Day_010_HW.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Import 需要的套件\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 設定 data_path\n",
    "dir_data = './data/'\n",
    "f_app_train = os.path.join(dir_data, 'application_train.csv')\n",
    "f_app_test = os.path.join(dir_data, 'application_test.csv')\n",
    "\n",
    "app_train = pd.read_csv(f_app_train)\n",
    "app_test = pd.read_csv(f_app_test)\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "\n",
    "# Create a label encoder object\n",
    "le = LabelEncoder()\n",
    "le_count = 0\n",
    "\n",
    "# Iterate through the columns\n",
    "for col in app_train:\n",
    "    if app_train[col].dtype == 'object':\n",
    "        # If 2 or fewer unique categories\n",
    "        if len(list(app_train[col].unique())) <= 2:\n",
    "            # Train on the training data\n",
    "            le.fit(app_train[col])\n",
    "            # Transform both training and testing data\n",
    "            app_train[col] = le.transform(app_train[col])\n",
    "            app_test[col] = le.transform(app_test[col])\n",
    "            \n",
    "            # Keep track of how many columns were label encoded\n",
    "            le_count += 1\n",
    "\n",
    "# one hot encoding\n",
    "app_train = pd.get_dummies(app_train)\n",
    "app_test = pd.get_dummies(app_test)\n",
    "\n",
    "# Create an anomalous(異常的) flag column\n",
    "# app_train[\"DAYS_EMPLOYED\"] == 365243 會回傳判斷的true/ false\n",
    "app_train['DAYS_EMPLOYED_ANOM'] = app_train[\"DAYS_EMPLOYED\"] == 365243\n",
    "app_train['DAYS_EMPLOYED'].replace({365243: np.nan}, inplace = True)\n",
    "# also apply to testing dataset\n",
    "app_test['DAYS_EMPLOYED_ANOM'] = app_test[\"DAYS_EMPLOYED\"] == 365243\n",
    "app_test[\"DAYS_EMPLOYED\"].replace({365243: np.nan}, inplace = True)\n",
    "\n",
    "# absolute the value of DAYS_BIRTH\n",
    "app_train['DAYS_BIRTH'] = abs(app_train['DAYS_BIRTH'])\n",
    "app_test['DAYS_BIRTH'] = abs(app_test['DAYS_BIRTH'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 相關係數\n",
    "一樣，pandas 很貼心地讓我們可以非常容易計算相關係數"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SK_ID_CURR                                 -0.002108\n",
       "TARGET                                      1.000000\n",
       "NAME_CONTRACT_TYPE                         -0.030896\n",
       "FLAG_OWN_CAR                               -0.021851\n",
       "FLAG_OWN_REALTY                            -0.006148\n",
       "CNT_CHILDREN                                0.019187\n",
       "AMT_INCOME_TOTAL                           -0.003982\n",
       "AMT_CREDIT                                 -0.030369\n",
       "AMT_ANNUITY                                -0.012817\n",
       "AMT_GOODS_PRICE                            -0.039645\n",
       "REGION_POPULATION_RELATIVE                 -0.037227\n",
       "DAYS_BIRTH                                 -0.078239\n",
       "DAYS_EMPLOYED                               0.074958\n",
       "DAYS_REGISTRATION                           0.041975\n",
       "DAYS_ID_PUBLISH                             0.051457\n",
       "OWN_CAR_AGE                                 0.037612\n",
       "FLAG_MOBIL                                  0.000534\n",
       "FLAG_EMP_PHONE                              0.045982\n",
       "FLAG_WORK_PHONE                             0.028524\n",
       "FLAG_CONT_MOBILE                            0.000370\n",
       "FLAG_PHONE                                 -0.023806\n",
       "FLAG_EMAIL                                 -0.001758\n",
       "CNT_FAM_MEMBERS                             0.009308\n",
       "REGION_RATING_CLIENT                        0.058899\n",
       "REGION_RATING_CLIENT_W_CITY                 0.060893\n",
       "HOUR_APPR_PROCESS_START                    -0.024166\n",
       "REG_REGION_NOT_LIVE_REGION                  0.005576\n",
       "REG_REGION_NOT_WORK_REGION                  0.006942\n",
       "LIVE_REGION_NOT_WORK_REGION                 0.002819\n",
       "REG_CITY_NOT_LIVE_CITY                      0.044395\n",
       "                                              ...   \n",
       "ORGANIZATION_TYPE_Trade: type 1             0.001032\n",
       "ORGANIZATION_TYPE_Trade: type 2            -0.003105\n",
       "ORGANIZATION_TYPE_Trade: type 3             0.008911\n",
       "ORGANIZATION_TYPE_Trade: type 4            -0.002621\n",
       "ORGANIZATION_TYPE_Trade: type 5            -0.000904\n",
       "ORGANIZATION_TYPE_Trade: type 6            -0.005788\n",
       "ORGANIZATION_TYPE_Trade: type 7             0.008170\n",
       "ORGANIZATION_TYPE_Transport: type 1        -0.003375\n",
       "ORGANIZATION_TYPE_Transport: type 2        -0.000839\n",
       "ORGANIZATION_TYPE_Transport: type 3         0.017552\n",
       "ORGANIZATION_TYPE_Transport: type 4         0.005929\n",
       "ORGANIZATION_TYPE_University               -0.007672\n",
       "ORGANIZATION_TYPE_XNA                      -0.045987\n",
       "FONDKAPREMONT_MODE_not specified           -0.002667\n",
       "FONDKAPREMONT_MODE_org spec account        -0.011285\n",
       "FONDKAPREMONT_MODE_reg oper account        -0.022587\n",
       "FONDKAPREMONT_MODE_reg oper spec account   -0.011257\n",
       "HOUSETYPE_MODE_block of flats              -0.040594\n",
       "HOUSETYPE_MODE_specific housing             0.005311\n",
       "HOUSETYPE_MODE_terraced house               0.000982\n",
       "WALLSMATERIAL_MODE_Block                   -0.006777\n",
       "WALLSMATERIAL_MODE_Mixed                   -0.001713\n",
       "WALLSMATERIAL_MODE_Monolithic              -0.009384\n",
       "WALLSMATERIAL_MODE_Others                   0.000628\n",
       "WALLSMATERIAL_MODE_Panel                   -0.033119\n",
       "WALLSMATERIAL_MODE_Stone, brick            -0.012657\n",
       "WALLSMATERIAL_MODE_Wooden                   0.007946\n",
       "EMERGENCYSTATE_MODE_No                     -0.042201\n",
       "EMERGENCYSTATE_MODE_Yes                     0.004829\n",
       "DAYS_EMPLOYED_ANOM                         -0.045987\n",
       "Name: TARGET, Length: 244, dtype: float64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "app_train.corr()['TARGET']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 練習時間\n",
    "列出目標 (TARGET) 與所有欄位之間相關係數，數值最大以及最小各 15 個\n",
    "\n",
    "通過相關係數的結果觀察有興趣的欄位與 TARGET 或其他欄位的相關係數，並嘗試找出有趣的訊息\n",
    "- 最好的方式當然是畫圖，舉例來說，我們知道  EXT_SOURCE_3 這個欄位和 TARGET 之間的相關係數是 -0.178919 (在已經這個資料集已經是最負的了！)，那我們可以 EXT_SOURCE_3  為 x 軸， TARGET 為 y 軸，把資料給畫出來"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TARGET與所有欄位之間相關係數，數值最大15 個:\n",
      "TARGET                                               1.000000\n",
      "DAYS_EMPLOYED                                        0.074958\n",
      "REGION_RATING_CLIENT_W_CITY                          0.060893\n",
      "REGION_RATING_CLIENT                                 0.058899\n",
      "NAME_INCOME_TYPE_Working                             0.057481\n",
      "DAYS_LAST_PHONE_CHANGE                               0.055218\n",
      "CODE_GENDER_M                                        0.054713\n",
      "DAYS_ID_PUBLISH                                      0.051457\n",
      "REG_CITY_NOT_WORK_CITY                               0.050994\n",
      "NAME_EDUCATION_TYPE_Secondary / secondary special    0.049824\n",
      "FLAG_EMP_PHONE                                       0.045982\n",
      "REG_CITY_NOT_LIVE_CITY                               0.044395\n",
      "FLAG_DOCUMENT_3                                      0.044346\n",
      "OCCUPATION_TYPE_Laborers                             0.043019\n",
      "DAYS_REGISTRATION                                    0.041975\n",
      "Name: TARGET, dtype: float64\n",
      "CPU times: user 40.3 s, sys: 582 ms, total: 40.9 s\n",
      "Wall time: 40.9 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "target_corr = app_train.corr()['TARGET']\n",
    "print('TARGET與所有欄位之間相關係數，數值最大15 個:')\n",
    "# target_corr是Series\n",
    "print(target_corr.sort_values(ascending=False).head(15))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TARGET與所有欄位之間相關係數，數值最小15 個\n",
      "EXT_SOURCE_3                           -0.178919\n",
      "EXT_SOURCE_2                           -0.160472\n",
      "EXT_SOURCE_1                           -0.155317\n",
      "DAYS_BIRTH                             -0.078239\n",
      "NAME_EDUCATION_TYPE_Higher education   -0.056593\n",
      "CODE_GENDER_F                          -0.054704\n",
      "NAME_INCOME_TYPE_Pensioner             -0.046209\n",
      "ORGANIZATION_TYPE_XNA                  -0.045987\n",
      "DAYS_EMPLOYED_ANOM                     -0.045987\n",
      "FLOORSMAX_AVG                          -0.044003\n",
      "FLOORSMAX_MEDI                         -0.043768\n",
      "FLOORSMAX_MODE                         -0.043226\n",
      "EMERGENCYSTATE_MODE_No                 -0.042201\n",
      "HOUSETYPE_MODE_block of flats          -0.040594\n",
      "AMT_GOODS_PRICE                        -0.039645\n",
      "Name: TARGET, dtype: float64\n",
      "CPU times: user 43.2 s, sys: 660 ms, total: 43.8 s\n",
      "Wall time: 43.9 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print('TARGET與所有欄位之間相關係數，數值最小15 個')\n",
    "print(target_corr.sort_values(ascending=True).head(15))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 圖示顯示"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'EXT_SOURCE_3')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(app_train['EXT_SOURCE_3'], app_train['TARGET'])\n",
    "plt.xlabel('TARGET')\n",
    "plt.ylabel('EXT_SOURCE_3')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "app_train.boxplot(column='EXT_SOURCE_3', by = 'TARGET', vert=False)\n",
    "# 避免預設的'boxplot grouped by target' title\n",
    "plt.suptitle('')\n",
    "plt.xlabel(None)\n",
    "plt.ylabel('TARGET')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10f1c99b0>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "sns.boxplot(x='TARGET', y='EXT_SOURCE_3', data=app_train, width=0.6)"
   ]
  },
  {
   "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": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}