{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": true,
    "pycharm": {
     "name": "#%%\n"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 150000 entries, 1 to 150000\n",
      "Data columns (total 11 columns):\n",
      "SeriousDlqin2yrs                        150000 non-null int64\n",
      "RevolvingUtilizationOfUnsecuredLines    150000 non-null float64\n",
      "age                                     150000 non-null int64\n",
      "NumberOfTime30-59DaysPastDueNotWorse    150000 non-null int64\n",
      "DebtRatio                               150000 non-null float64\n",
      "MonthlyIncome                           120269 non-null float64\n",
      "NumberOfOpenCreditLinesAndLoans         150000 non-null int64\n",
      "NumberOfTimes90DaysLate                 150000 non-null int64\n",
      "NumberRealEstateLoansOrLines            150000 non-null int64\n",
      "NumberOfTime60-89DaysPastDueNotWorse    150000 non-null int64\n",
      "NumberOfDependents                      146076 non-null float64\n",
      "dtypes: float64(4), int64(7)\n",
      "memory usage: 13.7 MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "data = pd.read_csv(\"Data/cs-training.csv\",index_col=0)\n",
    "print(data.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns',None)\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001B[7;37;41m\t 数据详细描述： \u001B[0m\n",
      "       SeriousDlqin2yrs  RevolvingUtilizationOfUnsecuredLines            age  \\\n",
      "count     150000.000000                         150000.000000  150000.000000   \n",
      "mean           0.066840                              6.048438      52.295207   \n",
      "std            0.249746                            249.755371      14.771866   \n",
      "min            0.000000                              0.000000       0.000000   \n",
      "25%            0.000000                              0.029867      41.000000   \n",
      "50%            0.000000                              0.154181      52.000000   \n",
      "75%            0.000000                              0.559046      63.000000   \n",
      "max            1.000000                          50708.000000     109.000000   \n",
      "\n",
      "       NumberOfTime30-59DaysPastDueNotWorse      DebtRatio  MonthlyIncome  \\\n",
      "count                         150000.000000  150000.000000   1.202690e+05   \n",
      "mean                               0.421033     353.005076   6.670221e+03   \n",
      "std                                4.192781    2037.818523   1.438467e+04   \n",
      "min                                0.000000       0.000000   0.000000e+00   \n",
      "25%                                0.000000       0.175074   3.400000e+03   \n",
      "50%                                0.000000       0.366508   5.400000e+03   \n",
      "75%                                0.000000       0.868254   8.249000e+03   \n",
      "max                               98.000000  329664.000000   3.008750e+06   \n",
      "\n",
      "       NumberOfOpenCreditLinesAndLoans  NumberOfTimes90DaysLate  \\\n",
      "count                    150000.000000            150000.000000   \n",
      "mean                          8.452760                 0.265973   \n",
      "std                           5.145951                 4.169304   \n",
      "min                           0.000000                 0.000000   \n",
      "25%                           5.000000                 0.000000   \n",
      "50%                           8.000000                 0.000000   \n",
      "75%                          11.000000                 0.000000   \n",
      "max                          58.000000                98.000000   \n",
      "\n",
      "       NumberRealEstateLoansOrLines  NumberOfTime60-89DaysPastDueNotWorse  \\\n",
      "count                 150000.000000                         150000.000000   \n",
      "mean                       1.018240                              0.240387   \n",
      "std                        1.129771                              4.155179   \n",
      "min                        0.000000                              0.000000   \n",
      "25%                        0.000000                              0.000000   \n",
      "50%                        1.000000                              0.000000   \n",
      "75%                        2.000000                              0.000000   \n",
      "max                       54.000000                             98.000000   \n",
      "\n",
      "       NumberOfDependents  \n",
      "count       146076.000000  \n",
      "mean             0.757222  \n",
      "std              1.115086  \n",
      "min              0.000000  \n",
      "25%              0.000000  \n",
      "50%              0.000000  \n",
      "75%              1.000000  \n",
      "max             20.000000  \n"
     ]
    }
   ],
   "source": [
    "pd.set_option('display.max_rows', None)\n",
    "print(\"\\033[7;37;41m\\t 数据详细描述： \\033[0m\")\n",
    "print(data.describe())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\programdata\\anaconda3\\envs\\py3.5\\lib\\site-packages\\ipykernel_launcher.py:4: FutureWarning: \n",
      ".ix is deprecated. Please use\n",
      ".loc for label based indexing or\n",
      ".iloc for positional indexing\n",
      "\n",
      "See the documentation here:\n",
      "http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#ix-indexer-is-deprecated\n",
      "  after removing the cwd from sys.path.\n",
      "c:\\programdata\\anaconda3\\envs\\py3.5\\lib\\site-packages\\pandas\\core\\indexing.py:822: FutureWarning: \n",
      ".ix is deprecated. Please use\n",
      ".loc for label based indexing or\n",
      ".iloc for positional indexing\n",
      "\n",
      "See the documentation here:\n",
      "http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#ix-indexer-is-deprecated\n",
      "  retval = getattr(retval, self.name)._getitem_axis(key, axis=i)\n",
      "c:\\programdata\\anaconda3\\envs\\py3.5\\lib\\site-packages\\ipykernel_launcher.py:6: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  \n",
      "c:\\programdata\\anaconda3\\envs\\py3.5\\lib\\site-packages\\ipykernel_launcher.py:7: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
      "  import sys\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.         0.02505911 0.0189291  0.00146759 0.85552865 0.03058563\n",
      " 0.         0.06842991 0.        ]\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 145563 entries, 1 to 150000\n",
      "Data columns (total 11 columns):\n",
      "好坏客户          145563 non-null int64\n",
      "可用额度比值        145563 non-null float64\n",
      "年龄            145563 non-null int64\n",
      "逾期30-59天笔数    145563 non-null int64\n",
      "负债率           145563 non-null float64\n",
      "月收入           145563 non-null float64\n",
      "信贷数量          145563 non-null int64\n",
      "逾期90天笔数       145563 non-null int64\n",
      "固定资产贷款量       145563 non-null int64\n",
      "逾期60-89天笔数    145563 non-null int64\n",
      "家属数量          145563 non-null float64\n",
      "dtypes: float64(4), int64(7)\n",
      "memory usage: 13.3 MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "def set_missing(df):\n",
    "    # 把已有的数值型特征取出来，MonthlyIncome位于第5列，NumberOfDependents位于第10列\n",
    "    process_df = df.ix[:, [5, 0, 1, 2, 3, 4, 6, 7, 8, 9]]\n",
    "    # 分成已知该特征和未知该特征两部分\n",
    "    known = process_df[process_df.MonthlyIncome.notnull()].as_matrix()\n",
    "    unknown = process_df[process_df.MonthlyIncome.isnull()].as_matrix()\n",
    "    X = known[:, 1:]    # X为特征属性值\n",
    "    y = known[:, 0]    # y为结果标签值\n",
    "    rfr = RandomForestRegressor(random_state=0, n_estimators=200, max_depth=3)\n",
    "    rfr.fit(X, y)\n",
    "    print(rfr.feature_importances_)\n",
    "\n",
    "    # 用得到的模型进行未知特征值预测月收入\n",
    "    predicted = rfr.predict(unknown[:, 1:]).round(0)\n",
    "    # 用得到的预测结果填补原缺失数据\n",
    "    df.loc[df.MonthlyIncome.isnull(), 'MonthlyIncome'] = predicted\n",
    "    return df\n",
    "\n",
    "#用随机森林填补比较多的缺失值\n",
    "data=set_missing(data)\n",
    "data = data.dropna()\n",
    "data = data.drop_duplicates()\n",
    "states={\n",
    "        'SeriousDlqin2yrs':'好坏客户',\n",
    "        'RevolvingUtilizationOfUnsecuredLines':'可用额度比值', #无担保放款循环利用比值\n",
    "        'age':'年龄',\n",
    "        'NumberOfTime30-59DaysPastDueNotWorse':'逾期30-59天笔数',\n",
    "        'DebtRatio':'负债率',\n",
    "        'MonthlyIncome':'月收入',\n",
    "        'NumberOfOpenCreditLinesAndLoans':'信贷数量',\n",
    "        'NumberOfTimes90DaysLate':'逾期90天笔数',\n",
    "        'NumberRealEstateLoansOrLines':'固定资产贷款量',\n",
    "        'NumberOfTime60-89DaysPastDueNotWorse':'逾期60-89天笔数',\n",
    "        'NumberOfDependents':'家属数量'\n",
    "         } #创建字典\n",
    "\n",
    "data.rename(columns=states,inplace=True)\n",
    "print(data.info())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "outputs": [
    {
     "data": {
      "text/plain": "{'boxes': [<matplotlib.lines.Line2D at 0x1d5a3e8c9b0>],\n 'caps': [<matplotlib.lines.Line2D at 0x1d5a3e8d2b0>,\n  <matplotlib.lines.Line2D at 0x1d5a3e8d6d8>],\n 'fliers': [<matplotlib.lines.Line2D at 0x1d5a3e8df28>],\n 'means': [],\n 'medians': [<matplotlib.lines.Line2D at 0x1d5a3e8db00>],\n 'whiskers': [<matplotlib.lines.Line2D at 0x1d5a3e8cb00>,\n  <matplotlib.lines.Line2D at 0x1d5a3e8ce48>]}"
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 1080x720 with 6 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "mpl.rcParams['font.sans-serif'] = ['KaiTi']\n",
    "mpl.rcParams['font.serif'] = ['KaiTi']\n",
    "# train_box = data.iloc[:,[3,7,9]]\n",
    "# train_box.boxplot(figsize=(10,4))\n",
    "# plt.show()\n",
    "fig=plt.figure(figsize=(15,10))\n",
    "a=fig.add_subplot(3,2,1)\n",
    "b=fig.add_subplot(3,2,2)\n",
    "c=fig.add_subplot(3,2,3)\n",
    "d=fig.add_subplot(3,2,4)\n",
    "e=fig.add_subplot(3,2,5)\n",
    "f=fig.add_subplot(3,2,6)\n",
    "\n",
    "a.boxplot(data['可用额度比值'],labels=['可用额度比值'])\n",
    "b.boxplot([data['年龄'],data['好坏客户']],labels=['年龄','好坏客户'])\n",
    "c.boxplot([data['逾期30-59天笔数'],data['逾期60-89天笔数'],data['逾期90天笔数']],labels=['逾期30-59天笔数','逾期60-89天笔数','逾期90天笔数'])\n",
    "d.boxplot([data['信贷数量'],data['固定资产贷款量'],data['家属数量']],labels=['信贷数量','固定资产贷款量','家属数量'])\n",
    "e.boxplot(data['月收入'],labels=['月收入'])\n",
    "f.boxplot(data['负债率'],labels=['负债率'])"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "54\n"
     ]
    }
   ],
   "source": [
    "print(data['固定资产贷款量'].max())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "outputs": [
    {
     "data": {
      "text/plain": "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001D5A3E9EE80>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D5965B4128>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D5965BB438>],\n       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001D5A57F3748>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D5A57D9CF8>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D59826A3C8>],\n       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001D598268A58>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D598245128>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D5A3DE3198>],\n       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001D593013F60>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D5A5D20E10>,\n        <matplotlib.axes._subplots.AxesSubplot object at 0x000001D593058B38>]],\n      dtype=object)"
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 1440x1080 with 12 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.hist(figsize=(20,15))\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5689025735\n",
      "62.0\n",
      "0.43679564175\n",
      "7912.25\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 105382 entries, 1 to 150000\n",
      "Data columns (total 11 columns):\n",
      "好坏客户          105382 non-null int64\n",
      "可用额度比值        105382 non-null float64\n",
      "年龄            105382 non-null int64\n",
      "逾期30-59天笔数    105382 non-null int64\n",
      "负债率           105382 non-null float64\n",
      "月收入           105382 non-null float64\n",
      "信贷数量          105382 non-null int64\n",
      "逾期90天笔数       105382 non-null int64\n",
      "固定资产贷款量       105382 non-null int64\n",
      "逾期60-89天笔数    105382 non-null int64\n",
      "家属数量          105382 non-null float64\n",
      "dtypes: float64(4), int64(7)\n",
      "memory usage: 9.6 MB\n"
     ]
    },
    {
     "data": {
      "text/plain": "<matplotlib.axes._subplots.AxesSubplot at 0x1d598c48ba8>"
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 1080x720 with 1 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for k in [1,2,4,5]: #遍历列\n",
    "    q1=data.iloc[:,k].quantile(0.25)  #计算上四分位数\n",
    "    q3=data.iloc[:,k].quantile(0.75)  #计算下四分位数\n",
    "    iqr=q3-q1\n",
    "    print(q3)\n",
    "    low=q1-1.5*iqr\n",
    "    up=q3+1.5*iqr\n",
    "    if k==1:\n",
    "        data1=data\n",
    "    data1=data1[(data1.iloc[:,k]>low) & (data1.iloc[:,k]< up)]  #保留正常值范围\n",
    "data=data1\n",
    "data.info()\n",
    "data.iloc[:,[1,2,4,5]].boxplot(figsize=(15,10))\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6948\n"
     ]
    }
   ],
   "source": [
    "data=data[data['逾期30-59天笔数']<80]\n",
    "data=data[data['逾期60-89天笔数']<80]\n",
    "data=data[data['逾期60-89天笔数']<80]\n",
    "data=data[data['逾期90天笔数']<80]\n",
    "data=data[data['固定资产贷款量']<50]\n",
    "data=data[data['家属数量']<15]\n",
    "print(data['好坏客户'].sum())"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           iv       woe  坏客户数   好客户数\n",
      "可用额度比值                                              \n",
      "(0.999, 21040.2]     0.185992 -1.244711   420  20620\n",
      "(21040.2, 42079.4]   0.219103 -1.387684   365  20674\n",
      "(42079.4, 63118.6]   0.088399 -0.784102   658  20381\n",
      "(63118.6, 84157.8]   0.000882  0.065475  1477  19562\n",
      "(84157.8, 105197.0]  0.491316  1.208402  4028  17012\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "年龄                                                  \n",
      "(0.999, 13150.5]     0.042872  0.522469  1401  11749\n",
      "(13150.5, 26300.0]   0.014351  0.316231  1163  11987\n",
      "(26300.0, 39449.5]   0.006996  0.225241  1070  12079\n",
      "(39449.5, 52599.0]   0.003940  0.171045  1018  12132\n",
      "(52599.0, 65748.5]   0.000035  0.016575   882  12267\n",
      "(65748.5, 78898.0]   0.008521 -0.277127   669  12481\n",
      "(78898.0, 92047.5]   0.050143 -0.740423   429  12720\n",
      "(92047.5, 105197.0]  0.089835 -1.055060   316  12834\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "负债率                                                 \n",
      "(0.999, 26300.0]     0.005562 -0.154208  1503  24797\n",
      "(26300.0, 52599.0]   0.002878 -0.109884  1567  24732\n",
      "(52599.0, 78898.0]   0.002674 -0.105819  1573  24726\n",
      "(78898.0, 105197.0]  0.026814  0.306328  2305  23994\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "月收入                                                 \n",
      "(0.999, 21040.2]     0.027856  0.346048  1912  19128\n",
      "(21040.2, 42079.4]   0.005916  0.165894  1621  19418\n",
      "(42079.4, 63118.6]   0.000019 -0.009729  1377  19662\n",
      "(63118.6, 84157.8]   0.009978 -0.234967  1114  19925\n",
      "(84157.8, 105197.0]  0.030964 -0.431508   924  20116\n",
      "                              iv       woe  坏客户数   好客户数\n",
      "信贷数量                                                   \n",
      "(0.999, 17533.667]      0.044260  0.465503  1775  15758\n",
      "(17533.667, 35066.333]  0.000004 -0.004643  1153  16380\n",
      "(35066.333, 52599.0]    0.005834 -0.195146   964  16569\n",
      "(52599.0, 70131.667]    0.007592 -0.223995   938  16594\n",
      "(70131.667, 87664.333]  0.002626 -0.129066  1026  16507\n",
      "(87664.333, 105197.0]   0.000638 -0.062716  1092  16441\n"
     ]
    }
   ],
   "source": [
    "from scipy import stats  \n",
    "import numpy as np\n",
    "#定义自动分箱函数，采用最优分段进行分箱\n",
    "'''采用斯皮尔曼等级相关系数进行变量相关分析，该相关系数对两个变量划分等级在进行分析，[-1,1]，\n",
    "当两个变量完全单调相关时，斯皮尔曼相关系数则为+1或−1'''\n",
    "\n",
    "def op(y,x,n=20): #定义函数，有三个参数，x为要分箱的变量，y对应关注的因变量，即好坏客户，n最大分组数为20，依次减小试验\n",
    "    r=0\n",
    "    bad=y.sum()         #计算坏客户个数，因为等于1时表示坏客户，所以求和即可得到好客户数\n",
    "    good=y.count()-bad   #count统计个数为所有客户数，减去好客户得到好客户人数\n",
    "    while np.abs(r)<1:   #判断，当绝对值小于1时继续执行，直到相关系数绝对值等于1停止\n",
    "\n",
    "        d1=pd.DataFrame({\"x\":x,\"y\":y,\"bucket\":pd.qcut(x,n)}) \n",
    "        d2=d1.groupby('bucket',as_index=True) #对数据框d1按照bucket变量分组，Ture则返回以组标签为索引的对象,reset_index()可以取消分组索引让其变成dataframe \n",
    "        r,p=stats.spearmanr(d2.mean().x,d2.mean().y)  #分组，数据为组间均值，计算此时x与y的斯皮尔曼等级相关系数，输出相关系数和P值\n",
    "        n=n-1               #减小分组数\n",
    "    d3=pd.DataFrame(d2.x.min(),columns=['min'])  #建立数据框\n",
    "    d3['min']=d2.min().x\n",
    "    d3['max']=d2.max().x\n",
    "    d3['sum']=d2.sum().y #对应分组的坏客户数\n",
    "    d3['total']=d2.count().y  #对应分组总客户数\n",
    "    d3['rate']=d2.mean().y     #坏客户数占该组总人数比\n",
    "    d3['woe']=np.log((d3['rate']/(1-d3['rate']))/(good/bad))   #求woe\n",
    "    # d3['iv']=\n",
    "    \n",
    "    d4=(d3.sort_index(by='min')).reset_index(drop=True)\n",
    "    print(\"=\"* 60)\n",
    "    print(d4)\n",
    "    return(d4)\n",
    "#利用所定义的函数依次对连续型变量进行最优分段分箱处理,满足条件的有以下\n",
    "# x2=op(data['好坏客户'],data['年龄'])\n",
    "# x4=op(data['好坏客户'],data['负债率'])\n",
    "# x5=op(data['好坏客户'],data['月收入'])\n",
    "\n",
    "#对于不能采用最优分段的变量采用等深分段\n",
    "\n",
    "def funqcut(y,x,n):\n",
    "    cut1=pd.qcut(x.rank(method='first'),n)  #进行等深分箱，分组\n",
    "\n",
    "    data=pd.DataFrame({\"x\":x,\"y\":y,\"cut1\":cut1})\n",
    "    cutbad=data.groupby(cut1).y.sum()   #求分组下的坏客户数\n",
    "    cutgood=data.groupby(cut1).y.count()-cutbad #求分组下好客户数\n",
    "    bad=data.y.sum() #求总的坏客户数\n",
    "    good=data.y.count()-bad #求总的好客户数\n",
    "\n",
    "    woe=np.log((cutbad/bad)/(cutgood/good)) #求各分组的woe\n",
    "    iv=(cutbad/bad-cutgood/good)*woe  #求各分组的iv\n",
    "\n",
    "    cut=pd.DataFrame({\"坏客户数\":cutbad,\"好客户数\":cutgood,\"woe\":woe,\"iv\":iv})\n",
    "    print(cut)\n",
    "\n",
    "    return cut#返回表格和对应分组列表\n",
    "\n",
    "#funqcut(train['好坏客户'],train['年龄'],6).reset_index()\n",
    "\n",
    "x1=funqcut(data['好坏客户'],data['可用额度比值'],5).reset_index()\n",
    "x2=funqcut(data['好坏客户'],data['年龄'],8).reset_index()\n",
    "x4=funqcut(data['好坏客户'],data['负债率'],4).reset_index()\n",
    "x5=funqcut(data['好坏客户'],data['月收入'],5).reset_index()\n",
    "x6=funqcut(data['好坏客户'],data['信贷数量'],6).reset_index()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           iv       woe  坏客户数   好客户数\n",
      "逾期30-59天笔数                                          \n",
      "(0.999, 21040.2]     0.041161 -0.505252   861  20179\n",
      "(21040.2, 42079.4]   0.039068 -0.490767   873  20166\n",
      "(42079.4, 63118.6]   0.047825 -0.549696   825  20214\n",
      "(63118.6, 84157.8]   0.043487 -0.521060   848  20191\n",
      "(84157.8, 105197.0]  0.348548  1.051317  3541  17499\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "逾期90天笔数                                             \n",
      "(0.999, 21040.2]     0.030816 -0.430376   925  20115\n",
      "(21040.2, 42079.4]   0.021164 -0.350731   998  20041\n",
      "(42079.4, 63118.6]   0.027123 -0.401313   951  20088\n",
      "(63118.6, 84157.8]   0.019221 -0.332992  1015  20024\n",
      "(84157.8, 105197.0]  0.225825  0.877824  3059  17981\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "固定资产贷款量                                             \n",
      "(0.999, 21040.2]     0.014273  0.252814  1756  19284\n",
      "(21040.2, 42079.4]   0.009963  0.213068  1693  19346\n",
      "(42079.4, 63118.6]   0.010522 -0.241622  1107  19932\n",
      "(63118.6, 84157.8]   0.006738 -0.191281  1161  19878\n",
      "(84157.8, 105197.0]  0.003160 -0.129258  1231  19809\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "逾期60-89天笔数                                          \n",
      "(0.999, 21040.2]     0.018677 -0.327878  1020  20020\n",
      "(21040.2, 42079.4]   0.008374 -0.214306  1136  19903\n",
      "(42079.4, 63118.6]   0.015560 -0.297341  1050  19989\n",
      "(63118.6, 84157.8]   0.010287 -0.238765  1110  19929\n",
      "(84157.8, 105197.0]  0.134785  0.704010  2632  18408\n",
      "                           iv       woe  坏客户数   好客户数\n",
      "家属数量                                                \n",
      "(0.999, 21040.2]     0.005350 -0.169662  1185  19855\n",
      "(21040.2, 42079.4]   0.005026 -0.164259  1191  19848\n",
      "(42079.4, 63118.6]   0.003670 -0.139609  1219  19820\n",
      "(63118.6, 84157.8]   0.001996  0.097791  1522  19517\n",
      "(84157.8, 105197.0]  0.020305  0.298535  1831  19209\n"
     ]
    }
   ],
   "source": [
    "x3=funqcut(data['好坏客户'],data['逾期30-59天笔数'],5).reset_index()\n",
    "x7=funqcut(data['好坏客户'],data['逾期90天笔数'],5).reset_index()\n",
    "x8=funqcut(data['好坏客户'],data['固定资产贷款量'],5).reset_index()\n",
    "x9=funqcut(data['好坏客户'],data['逾期60-89天笔数'],5).reset_index()\n",
    "x10=funqcut(data['好坏客户'],data['家属数量'],5).reset_index()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "outputs": [
    {
     "data": {
      "text/plain": "Text(0.5,1,'特征变量的IV值分布')"
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 1080x720 with 1 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ivx1=x1.iv.sum()\n",
    "ivx2=x2.iv.sum()\n",
    "ivx3=x3.iv.sum()\n",
    "ivx4=x4.iv.sum()\n",
    "ivx5=x5.iv.sum()\n",
    "ivx6=x6.iv.sum()\n",
    "ivx7=x7.iv.sum()\n",
    "ivx8=x8.iv.sum()\n",
    "ivx9=x9.iv.sum()\n",
    "ivx10=x10.iv.sum()\n",
    "IV=pd.DataFrame({\"可用额度比值\":ivx1,\n",
    "                 \"年龄\":ivx2,\n",
    "                 \"逾期30-59天笔数\":ivx3,\n",
    "                 \"负债率\":ivx4,\n",
    "                 \"月收入\":ivx5,\n",
    "                 \"信贷数量\":ivx6,\n",
    "                 \"逾期90天笔数\":ivx7,\n",
    "                 \"固定资产贷款量\":ivx8,\n",
    "                 \"逾期60-89天笔数\":ivx9,\n",
    "                 \"家属数量\":ivx10},index=[0])\n",
    "\n",
    "ivplot=IV.plot.bar(figsize=(15,10))\n",
    "ivplot.set_title('特征变量的IV值分布')"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.series.Series'>\n"
     ]
    }
   ],
   "source": [
    "#利用等深分段分箱设定，将各个变量的分箱后情况保存为cut1,cut2，cut3...与所得到的分箱情况总结表相对应。其中采用x2/x4/x5采用最优分段的结果\n",
    "def cutdata(x,n):\n",
    "    a=pd.qcut(x.rank(method='first'),n,labels=False)#等深分组，label=False返回整数值（0,1,2,3...）对应第一类、第二类..\n",
    "    return a\n",
    "#连续变量均被等深分为了5类\n",
    "\n",
    "#应用函数，求出各变量分类情况\n",
    "cut1=cutdata(data['可用额度比值'],5)\n",
    "cut2=cutdata(data['年龄'],8)\n",
    "cut3=cutdata(data['逾期30-59天笔数'],5)\n",
    "cut4=cutdata(data['负债率'],4)\n",
    "cut5=cutdata(data['月收入'],5)\n",
    "cut6=cutdata(data['信贷数量'],6)\n",
    "cut7=cutdata(data['逾期90天笔数'],5)\n",
    "cut8=cutdata(data['固定资产贷款量'],5)\n",
    "cut9=cutdata(data['逾期60-89天笔数'],5)\n",
    "cut10=cutdata(data['家属数量'],5)\n",
    "\n",
    "print(type(cut2))"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "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>好坏客户</th>\n      <th>可用额度比值</th>\n      <th>年龄</th>\n      <th>逾期30-59天笔数</th>\n      <th>负债率</th>\n      <th>月收入</th>\n      <th>信贷数量</th>\n      <th>逾期90天笔数</th>\n      <th>固定资产贷款量</th>\n      <th>逾期60-89天笔数</th>\n      <th>家属数量</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>1.208402</td>\n      <td>0.225241</td>\n      <td>1.051317</td>\n      <td>0.306328</td>\n      <td>-0.431508</td>\n      <td>-0.129066</td>\n      <td>-0.430376</td>\n      <td>-0.129258</td>\n      <td>-0.327878</td>\n      <td>0.097791</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0</td>\n      <td>1.208402</td>\n      <td>0.316231</td>\n      <td>-0.505252</td>\n      <td>-0.154208</td>\n      <td>0.346048</td>\n      <td>0.465503</td>\n      <td>-0.430376</td>\n      <td>0.252814</td>\n      <td>-0.327878</td>\n      <td>-0.139609</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0</td>\n      <td>0.065475</td>\n      <td>0.316231</td>\n      <td>1.051317</td>\n      <td>-0.154208</td>\n      <td>0.346048</td>\n      <td>0.465503</td>\n      <td>0.877824</td>\n      <td>0.252814</td>\n      <td>-0.327878</td>\n      <td>-0.169662</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0</td>\n      <td>-0.784102</td>\n      <td>0.522469</td>\n      <td>-0.505252</td>\n      <td>-0.154208</td>\n      <td>0.165894</td>\n      <td>-0.004643</td>\n      <td>-0.430376</td>\n      <td>0.252814</td>\n      <td>-0.327878</td>\n      <td>-0.169662</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>0</td>\n      <td>-0.784102</td>\n      <td>-1.055060</td>\n      <td>-0.505252</td>\n      <td>-0.105819</td>\n      <td>0.165894</td>\n      <td>0.465503</td>\n      <td>-0.430376</td>\n      <td>0.213068</td>\n      <td>-0.327878</td>\n      <td>-0.139609</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "   好坏客户    可用额度比值        年龄  逾期30-59天笔数       负债率       月收入      信贷数量  \\\n1     1  1.208402  0.225241    1.051317  0.306328 -0.431508 -0.129066   \n2     0  1.208402  0.316231   -0.505252 -0.154208  0.346048  0.465503   \n3     0  0.065475  0.316231    1.051317 -0.154208  0.346048  0.465503   \n4     0 -0.784102  0.522469   -0.505252 -0.154208  0.165894 -0.004643   \n6     0 -0.784102 -1.055060   -0.505252 -0.105819  0.165894  0.465503   \n\n    逾期90天笔数   固定资产贷款量  逾期60-89天笔数      家属数量  \n1 -0.430376 -0.129258   -0.327878  0.097791  \n2 -0.430376  0.252814   -0.327878 -0.139609  \n3  0.877824  0.252814   -0.327878 -0.169662  \n4 -0.430376  0.252814   -0.327878 -0.169662  \n6 -0.430376  0.213068   -0.327878 -0.139609  "
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#依据变量值的分类替换成对应的woe值\n",
    "def replace_train(cut,cut_woe):  #定义替换函数，cut为分组情况，cut_woe为分组对应woe值\n",
    "    a=[]\n",
    "    for i in cut.unique(): #unique为去重，保留唯一值\n",
    "        a.append(i)\n",
    "        a.sort()  #排序，默认小到大，得到类别列表并排序，实则为[0,1,2,3,4]\n",
    "    for m in range(len(a)):\n",
    "        cut.replace(a[m],cut_woe.values[m],inplace=True) #替换函数，把cut中旧数值a[m]即分类替换为对应woe,cut_woe中的woe也是从小到大排序，因此与a[m]对应，正如把cut中的0替换为woe值，没有改变cut的数值顺序\n",
    "    return cut  #返回被替换后的列表\n",
    "\n",
    "train_new=pd.DataFrame() #创建新数据框保存替换后的新数据\n",
    "train_new['好坏客户']=data['好坏客户']\n",
    "train_new['可用额度比值']=replace_train(cut1,x1.woe)\n",
    "train_new['年龄']=replace_train(cut2,x2.woe)\n",
    "train_new['逾期30-59天笔数']=replace_train(cut3,x3.woe)\n",
    "train_new['负债率']=replace_train(cut4,x4.woe)\n",
    "train_new['月收入']=replace_train(cut5,x5.woe)\n",
    "train_new['信贷数量']=replace_train(cut6,x6.woe)\n",
    "train_new['逾期90天笔数']=replace_train(cut7,x7.woe)\n",
    "train_new['固定资产贷款量']=replace_train(cut8,x8.woe)\n",
    "train_new['逾期60-89天笔数']=replace_train(cut9,x9.woe)\n",
    "train_new['家属数量']=replace_train(cut10,x10.woe)\n",
    "train_new.head()\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "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>好坏客户</th>\n      <th>可用额度比值</th>\n      <th>年龄</th>\n      <th>逾期30-59天笔数</th>\n      <th>逾期90天笔数</th>\n      <th>逾期60-89天笔数</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>1.208402</td>\n      <td>0.225241</td>\n      <td>1.051317</td>\n      <td>-0.430376</td>\n      <td>-0.327878</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>0</td>\n      <td>1.208402</td>\n      <td>0.316231</td>\n      <td>-0.505252</td>\n      <td>-0.430376</td>\n      <td>-0.327878</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>0</td>\n      <td>0.065475</td>\n      <td>0.316231</td>\n      <td>1.051317</td>\n      <td>0.877824</td>\n      <td>-0.327878</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>0</td>\n      <td>-0.784102</td>\n      <td>0.522469</td>\n      <td>-0.505252</td>\n      <td>-0.430376</td>\n      <td>-0.327878</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>0</td>\n      <td>-0.784102</td>\n      <td>-1.055060</td>\n      <td>-0.505252</td>\n      <td>-0.430376</td>\n      <td>-0.327878</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "   好坏客户    可用额度比值        年龄  逾期30-59天笔数   逾期90天笔数  逾期60-89天笔数\n1     1  1.208402  0.225241    1.051317 -0.430376   -0.327878\n2     0  1.208402  0.316231   -0.505252 -0.430376   -0.327878\n3     0  0.065475  0.316231    1.051317  0.877824   -0.327878\n4     0 -0.784102  0.522469   -0.505252 -0.430376   -0.327878\n6     0 -0.784102 -1.055060   -0.505252 -0.430376   -0.327878"
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression  #导入logistic回归模块\n",
    "from sklearn.cross_validation import train_test_split #导入数据切分函数\n",
    "\n",
    "'''根据前文的变量选择分析，将负债率、月收入、信贷数量、\n",
    "固定资产贷款量、家属数量变量舍弃，不纳入模型中'''\n",
    "train_new1=train_new.drop([\"负债率\",\"月收入\",\"信贷数量\",\"固定资产贷款量\",\"家属数量\"],axis=1)\n",
    "train_new1.head()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "outputs": [
    {
     "data": {
      "text/plain": "0.9355513307984791"
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x=train_new1.iloc[:,1:] #设定自变量\n",
    "y=train_new.iloc[:,0] #设定因变量\n",
    "\n",
    "#将数据集进行切割，分成训练集和测试集，其中样本占比0.8，采用随机抽样\n",
    "train_x,test_x,train_y,test_y=train_test_split(x,y,train_size=0.8,random_state=4)\n",
    "\n",
    "#建立模型\n",
    "model=LogisticRegression()\n",
    "result=model.fit(train_x,train_y) #训练模型，将结果保存为result\n",
    "pred_y=model.predict(test_x)  #预测测试集的y\n",
    "result.score(test_x,test_y)    #计算预测精度 正确率\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "516.121509491651\n",
      "43.2808512266689\n",
      "[-2.64317602]\n",
      "[[ 0.82927648  0.51405853  0.64271826  0.61278088 -0.01169952]]\n",
      "630.5204174451445\n"
     ]
    }
   ],
   "source": [
    "'''假设比例即违约与正常比v为1/70,此时预期分值Z为700，PDD（比率翻倍的分数）为30\n",
    "B=PDD/log(2)\n",
    "A=Z+B*log(v) '''\n",
    "#计算A、B\n",
    "B=30/np.log(2)\n",
    "A=700+B*np.log(1/70)\n",
    "print(A)\n",
    "print(B)\n",
    "#计算基础分值A-BP0，参考上文\n",
    "c=result.intercept_ #输出logistic模型的截距项\n",
    "print(c)\n",
    "coef=result.coef_   #输出回归参数\n",
    "print(coef)\n",
    "BaseScore=A-B*c[0]    #计算基础分值\n",
    "print(BaseScore)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "可用额度比值\n",
      "                     分组    得分\n",
      "0     (0.999, 21040.2] -45.0\n",
      "1   (21040.2, 42079.4] -50.0\n",
      "2   (42079.4, 63118.6] -28.0\n",
      "3   (63118.6, 84157.8]   2.0\n",
      "4  (84157.8, 105197.0]  43.0\n",
      "年龄\n",
      "                     分组    得分\n",
      "0     (0.999, 13150.5]  12.0\n",
      "1   (13150.5, 26300.0]   7.0\n",
      "2   (26300.0, 39449.5]   5.0\n",
      "3   (39449.5, 52599.0]   4.0\n",
      "4   (52599.0, 65748.5]   0.0\n",
      "5   (65748.5, 78898.0]  -6.0\n",
      "6   (78898.0, 92047.5] -16.0\n",
      "7  (92047.5, 105197.0] -23.0\n",
      "逾期30-59天笔数\n",
      "                     分组    得分\n",
      "0     (0.999, 21040.2] -14.0\n",
      "1   (21040.2, 42079.4] -14.0\n",
      "2   (42079.4, 63118.6] -15.0\n",
      "3   (63118.6, 84157.8] -14.0\n",
      "4  (84157.8, 105197.0]  29.0\n",
      "逾期90天笔数\n",
      "                     分组    得分\n",
      "0     (0.999, 21040.2] -11.0\n",
      "1   (21040.2, 42079.4]  -9.0\n",
      "2   (42079.4, 63118.6] -11.0\n",
      "3   (63118.6, 84157.8]  -9.0\n",
      "4  (84157.8, 105197.0]  23.0\n",
      "逾期60-89天笔数\n",
      "                     分组   得分\n",
      "0     (0.999, 21040.2]  0.0\n",
      "1   (21040.2, 42079.4]  0.0\n",
      "2   (42079.4, 63118.6]  0.0\n",
      "3   (63118.6, 84157.8]  0.0\n",
      "4  (84157.8, 105197.0] -0.0\n"
     ]
    }
   ],
   "source": [
    "def get_score(x,coef,B):\n",
    "    score=[]\n",
    "    for w in x.woe:\n",
    "        a=round(B*coef*w,0)#四舍五入返回整数\n",
    "        score.append(a)\n",
    "    datascore=pd.DataFrame({\"分组\":x.iloc[:,0],\"得分\":score})\n",
    "    return datascore\n",
    "\n",
    "scorex1=get_score(x1[:10],coef[0][0],B)\n",
    "scorex2=get_score(x2,coef[0][1],B)\n",
    "scorex3=get_score(x3,coef[0][2],B)\n",
    "scorex7=get_score(x7,coef[0][3],B)\n",
    "scorex9=get_score(x9,coef[0][4],B)\n",
    "print(\"可用额度比值\\n\",scorex1)\n",
    "print(\"年龄\\n\",scorex2)\n",
    "print(\"逾期30-59天笔数\\n\",scorex3)\n",
    "print(\"逾期90天笔数\\n\",scorex7)\n",
    "print(\"逾期60-89天笔数\\n\",scorex9)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 432x288 with 2 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "corr = data.corr()#计算各变量的相关性系数\n",
    "xticks = ['x0','x1','x2','x3','x4','x5','x6','x7','x8','x9','x10']#x轴标签\n",
    "yticks = list(corr.index)#y轴标签\n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(1, 1, 1)\n",
    "sns.heatmap(corr, annot=True, cmap='rainbow', ax=ax1, annot_kws={'size': 8, 'weight': 'bold', 'color': 'blue'})#绘制相关性系数热力图\n",
    "ax1.set_xticklabels(xticks, rotation=0, fontsize=10)\n",
    "ax1.set_yticklabels(yticks, rotation=0, fontsize=10)\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}