEX"
          },
          {
           "encoding": {
            "x": {
             "field": "EDUCATION",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of EDUCATION"
          },
          {
           "encoding": {
            "x": {
             "field": "MARRIAGE",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of MARRIAGE"
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "PAY_1",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_1"
          },
          {
           "encoding": {
            "x": {
             "field": "PAY_2",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_2"
          },
          {
           "encoding": {
            "x": {
             "field": "PAY_3",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_3"
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "PAY_4",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_4"
          },
          {
           "encoding": {
            "x": {
             "field": "PAY_5",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_5"
          },
          {
           "encoding": {
            "x": {
             "field": "PAY_6",
             "type": "nominal"
            },
            "y": {
             "aggregate": "count",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "title": "Count of PAY_6"
          }
         ]
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_bar_chart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The response variable  \n",
    "The response variable has two classes. 1 - Default payment is about a third of 0 - Non default. It's imbalanced dataset, but there are still around 5000 examples for default payment to train the model, I don't think it's too bad."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    17519\n",
       "1     4981\n",
       "Name: DEFAULT_NEXT_MONTH, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-c43837ad50d1dd53c60fec530b24704a.json"
       },
       "encoding": {
        "x": {
         "field": "DEFAULT_NEXT_MONTH",
         "type": "nominal"
        },
        "y": {
         "aggregate": "count",
         "type": "quantitative"
        }
       },
       "height": 200,
       "mark": {
        "opacity": 0.7,
        "type": "bar"
       },
       "title": "Count of DEFAULT_NEXT_MONTH"
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_bar(y_train.to_frame(), \n",
    "        'DEFAULT_NEXT_MONTH')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Possible relationships  \n",
    "Let's see if we can find some relationships or patterns through plots, so that it makes sense to predict based on these predictor variables.  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Correlations among numeric predictors  \n",
    "Collinearity among predictors would be a problem for some models, for example, logistic regression. If they are strongly correlated, the weights we get from the fitted logistic regression model cannot be trusted.  \n",
    "According to the below plot, bill statement amounts are highly correlated. It makes sense, because the amount owed this month is dependent on the amount owed from last month. So we might need to drop some columns when fitting the model or choose the model that won't be affected by the collinearity of predictors. Other than that, payment amounts, bill statement amounts and age are relatively independent of each other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "##TOO SLOW to run\n",
    "\n",
    "#num_corr = alt.Chart(X_train).mark_circle().encode(\n",
    "#     alt.X(alt.repeat('column'), type='quantitative'),\n",
    "#     alt.Y(alt.repeat('row'), type='quantitative')\n",
    "# ).properties(width=150, height=150).repeat(\n",
    "#     row=numeric_features,\n",
    "#     column=numeric_features\n",
    "# )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculate correlations\n",
    "corr = X_train[numeric_features].corr()\n",
    "corr_df = corr.reset_index().melt(id_vars='index', var_name='yaxis', value_name='correlation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-894b3c5dc92c27ae8615b7830d08f111.json"
       },
       "encoding": {
        "color": {
         "field": "correlation",
         "type": "quantitative"
        },
        "x": {
         "axis": {
          "title": null
         },
         "field": "index",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "title": null
         },
         "field": "yaxis",
         "type": "nominal"
        }
       },
       "mark": "rect",
       "title": "Correlations between numeric features"
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr = X_train[numeric_features].corr()\n",
    "corr_df = corr.reset_index().melt(id_vars='index', var_name='yaxis', value_name='correlation')\n",
    "alt.Chart(corr_df).mark_rect().encode(\n",
    "    alt.X('index:N', axis=alt.Axis(title=None)),\n",
    "    alt.Y('yaxis:N', axis=alt.Axis(title=None)),\n",
    "    alt.Color('correlation:Q')\n",
    ").properties(title='Correlations between numeric features')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Possible relationships between categorical and numeric predictors  \n",
    "\n",
    "If one class of the categorical feature amount is consistently higher than some other class of the categorical feature. For example, if women tend to have higher credit limit than men. Because the distribution of bill statement amounts and payment amounts are similar among different months, I selected one of them randomly to draw the graphs.  \n",
    "\n",
    "For example, we can see that education probably influences the credit limit, because 3-high school boxplot is lower than 4-Others boxplot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "def box_plot(data, cat_var, cont_var):\n",
    "    return alt.Chart(data).mark_boxplot().encode(\n",
    "    alt.X(cat_var, type='nominal'),\n",
    "    alt.Y(cont_var, type='quantitative')\n",
    ").properties(width=150, height=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "cont_list = ['PAY_AMT3', 'BILL_AMT3', 'LIMIT_BAL']\n",
    "cat_list = ['SEX', 'EDUCATION', 'MARRIAGE']\n",
    "\n",
    "row = alt.vconcat()\n",
    "for cont_var in cont_list:\n",
    "    col = alt.hconcat()\n",
    "    for cat_var in cat_list:\n",
    "        col |= box_plot(X_train, cat_var, cont_var)\n",
    "    row &= col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "title": {
         "anchor": "middle"
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-ffdf232926808bd25ec5ef105eb570ec.json"
       },
       "title": "Boxplot",
       "vconcat": [
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "SEX",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "EDUCATION",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "MARRIAGE",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "SEX",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "EDUCATION",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "MARRIAGE",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "SEX",
             "type": "nominal"
            },
            "y": {
             "field": "LIMIT_BAL",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "EDUCATION",
             "type": "nominal"
            },
            "y": {
             "field": "LIMIT_BAL",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "MARRIAGE",
             "type": "nominal"
            },
            "y": {
             "field": "LIMIT_BAL",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row.properties(title=\"Boxplot distribution\").configure_title(anchor='middle')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Possible relationships between predictors and the response variable  \n",
    "\n",
    "Our research question is what features predict default payment? Let's see if we can find any strong associations between predictors and the response variable visually."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Numeric predictors to the response variable  \n",
    "According to the boxplots below, credit limit for 0-Non default is slightly higher than 1-default. It's probably a good predictor for default payment. It also makes sense because for people able to get higher credit limit, they must have good credit history, so that banks are willing to allow them higher credit amount. Also, payment amounts seem are also good predictors to me, because non default tends to have higher and more higher outliers than default payment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = X_train.copy()\n",
    "df_train['DEFAULT_NEXT_MONTH'] = y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = alt.vconcat()\n",
    "col = alt.hconcat()\n",
    "ncol = 4  #how many to display in each row\n",
    "\n",
    "for i in range(len(numeric_features)):\n",
    "    col |= box_plot(df_train, 'DEFAULT_NEXT_MONTH', numeric_features[i])\n",
    "    if (i+1)%ncol==0:\n",
    "        row &= col\n",
    "        col = alt.hconcat()\n",
    "num_res_chart = row & col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-aeb87ad608560432c0dd4b58db70979f.json"
       },
       "vconcat": [
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "LIMIT_BAL",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "AGE",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT1",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT2",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT4",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT5",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "BILL_AMT6",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT1",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT2",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT3",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT4",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT5",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          },
          {
           "encoding": {
            "x": {
             "field": "DEFAULT_NEXT_MONTH",
             "type": "nominal"
            },
            "y": {
             "field": "PAY_AMT6",
             "type": "quantitative"
            }
           },
           "height": 200,
           "mark": "boxplot",
           "width": 150
          }
         ]
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_res_chart"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Categorical predictors to the response variable  \n",
    "According to the below graphs, gender, education and marriage's proportions of different classes are pretty similar between non default and default payment. For payment status, 2-payment delay for two months accounted for more in 1-default payment than in 0-non default. Therefore, they are probably good predictor variables to predict default payment as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def stack_bar(data, cat_var, res_var_cat):\n",
    "    return alt.Chart(data).mark_bar(opacity=0.7).encode(\n",
    "    alt.Y(res_var_cat, type='nominal', title='Default'),\n",
    "    alt.X('count()', axis=alt.Axis(grid=False), stack='normalize', title=f'Count of {cat_var} - Normalized'),\n",
    "    alt.Color(cat_var, type='nominal')\n",
    ").properties(width=400)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "row = alt.vconcat()\n",
    "col = alt.hconcat()\n",
    "ncol = 2  #how many to display in each row\n",
    "\n",
    "for i in range(len(categorical_features)):\n",
    "    col |= stack_bar(df_train, categorical_features[i], 'DEFAULT_NEXT_MONTH')\n",
    "    if (i+1)%ncol==0:\n",
    "        row &= col\n",
    "        col = alt.hconcat()  \n",
    "cat_res_chart = row & col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-aeb87ad608560432c0dd4b58db70979f.json"
       },
       "title": "Normalized stacked bar",
       "vconcat": [
        {
         "hconcat": [
          {
           "encoding": {
            "color": {
             "field": "SEX",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of SEX - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          },
          {
           "encoding": {
            "color": {
             "field": "EDUCATION",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of EDUCATION - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "color": {
             "field": "MARRIAGE",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of MARRIAGE - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          },
          {
           "encoding": {
            "color": {
             "field": "PAY_1",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_1 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "color": {
             "field": "PAY_2",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_2 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          },
          {
           "encoding": {
            "color": {
             "field": "PAY_3",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_3 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "color": {
             "field": "PAY_4",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_4 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          },
          {
           "encoding": {
            "color": {
             "field": "PAY_5",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_5 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          }
         ]
        },
        {
         "hconcat": [
          {
           "encoding": {
            "color": {
             "field": "PAY_6",
             "type": "nominal"
            },
            "x": {
             "aggregate": "count",
             "axis": {
              "grid": false
             },
             "stack": "normalize",
             "title": "Count of PAY_6 - Normalized",
             "type": "quantitative"
            },
            "y": {
             "field": "DEFAULT_NEXT_MONTH",
             "title": "Default",
             "type": "nominal"
            }
           },
           "mark": {
            "opacity": 0.7,
            "type": "bar"
           },
           "width": 400
          }
         ]
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_res_chart.properties(title='Normalized stacked bar')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "According to the above exploratory data analysis, I think the data entries are reasonable and valid. Data size is good enough for fitting models.\n",
    "\n",
    "We see collinearity between `BILL_AMT1`, `BILL_AMT2`, ..., `BILL_AMT6` variables. We should consider that when fitting the prediction models. According to the plots, `LIMIT_BAL`, `PAY_AMT1`, `PAY_AMT2`, ..., `PAY_AMT6`, `PAY_1`, `PAY_2`, ..., `PAY_6` are probably good predictor variables to predict if customers will fail to pay next month or not.\n",
    "\n",
    "Overall, this would be a good dataset to fit the model and to predict the default payment. And \"What features strongly predict default payment?\" is a good and interesting research question for this dataset."
   ]
  },
  {
   "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": 4
}