[Kaggle] EDA on application dataset.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Home Credit Default Risk\n",
    "https://www.kaggle.com/c/home-credit-default-risk#description"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 배경 이해하기:\n",
    "Many people struggle to get loans due to insufficient or non-existent credit histories. And, unfortunately, this population is often taken advantage of by untrustworthy lenders.\n",
    "\n",
    "Home Credit Group\n",
    "\n",
    "Home Credit strives to broaden financial inclusion for the unbanked population by providing a positive and safe borrowing experience. In order to make sure this underserved population has a positive loan experience, Home Credit makes use of a variety of alternative data--including telco and transactional information--to predict their clients' repayment abilities.\n",
    "\n",
    "While Home Credit is currently using various statistical and machine learning methods to make these predictions, they're challenging Kagglers to help them unlock the full potential of their data. Doing so will ensure that clients capable of repayment are not rejected and that loans are given with a principal, maturity, and repayment calendar that will empower their clients to be successful."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- 신용 기록이 없거나 충분하지 않는 사람들은 대출을 받는 데 어려움이 있음\n",
    "- 대출을 받는다 해도 신뢰할 수 없는 채권자(untrustworthy lenders, 아마 고리대금업자 등?)에게 돈을 빌리기 쉬움\n",
    "- HomeCredit은 금융 서비스를 충분히 받지 못하고 있는 이런 사람들에게 안전하고 긍정적인 대출 경험을 제공하고자, 다양한 대체 데이터(통신, 거래 정보 등)를 활용해 고객의 상환 능력을 예측하고 있음\n",
    "- 기존에도 다양한 통계 및 머신러닝 기법을 사용하고 있지만 본 대회를 통해 그들이 가진 데이터의 잠재력을 완전히 끌어내고자 함\n",
    "- **목표**: 상환할 수 있는 고객을 거절(reject)하지 않을 것 & 고객이 잘 상환할 수 있는 적절한 원금(principal), 만기(maturity) 및 상환(repayment) 일정 등을 정해 대출을 해주는 것 \n",
    "\n",
    "--> 이러한 배경으로 미루어보아 1) 신용이 좋지 않은 사람들의 데이터일 확률이 높음. 2) 현재 신용도는 낮더라도 대출 상환을 잘 할 사람을 어떻게 예측할 것인가? (성실성 등을 계량화? 과거에 비해 신용도가 높아지고 있는 사람들 파악하는 것이 중요해 보임)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 대회 평가 방법\n",
    "- Submissions are evaluated on area under the ROC curve between the predicted probability and the observed target.\n",
    "\n",
    "#### 일정\n",
    "- August 22, 2018 - Entry deadline. You must accept the competition rules before this date in order to compete.\n",
    "- August 22, 2018 - Team Merger deadline. This is the last day participants may join or merge teams.\n",
    "- August 29, 2018 - Final submission deadline."
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data Structure\n",
    "![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.stats as stats\n",
    "import sklearn.linear_model as linear_model\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import KFold\n",
    "from IPython.display import HTML, display\n",
    "from sklearn.manifold import TSNE\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import StratifiedShuffleSplit\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## EDA on application_train.csv & application_test.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "ename": "FileNotFoundError",
     "evalue": "File b'application_train.csv' does not exist",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-4-88f40cba62f8>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mdf1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'application_train.csv'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindex_col\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m\"SK_ID_CURR\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mdf2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'application_test.csv'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindex_col\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m\"SK_ID_CURR\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36mparser_f\u001b[1;34m(filepath_or_buffer, sep, delimiter, header, names, index_col, usecols, squeeze, prefix, mangle_dupe_cols, dtype, engine, converters, true_values, false_values, skipinitialspace, skiprows, nrows, na_values, keep_default_na, na_filter, verbose, skip_blank_lines, parse_dates, infer_datetime_format, keep_date_col, date_parser, dayfirst, iterator, chunksize, compression, thousands, decimal, lineterminator, quotechar, quoting, escapechar, comment, encoding, dialect, tupleize_cols, error_bad_lines, warn_bad_lines, skipfooter, skip_footer, doublequote, delim_whitespace, as_recarray, compact_ints, use_unsigned, low_memory, buffer_lines, memory_map, float_precision)\u001b[0m\n\u001b[0;32m    707\u001b[0m                     skip_blank_lines=skip_blank_lines)\n\u001b[0;32m    708\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 709\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0m_read\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfilepath_or_buffer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    710\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    711\u001b[0m     \u001b[0mparser_f\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__name__\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mname\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36m_read\u001b[1;34m(filepath_or_buffer, kwds)\u001b[0m\n\u001b[0;32m    447\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    448\u001b[0m     \u001b[1;31m# Create the parser.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 449\u001b[1;33m     \u001b[0mparser\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mTextFileReader\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfilepath_or_buffer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    450\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    451\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mchunksize\u001b[0m \u001b[1;32mor\u001b[0m \u001b[0miterator\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, f, engine, **kwds)\u001b[0m\n\u001b[0;32m    816\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0moptions\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'has_index_names'\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mkwds\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'has_index_names'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    817\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 818\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_make_engine\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    819\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    820\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mclose\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36m_make_engine\u001b[1;34m(self, engine)\u001b[0m\n\u001b[0;32m   1047\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_make_engine\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mengine\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'c'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1048\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mengine\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'c'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1049\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_engine\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mCParserWrapper\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mf\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0moptions\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1050\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1051\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mengine\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m'python'\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mC:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\io\\parsers.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, src, **kwds)\u001b[0m\n\u001b[0;32m   1693\u001b[0m         \u001b[0mkwds\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'allow_leading_cols'\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mindex_col\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1694\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1695\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_reader\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mparsers\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTextReader\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msrc\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1696\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1697\u001b[0m         \u001b[1;31m# XXX\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mpandas/_libs/parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader.__cinit__\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;32mpandas/_libs/parsers.pyx\u001b[0m in \u001b[0;36mpandas._libs.parsers.TextReader._setup_parser_source\u001b[1;34m()\u001b[0m\n",
      "\u001b[1;31mFileNotFoundError\u001b[0m: File b'application_train.csv' does not exist"
     ]
    }
   ],
   "source": [
    "df1 = pd.read_csv('application_train.csv', index_col = \"SK_ID_CURR\")\n",
    "df2 = pd.read_csv('application_test.csv', index_col = \"SK_ID_CURR\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'df1' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-6-2952d8bf897d>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mdf1\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m\"DataType\"\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m\"train\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mdf2\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m\"DataType\"\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m=\u001b[0m \u001b[1;34m\"test\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mdf\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mdf1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdf2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mignore_index\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mFalse\u001b[0m \u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mdf_original\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdf\u001b[0m \u001b[1;31m# 원본 데이터 df_original에 저장해둠\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'df1' is not defined"
     ]
    }
   ],
   "source": [
    "df1[\"DataType\"] = \"train\"\n",
    "df2[\"DataType\"]= \"test\"\n",
    "\n",
    "df = pd.concat([df1, df2], ignore_index = False )\n",
    "df_original = df # 원본 데이터 df_original에 저장해둠"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### target variable 분포 확인하기: \n",
    "- 1 - client with payment difficulties: he/she had late payment more than X days on at least one of the first Y installments of the loan in our sample,\n",
    "- 0 - all other cases\n",
    "- highly unbalance 데이터임을 알 수 있음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df1[\"TARGET\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp = df1[\"TARGET\"].value_counts()\n",
    "df = pd.DataFrame({'labels': temp.index,\n",
    "                   'values': temp.values\n",
    "                  })\n",
    "plt.figure(figsize = (6,6))\n",
    "plt.title('Application loans repayed - train dataset: \\n 0 = Others & 1= Having difficulties with repayment')\n",
    "sns.set_color_codes(\"pastel\")\n",
    "sns.barplot(x = 'labels', y=\"values\", data=df)\n",
    "locs, labels = plt.xticks()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(df1.shape)\n",
    "print(df2.shape)\n",
    "print(df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Kaggle Kernel에서 참고한 변수 유형별 시각화 함수\n",
    "- reference : https://www.kaggle.com/gpreda/home-credit-default-risk-extensive-eda\n",
    "- reference 참고하여 barplot의 bar 보여주는 순서 설정 추가함\n",
    "- **plot_stats(feature)**: \n",
    "    - 명목형 변수 대상\n",
    "    - 왼쪽 = 전체 데이터에서 category 별 빈도(n), 오른쪽 = 상환되지 않은 대출 중 category 별 비중(%) 표시\n",
    "- **plot_distribution(feature)**:\n",
    "    - 연속형 변수 대상\n",
    "- **def plot_distribution_hue(var)**:\n",
    "    - 연속형 변수 대상. target = 1, 0 그룹 나누어서 표시"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_stats(feature,label_rotation=False,horizontal_layout=True):\n",
    "    temp = df1[feature].value_counts()\n",
    "    df_temp = pd.DataFrame({feature: temp.index,'Number of contracts': temp.values})\n",
    "   # Calculate the percentage of target=1 per category value\n",
    "    cat_perc = df1[[feature, 'TARGET']].groupby([feature],as_index=False).mean()\n",
    "    cat_perc.sort_values(by='TARGET', ascending=False, inplace=True)\n",
    "   # --> 비율 큰 순서로 막대 순서를 설정해줌. 오히려 헷갈림....???\n",
    "    \n",
    "    if(horizontal_layout):\n",
    "        fig, (ax1, ax2) = plt.subplots(ncols=2, figsize=(12,6))\n",
    "    else:\n",
    "        fig, (ax1, ax2) = plt.subplots(nrows=2, figsize=(12,14))\n",
    "    sns.set_color_codes(\"pastel\")\n",
    "    \n",
    "    s = sns.barplot(ax=ax1, x = feature, y=\"Number of contracts\", data=df_temp)\n",
    "    if(label_rotation):\n",
    "        s.set_xticklabels(s.get_xticklabels(),rotation=90)\n",
    "\n",
    "    s = sns.barplot(ax=ax2, x = feature, y='TARGET', order=cat_perc[feature], data=cat_perc)\n",
    "    s = sns.barplot(ax=ax2, x = feature, y='TARGET',data=cat_perc)\n",
    "    if(label_rotation):\n",
    "        s.set_xticklabels(s.get_xticklabels(),rotation=90)\n",
    "        \n",
    "    plt.ylabel('Percent of target with value 1 [%]', fontsize=10)\n",
    "    plt.tick_params(axis='both', which='major', labelsize=10)\n",
    "\n",
    "    plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_distribution(feature,color):\n",
    "    plt.figure(figsize=(10,6))\n",
    "    plt.title(\"Distribution of %s\" % feature)\n",
    "    sns.distplot(df1[feature].dropna(),color=color, kde=True,bins=100)\n",
    "    plt.show()   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_distribution_hue(var):\n",
    "    sns.kdeplot(df1.loc[df1[\"TARGET\"] == 1, var], label = \"Target = 1\", shade = True)\n",
    "    sns.kdeplot(df1.loc[df1[\"TARGET\"] == 0, var], label = \"Target = 0\", shade = True)\n",
    "    plt.xlabel(var); plt.ylabel('Density'); plt.title(\"Distribution of \"+var+\" by Target\")\n",
    "    plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 목차\n",
    "1. 신청한 대출 특성\n",
    "2. 인구통계적 특성\n",
    "3. 직업 및 소득 관련\n",
    "4. 기본 자산 현황\n",
    "5. 거주지 관련\n",
    "6. 외부신용평가 자료"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-----------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 신청한 대출 특성\n",
    "    1.1. 대출 유형(Contract Type)\n",
    "    1.2. 대출 규모(\n",
    "    1.3. 대출 당시 제출 서류\n",
    "    1.4. 대출 당시 동반자(NAME_TYPE_SUITE)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1. 대출 유형(Contract Type)\n",
    "- 전반적으로 Revolving Loan보단 Cash Loan이 더 많음\n",
    "- 상환문제 발생 비율(Target == 1)은 Cash Loan을 받은 경우가 더 높음\n",
    "- encoding 해야 함"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('NAME_CONTRACT_TYPE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"NAME_CONTRACT_TYPE\"], df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"NAME_CONTRACT_TYPE\"],df1[\"TARGET\"], margins=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.2. 대출 규모"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('AMT_CREDIT')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('AMT_ANNUITY')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.4. 대출 당시 동반자\n",
    "- NAME_TYPE_SUITE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"NAME_TYPE_SUITE\"], df1[\"TARGET\"], normalize='index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 인구통계적 특성\n",
    "    2.1. 성별 (CODE_GENDER)\n",
    "    2.2. 나이 (DAYS_BIRTH)\n",
    "    2.3. 가족 구성 (NAME_FAMILY_STATUS, CNT_CHILDREN)\n",
    "    2.4. 교육 수준 (NAME_EDUCATION_TYPE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1. 성별\n",
    "- CODE_GENDER\n",
    "- 대출은 여자가 더 많이 받음. 그런데,미상환하는 비율은 남자 그룹에서 더 높게 나타남.(막대 색상에 주의)\n",
    "- XNA 4명 모두 TARGET == 0, Revolving Laoan 받은 그룹에 속함\n",
    "- test set에는 \"CODE_GENDER\" = \"XNA\"인 경우 없음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('CODE_GENDER')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# pd.crosstab(df1[\"CODE_GENDER\"],df1[\"TARGET\"], margins=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"CODE_GENDER\"], df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1.loc[df1[\"CODE_GENDER\"] == \"XNA\"] #XNA 4명 모두 TARGET == 0, Revolving Laoan 받은 그룹에 속함"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df2[\"CODE_GENDER\"].value_counts() #test set에는 \"CODE_GENDER\" = \"XNA\"인 경우 없음"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2. 나이\n",
    "- DAYS_BIRTH\n",
    "- 연령별 그룹으로 나누기 --> \"AgeGroup\" 변수 생성\n",
    "- 미상환 비중이 높은 그룹은 20대."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plot_distribution('DAYS_BIRTH','Orange')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_distribution_hue('DAYS_BIRTH')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1[\"Age\"]=-df1[\"DAYS_BIRTH\"]/365\n",
    "plot_distribution_hue('Age')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1[\"Age\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def age_grouping(x):\n",
    "    if x < 30:\n",
    "        return 20\n",
    "    elif x < 40:\n",
    "        return 30\n",
    "    elif x < 50:\n",
    "        return 40\n",
    "    elif x < 60:\n",
    "         return 50\n",
    "    elif x < 70:\n",
    "         return 60\n",
    "    return 70\n",
    "df1['AgeGroup'] = df1['Age'].apply(age_grouping)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('AgeGroup')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"AgeGroup\"], df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"TARGET\"],df1[\"AgeGroup\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3. 가족 구성\n",
    "- 결혼 여부(\"NAME_FAMILY_STATUS\"): \n",
    "    - 결혼 한 사람이 대출을 많이 신청함, 비교적 잘 갚음\n",
    "    - Civil marriage의 정확한 뜻 파악해야 함 왜 이 그룹에서 미상환 고객의 비율이 높게 나타나는지?  \n",
    "- 가족의 수(\"CNT_FAM_MEMBERS\"):\n",
    "    - 1인 가구보다 2인 가구의 대출 수가 더 많음\n",
    "- 자녀의 수(\"CNT_CHILDREN\"): \n",
    "    - 자녀가 없는 사람들이 대출을 신청하는 경우가 매우 많음\n",
    "    - 자녀가 6명 이하인 경우 상환율은 매우 높음(71%~92%)\n",
    "    - 자녀가 7명 이상인 그룹 중 내에서도 잘 갚는 경우, 못갚는 경우 혼재함. 다른 변수들(자산현황, 소득 등)과 같이 추가적으로 분석해야 함"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fam_var = ['NAME_FAMILY_STATUS', 'CNT_FAM_MEMBERS','CNT_CHILDREN']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1[fam_var].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('NAME_FAMILY_STATUS',True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('CNT_FAM_MEMBERS')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('CNT_CHILDREN')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"CNT_FAM_MEMBERS\"], df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"CNT_CHILDREN\"], df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('CNT_CHILDREN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.4. 교육 수준 \n",
    "- NAME_EDUCATION_TYPE\n",
    "- 전반적으로 중등교육(Secondary / secondary special) 정도의 수준을 받은 사람들이 대출을 많이 신청함\n",
    "- 미상환률이 높은 그룹: Incomplete higher, Lower secondary, Secondary / secondary special\n",
    "- 상환율이 높은 그룹: Higher education, Academic degree\n",
    "- Secondary / secondary special 그룹의 직업, 성별, 소득 등의 분포를 볼 필요가 있음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('NAME_EDUCATION_TYPE',True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"NAME_EDUCATION_TYPE\"], df1[\"TARGET\"], normalize='index')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df1[\"CNT_CHILDREN\"].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "--------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 직업 및 소득 관련\n",
    "    3.1. 소득 유형(NAME_INCOME_TYPE)\n",
    "    3.2. 소득 수준(AMT_INCOME_TOTAL)\n",
    "    3.3. 직업 유형(OCCUPATION_TYPE)\n",
    "    3.4. 조직 유형(ORGANIZATION_TYPE)\n",
    "    3.5. 고용 기간(DAYS_EMPLOYED)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1. 소득 유형\n",
    "- NAME_INCOME_TYPE\n",
    "- 미상환비율이 40% 이상인 그룹 위험군(Maternity leave,Unemployed) 으로 분리? \n",
    "- 이 그룹은 고용상태가 안정적이지 않음에도 어떻게 대출을 받았을까? 자산이 많은 그룹인가? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('NAME_INCOME_TYPE',True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"NAME_INCOME_TYPE\"], df1[\"TARGET\"], normalize='index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2. 소득 수준\n",
    "- AMT_INCOME_TOTAL\n",
    "- 한쪽으로 매우 치우침.. 로그 변환"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1[\"AMT_INCOME_TOTAL\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_distribution('AMT_INCOME_TOTAL','blue')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1['ln_AMT_INCOME_TOTAL'] = np.log(df1['AMT_INCOME_TOTAL'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('ln_AMT_INCOME_TOTAL')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3. 직업 유형\n",
    "- OCCUPATION_TYPE\n",
    "- Low-skill Laborers가 전체 대출자에서 차지하는 비중은 많지 않으나, 미상환율은 가장 높은 그룹임"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('OCCUPATION_TYPE',True, False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.4. 조직 유형\n",
    "- ORGANIZATION_TYPE \n",
    "- type3 의 미상환율이 특히 높게 나타남\n",
    "- 좀 더 그룹지어서 봐야함.."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('ORGANIZATION_TYPE',True, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab(df1[\"ORGANIZATION_TYPE\"], df1[\"TARGET\"], normalize='index')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.5. 고용 기간\n",
    "- DAYS_EMPLOYED\n",
    "- \"365243\"은 은퇴한 그룹으로 예상\n",
    "- +로 바꾸고 365로 나누어서 년수로 계산(\"Years_EMPLOYED\")\n",
    "- 미상환 그룹 중 근속연수가 낮은(7년 이하 정도?) 비율이 높게 나타남"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1[\"DAYS_EMPLOYED\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('DAYS_EMPLOYED')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1[\"Years_EMPLOYED\"]= (-df1['DAYS_EMPLOYED'])/365\n",
    "df1[\"Years_EMPLOYED\"].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('Years_EMPLOYED')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_retired = df1.loc[df1[\"Years_EMPLOYED\"]<0]\n",
    "test_retired[\"Years_EMPLOYED\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_working= df1.loc[df1[\"Years_EMPLOYED\"]>=0]\n",
    "sns.kdeplot(test_working.loc[test_working[\"TARGET\"] == 1, \"Years_EMPLOYED\"], label = \"TARGET = 1\", shade = True)\n",
    "sns.kdeplot(test_working.loc[test_working[\"TARGET\"] == 0, \"Years_EMPLOYED\"], label = \"TARGET = 0\", shade = True)\n",
    "plt.xlabel('Years_EMPLOYED'); plt.ylabel('Density'); plt.title(\"Density Plot of Years_Employed of working group by Target\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 기본 자산 현황\n",
    "    4.1. 차량소유 여부(FLAG_OWN_CAR)\n",
    "    4.2. 부동산 소유 여부(FLAG_OWN_REALTY)\n",
    "    4.3. 자동차 연식(OWN_CAR_AGE)\n",
    "    4.4. 연락망(FLAG_MOBIL, FLAG_EMP_PHONE, FLAG_WORK_PHONE, FLAG_CONT_MOBILE, FLAG_PHONE, FLAG_EMAIL) - 이것도 여기에 넣는 것이 맞는 건지 모르겠습니다;\n",
    "- 자산 상태(차량, 부동산) 확인\n",
    "- y = 1, n = 0\n",
    "- 자동차는 안가진 경우가 더 많고, 부동산은 가진 경우가 더 많음. 특이하다고 생각됩니다. 보통은 자동차가 더 많이 보급되어 있고, 부동산을 가진 경우는 더 드물지 않나요? \n",
    "- (Y,N 주의해서 읽어야함. 양쪽 그래프 레이블에 따라 동일한 컬러를 넣고 싶은데 ㅠㅠ 어떻게 하는 건지 모르겠습니다)\n",
    "- 자산 유무에 따른 상환 문제 발생 여부가 거의 비슷하긴 함\n",
    "- 다만, 자동차(또는 부동산)가 없는 그룹에서 미상환 비율이 더 높게 나타남\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1. 차량 소유 여부"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_OWN_CAR')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pd.crosstab(df1[\"FLAG_OWN_CAR\"],df1[\"TARGET\"], margins=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab( df1[\"FLAG_OWN_CAR\"],df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2. 부동산 소유 여부"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_OWN_REALTY')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# pd.crosstab(df1[\"FLAG_OWN_REALTY\"], df1[\"TARGET\"], margins=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.crosstab( df1[\"FLAG_OWN_REALTY\"],df1[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 자동차 &부동산, 둘 다 가진 경우, 하나라도 있는 경우, 둘 다 없는 경우, 따라 상환율이 다르게 나타나지 않을까요?\n",
    "- 실제 둘 다 없는 그룹 = 미상환 발생 비율 8.99%로 가장 높게 나타남\n",
    "- 차가 없고 집이 있는 경우 = 미상환 발생 비율 8.28%\n",
    "- 차가 있고, 집이 없는 경우 = 미상환 발생 비율 7.04%\n",
    "- 둘 다 있는 그룹 = 미상환 발생 비율 7.33%로 나타남\n",
    "- 4가지 그룹으로 인코딩하면 어떨까요?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_CR = pd.DataFrame(df1[\"FLAG_OWN_CAR\"])\n",
    "test_CR[\"FLAG_OWN_REALTY\"] = df1[\"FLAG_OWN_REALTY\"]\n",
    "test_CR[\"CARorRealty\"] = test_CR[\"FLAG_OWN_CAR\"] + test_CR[\"FLAG_OWN_REALTY\"]\n",
    "test_CR[\"TARGET\"] = df1[\"TARGET\"]\n",
    "test_CR= pd.DataFrame(test_CR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_CR.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "pd.crosstab( test_CR[\"CARorRealty\"],test_CR[\"TARGET\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pd.crosstab(test[\"TARGET\"], test[\"CARorRealty\"]).apply(lambda r: r/r.sum(), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.3. 자동차 연식\n",
    "- 10년 이하 된 연식의 자동차를 가진 그룹은 상환을 잘 함\n",
    "- 10년 이상 된 연식의 자동차를 가진 그룹은 상환을 잘 못함\n",
    "- 45년 이상 된 연식의 자동차를 가진 그룹과 그 이하 자동차를 가진 그룹으로 분류?\n",
    "- ;; 갑자기 그래프 해석하는 방법이 헷갈리네요 @_@,,,,  상환을 잘 하는 그룹 중, 새로운 자동차를 보유한 사람들이 많다고 해야 하는 건가요? 아님.. 저렇게 해석해도 큰 무리가 없으려나요..;"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('OWN_CAR_AGE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.4. 각종 연락망 보유 여부\n",
    "- FLAG_MOBIL, FLAG_EMP_PHONE, FLAG_WORK_PHONE, FLAG_CONT_MOBILE, FLAG_PHONE, FLAG_EMAIL\n",
    "- 크게 영향을 미치지 않는 것으로 판단\n",
    "- 연락망 수를 합쳐서 보아도 차이 없어보임"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_MOBIL')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_EMP_PHONE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_WORK_PHONE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_CONT_MOBILE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('FLAG_PHONE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df1['n_of_contacts'] = df1['FLAG_MOBIL']+df1['FLAG_EMP_PHONE']+df1['FLAG_WORK_PHONE']+df1['FLAG_CONT_MOBILE']+df1['FLAG_PHONE']+df1['FLAG_EMAIL']\n",
    "df1['n_of_contacts'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_stats('n_of_contacts')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. 거주지 관련"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    5.1. 집 특성 변수\n",
    "    - APARTMENTS, BASEMENTAREA, COMMONAREA, ELEVATORS, ENTRANCES, FLOORSMAX, FLOORSMIN, LANDAREA, LIVINGAPARTMENTS, LIVINGAREA, NONLIVINGAPARTMENTS , NONLIVINGAREA, YEARS_BEGINEXPLUATATION, YEARS_BUILD\n",
    "       - AVG, MEDI, MODE 세 가지 값 있음\n",
    "    - TOTALAREA_MODE: AVG, MEDI 없음. 연속형\n",
    "    - FONDKAPREMONT_MODE: 범주형\n",
    "    - HOUSETYPE_MODE: 범주형\n",
    "    - WALLSMATERIAL_MODE: 범주형\n",
    "\n",
    "    5.2. 거주 지역 특성\n",
    "    - REGION_POPULATION_RELATIVE\n",
    "    - REGION_RATING_CLIENT\n",
    "    - REGION_RATING_CLIENT_W_CITY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df[\"FLOORSMAX_AVG\"].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df[\"FLOORSMAX_MODE\"].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df[\"FLOORSMAX_MEDI\"].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df1[\"Landarea_skew\"] = df1[\"LANDAREA_MODE\"] - df1[\"LANDAREA_AVG\"]\n",
    "# df1[\"APARTMENTS_skew\"] = df1[\"APARTMENTS_MODE\"] - df1[\"APARTMENTS_AVG\"]\n",
    "# df1[\"BASEMENTAREA_skew\"] = df1[\"BASEMENTAREA_MODE\"] - df1[\"BASEMENTAREA_AVG\"]\n",
    "# df1[\"COMMONAREA_skew\"] = df1[\"COMMONAREA_MODE\"] - df1[\"COMMONAREA_AVG\"]\n",
    "# df1[\"ELEVATORS_skew\"] = df1[\"ELEVATORS_MODE\"] - df1[\"ELEVATORS_AVG\"]\n",
    "# df1[\"ENTRANCES_skew\"] = df1[\"ENTRANCES_MODE\"] - df1[\"ENTRANCES_AVG\"]\n",
    "# df1[\"FLOORSMAX_skew\"] = df1[\"FLOORSMAX_MODE\"] - df1[\"FLOORSMAX_AVG\"]\n",
    "# df1[\"FLOORSMIN_skew\"] = df1[\"FLOORSMIN_MODE\"] - df1[\"FLOORSMIN_AVG\"]\n",
    "# df1[\"LANDAREA_skew\"] = df1[\"LANDAREA_MODE\"] - df1[\"LANDAREA_AVG\"]\n",
    "# df1[\"LIVINGAPARTMENTS_skew\"] = df1[\"LIVINGAPARTMENTS_MODE\"] - df1[\"LIVINGAPARTMENTS_AVG\"]\n",
    "# df1[\"LIVINGAREA_skew\"] = df1[\"LIVINGAREA_MODE\"] - df1[\"LIVINGAREA_AVG\"]\n",
    "# df1[\"NONLIVINGAPARTMENTS_skew\"] = df1[\"NONLIVINGAPARTMENTS_MODE\"] - df1[\"NONLIVINGAPARTMENTS_AVG\"]\n",
    "# df1[\"NONLIVINGAREA_skew\"] = df1[\"NONLIVINGAREA_MODE\"] - df1[\"NONLIVINGAREA_AVG\"]\n",
    "# df1[\"YEARS_BEGINEXPLUATATION_skew\"] = df1[\"YEARS_BEGINEXPLUATATION_MODE\"] - df1[\"YEARS_BEGINEXPLUATATION_AVG\"]\n",
    "# df1[\"YEARS_BUILD_skew\"] = df1[\"YEARS_BUILD_MODE\"] - df1[\"YEARS_BUILD_AVG\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# df1.loc[:,[\"Landarea_skew\", \"APARTMENTS_skew\",\"BASEMENTAREA_skew\",\"COMMONAREA_skew\",\"ELEVATORS_skew\",\"ENTRANCES_skew\",\"FLOORSMAX_skew\",\"FLOORSMIN_skew\",\"LANDAREA_skew\",\"LIVINGAPARTMENTS_skew\",\"LIVINGAREA_skew\",\"NONLIVINGAPARTMENTS_skew\",\"NONLIVINGAREA_skew\",\"YEARS_BEGINEXPLUATATION_skew\",\"YEARS_BUILD_skew\"]].hist(bins = 30, figsize = (30,25))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#df1_house = df1.loc[:,[\"LANDAREA_AVG\", \"APARTMENTS_AVG\",\"BASEMENTAREA_AVG\",\"COMMONAREA_AVG\",\"ELEVATORS_AVG\",\"ENTRANCES_AVG\",\"FLOORSMAX_AVG\",\"FLOORSMIN_AVG\",\"LANDAREA_AVG\",\"LIVINGAPARTMENTS_AVG\",\"LIVINGAREA_AVG\",\"NONLIVINGAPARTMENTS_AVG\",\"NONLIVINGAREA_AVG\",\"YEARS_BEGINEXPLUATATION_AVG\",\"YEARS_BUILD_AVG\",\"TARGET\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df1_house.hist(bins = 30, figsize = (30,25))\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df1_house.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. 외부 신용평가 자료\n",
    "    6.1. 외부 데이터의 점수(EXT_SOURCE_1 ~ 3)\n",
    "    6.2. CreditBureau 과거 기록\n",
    "           - 대출 신청 이전에 Credit Bureau에 접수된 해당 고객에 대한 enquiries (?)\n",
    "           - AMT_REQ_CREDIT_BUREAU_HOUR, AMT_REQ_CREDIT_BUREAU_DAY, AMT_REQ_CREDIT_BUREAU_WEEK, AMT_REQ_CREDIT_BUREAU_MON, AMT_REQ_CREDIT_BUREAU_QRT, AMT_REQ_CREDIT_BUREAU_YEAR\n",
    "\n",
    "    6.3.주변인신용평가 자료\n",
    "          - OBS_30_CNT_SOCIAL_CIRCLE, DEF_30_CNT_SOCIAL_CIRCLE, OBS_60_CNT_SOCIAL_CIRCLE, DEF_60_CNT_SOCIAL_CIRCLE, DAYS_LAST_PHONE_CHANGE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.1. 외부 데이터의 점수(EXT_SOURCE_1 ~ 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('EXT_SOURCE_1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_distribution_hue('EXT_SOURCE_2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_distribution_hue('EXT_SOURCE_3')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5.2. CreditBureau 과거 기록"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plot_stats('AMT_REQ_CREDIT_BUREAU_HOUR', True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_stats('AMT_REQ_CREDIT_BUREAU_DAY', True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_stats('AMT_REQ_CREDIT_BUREAU_WEEK', True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_stats('AMT_REQ_CREDIT_BUREAU_MON', True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_stats('AMT_REQ_CREDIT_BUREAU_YEAR', True, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#AMT_REQ = [\"AMT_REQ_CREDIT_BUREAU_DAY\", \"AMT_REQ_CREDIT_BUREAU_HOUR\", \"AMT_REQ_CREDIT_BUREAU_MON\", \"AMT_REQ_CREDIT_BUREAU_QRT\",\"AMT_REQ_CREDIT_BUREAU_WEEK\",\"AMT_REQ_CREDIT_BUREAU_YEAR\"]"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}