data_exploration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:24.804710Z",
     "start_time": "2020-08-17T19:15:24.645368Z"
    }
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:25.513456Z",
     "start_time": "2020-08-17T19:15:25.355717Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data.csv  test.csv  train.csv\r\n"
     ]
    }
   ],
   "source": [
    "!ls data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:26.074842Z",
     "start_time": "2020-08-17T19:15:25.887297Z"
    }
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"data/data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:26.351538Z",
     "start_time": "2020-08-17T19:15:26.340421Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(150000, 12)"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:28.722309Z",
     "start_time": "2020-08-17T19:15:28.699861Z"
    }
   },
   "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>Unnamed: 0</th>\n",
       "      <th>SeriousDlqin2yrs</th>\n",
       "      <th>RevolvingUtilizationOfUnsecuredLines</th>\n",
       "      <th>age</th>\n",
       "      <th>NumberOfTime30-59DaysPastDueNotWorse</th>\n",
       "      <th>DebtRatio</th>\n",
       "      <th>MonthlyIncome</th>\n",
       "      <th>NumberOfOpenCreditLinesAndLoans</th>\n",
       "      <th>NumberOfTimes90DaysLate</th>\n",
       "      <th>NumberRealEstateLoansOrLines</th>\n",
       "      <th>NumberOfTime60-89DaysPastDueNotWorse</th>\n",
       "      <th>NumberOfDependents</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.766127</td>\n",
       "      <td>45</td>\n",
       "      <td>2</td>\n",
       "      <td>0.802982</td>\n",
       "      <td>9120.0</td>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.957151</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "      <td>0.121876</td>\n",
       "      <td>2600.0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0.658180</td>\n",
       "      <td>38</td>\n",
       "      <td>1</td>\n",
       "      <td>0.085113</td>\n",
       "      <td>3042.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0.233810</td>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "      <td>0.036050</td>\n",
       "      <td>3300.0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0.907239</td>\n",
       "      <td>49</td>\n",
       "      <td>1</td>\n",
       "      <td>0.024926</td>\n",
       "      <td>63588.0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  SeriousDlqin2yrs  RevolvingUtilizationOfUnsecuredLines  age  \\\n",
       "0           1                 1                              0.766127   45   \n",
       "1           2                 0                              0.957151   40   \n",
       "2           3                 0                              0.658180   38   \n",
       "3           4                 0                              0.233810   30   \n",
       "4           5                 0                              0.907239   49   \n",
       "\n",
       "   NumberOfTime30-59DaysPastDueNotWorse  DebtRatio  MonthlyIncome  \\\n",
       "0                                     2   0.802982         9120.0   \n",
       "1                                     0   0.121876         2600.0   \n",
       "2                                     1   0.085113         3042.0   \n",
       "3                                     0   0.036050         3300.0   \n",
       "4                                     1   0.024926        63588.0   \n",
       "\n",
       "   NumberOfOpenCreditLinesAndLoans  NumberOfTimes90DaysLate  \\\n",
       "0                               13                        0   \n",
       "1                                4                        0   \n",
       "2                                2                        1   \n",
       "3                                5                        0   \n",
       "4                                7                        0   \n",
       "\n",
       "   NumberRealEstateLoansOrLines  NumberOfTime60-89DaysPastDueNotWorse  \\\n",
       "0                             6                                     0   \n",
       "1                             0                                     0   \n",
       "2                             0                                     0   \n",
       "3                             0                                     0   \n",
       "4                             1                                     0   \n",
       "\n",
       "   NumberOfDependents  \n",
       "0                 2.0  \n",
       "1                 1.0  \n",
       "2                 0.0  \n",
       "3                 0.0  \n",
       "4                 0.0  "
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:31.401024Z",
     "start_time": "2020-08-17T19:15:31.387673Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unnamed: 0                                  0\n",
       "SeriousDlqin2yrs                            0\n",
       "RevolvingUtilizationOfUnsecuredLines        0\n",
       "age                                         0\n",
       "NumberOfTime30-59DaysPastDueNotWorse        0\n",
       "DebtRatio                                   0\n",
       "MonthlyIncome                           29731\n",
       "NumberOfOpenCreditLinesAndLoans             0\n",
       "NumberOfTimes90DaysLate                     0\n",
       "NumberRealEstateLoansOrLines                0\n",
       "NumberOfTime60-89DaysPastDueNotWorse        0\n",
       "NumberOfDependents                       3924\n",
       "dtype: int64"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:37.262768Z",
     "start_time": "2020-08-17T19:15:32.505066Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x1221d8390>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x1225b64d0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x1225b4b50>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x122272210>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x1222ad890>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x122295f10>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x1227145d0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x1221a9c50>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x121fa3350>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x1220959d0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x122673050>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x1220e06d0>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.hist(bins=50, figsize=(15, 10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Unbalanced dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:37.275584Z",
     "start_time": "2020-08-17T19:15:37.265443Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.06684"
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.SeriousDlqin2yrs.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## DebtRatio can either be very small (<1) or very large (>>1):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:37.298180Z",
     "start_time": "2020-08-17T19:15:37.281454Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7642266666666667"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df.DebtRatio < 1).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:37.644748Z",
     "start_time": "2020-08-17T19:15:37.302632Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1223a3890>"
      ]
     },
     "execution_count": 154,
     "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": [
    "df[df.DebtRatio < 1].DebtRatio.hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:38.238724Z",
     "start_time": "2020-08-17T19:15:37.879561Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x122d56f90>"
      ]
     },
     "execution_count": 155,
     "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": [
    "df[(df.DebtRatio > 1) & (df.DebtRatio<100)].DebtRatio.hist(bins=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MonthlyIncome is mostly missing when DebtRatio > 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:40.098323Z",
     "start_time": "2020-08-17T19:15:40.087793Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.19820666666666667"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.MonthlyIncome.isnull().mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:40.567088Z",
     "start_time": "2020-08-17T19:15:40.548302Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.014149379765165658"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(df.DebtRatio<1)].MonthlyIncome.isnull().mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:41.041851Z",
     "start_time": "2020-08-17T19:15:41.029245Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7941486182656459"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[(df.DebtRatio>1)].MonthlyIncome.isnull().mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prepare data for ML"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:43.644332Z",
     "start_time": "2020-08-17T19:15:43.636648Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn.base import BaseEstimator, TransformerMixin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:43.958983Z",
     "start_time": "2020-08-17T19:15:43.954301Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn_pandas import DataFrameMapper"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imputer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:44.792282Z",
     "start_time": "2020-08-17T19:15:44.781719Z"
    }
   },
   "outputs": [],
   "source": [
    "class DFImputer(BaseEstimator, TransformerMixin):\n",
    "    def __init__(self, strategy=\"median\", fill_value=None):\n",
    "        self.imputer = SimpleImputer(strategy=strategy, fill_value=fill_value)\n",
    "        self.fitted = False\n",
    "    \n",
    "    def fit(self, X, y=None):\n",
    "        self._feature_names = X.columns\n",
    "        self.imputer.fit(X)\n",
    "        self.fitted = True\n",
    "        return self\n",
    "    \n",
    "    def transform(self, X):\n",
    "        assert self.fitted, \"Need to cal .fit(X) function first!\"\n",
    "        return pd.DataFrame(self.imputer.transform(X[self._feature_names]), \n",
    "                            columns=self._feature_names, index=X.index)\n",
    "    \n",
    "    def get_feature_names(self):\n",
    "        return self._feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:46.407896Z",
     "start_time": "2020-08-17T19:15:46.399051Z"
    }
   },
   "outputs": [],
   "source": [
    "transformer = DataFrameMapper([\n",
    "    (['MonthlyIncome'], DFImputer(strategy=\"constant\", fill_value=-1)),\n",
    "    (['age'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberOfDependents'], DFImputer(strategy=\"median\")),\n",
    "    (['DebtRatio'], DFImputer(strategy=\"median\")),\n",
    "    (['RevolvingUtilizationOfUnsecuredLines'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberRealEstateLoansOrLines'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberOfOpenCreditLinesAndLoans'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberOfTime30-59DaysPastDueNotWorse'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberOfTime60-89DaysPastDueNotWorse'], DFImputer(strategy=\"median\")),\n",
    "    (['NumberOfTimes90DaysLate'], DFImputer(strategy=\"median\")),   \n",
    "], input_df=True, df_out=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## train-test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:49.287481Z",
     "start_time": "2020-08-17T19:15:49.229622Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "train_df, test_df = train_test_split(df, test_size=0.1, random_state=42)\n",
    "\n",
    "TARGET = 'SeriousDlqin2yrs'\n",
    "X_train = train_df.drop(TARGET, axis=1)\n",
    "y_train = train_df[TARGET]\n",
    "X_test = test_df.drop(TARGET, axis=1)\n",
    "y_test = test_df[TARGET]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## build transformer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:15:51.302675Z",
     "start_time": "2020-08-17T19:15:51.060739Z"
    }
   },
   "outputs": [],
   "source": [
    "transformer.fit(X_train)\n",
    "\n",
    "X_train = transformer.transform(X_train)\n",
    "X_test = transformer.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:17:32.333792Z",
     "start_time": "2020-08-17T19:17:32.328997Z"
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score, roc_auc_score, precision_score, recall_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:17:37.486017Z",
     "start_time": "2020-08-17T19:17:33.010603Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:  0.9392666666666667\n",
      "roc auc: 0.8576604536833953\n",
      "precision:  0.6288659793814433\n",
      "recall:  0.12695109261186263\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestClassifier(n_estimators=50, max_depth=6, max_leaf_nodes=30)\n",
    "model.fit(X_train, y_train)\n",
    "y_preds = model.predict(X_test)\n",
    "y_pred_probas = model.predict_proba(X_test)[:,1]\n",
    "\n",
    "print(\"accuracy: \", accuracy_score(y_test, y_preds))\n",
    "print(\"roc auc:\", roc_auc_score(y_test, y_pred_probas))\n",
    "print(\"precision: \", precision_score(y_test, y_preds))\n",
    "print(\"recall: \", recall_score(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:18:26.477645Z",
     "start_time": "2020-08-17T19:17:53.264527Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:  0.9394666666666667\n",
      "roc auc: 0.8595484231195112\n",
      "precision:  0.5898305084745763\n",
      "recall:  0.18106139438085328\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestClassifier(max_features=2, min_samples_leaf=4, min_samples_split=10, n_estimators=200)\n",
    "model.fit(X_train, y_train)\n",
    "y_preds = model.predict(X_test)\n",
    "y_pred_probas = model.predict_proba(X_test)[:,1]\n",
    "\n",
    "print(\"accuracy: \", accuracy_score(y_test, y_preds))\n",
    "print(\"roc auc:\", roc_auc_score(y_test, y_pred_probas))\n",
    "print(\"precision: \", precision_score(y_test, y_preds))\n",
    "print(\"recall: \", recall_score(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T14:09:22.833213Z",
     "start_time": "2020-08-17T14:09:22.805297Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:  0.9398666666666666\n",
      "roc auc: 0.8593324719995489\n",
      "precision:  0.6006825938566553\n",
      "recall:  0.18314255983350677\n"
     ]
    }
   ],
   "source": [
    "print(\"accuracy: \", accuracy_score(y_test, y_preds))\n",
    "print(\"roc auc:\", roc_auc_score(y_test, y_pred_probas))\n",
    "print(\"precision: \", precision_score(y_test, y_preds))\n",
    "print(\"recall: \", recall_score(y_test, y_preds))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Start explainerdashboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:21:30.332217Z",
     "start_time": "2020-08-17T19:21:29.033288Z"
    }
   },
   "outputs": [],
   "source": [
    "from explainerdashboard.explainers import *\n",
    "from explainerdashboard.dashboards import *\n",
    "from explainerdashboard.dashboard_tabs import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:21:31.430292Z",
     "start_time": "2020-08-17T19:21:31.186778Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Note: shap=='guess' so guessing for RandomForestClassifier shap='tree'...\n",
      "Note: model_output=='probability', so assuming that raw shap output of RandomForestClassifier is in probability space...\n",
      "Generating self.shap_explainer = shap.TreeExplainer(model)\n",
      "Final note: You can always monkeypatch self.shap_explainer if desired...\n"
     ]
    }
   ],
   "source": [
    "explainer = ClassifierExplainer(model, X_test, y_test, labels=[\"Not delinquent\", \"Delinquent\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2020-08-17T19:21:35.793Z"
    }
   },
   "outputs": [],
   "source": [
    "db = ExplainerDashboard(explainer).run(8052)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Start API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-06T10:20:30.752068Z",
     "start_time": "2020-08-06T10:20:30.567775Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Setting feature_perturbation = \"tree_path_dependent\" because no background data was given.\n"
     ]
    }
   ],
   "source": [
    "import shap\n",
    "shap_explainer = shap.TreeExplainer(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-08-17T19:21:22.024627Z",
     "start_time": "2020-08-17T19:21:19.956641Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[33m * Tip: There are .env or .flaskenv files present. Do \"pip install python-dotenv\" to use them.\u001b[0m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " * Serving Flask app \"__main__\" (lazy loading)\n",
      " * Environment: production\n",
      "\u001b[31m   WARNING: This is a development server. Do not use it in a production deployment.\u001b[0m\n",
      "\u001b[2m   Use a production WSGI server instead.\u001b[0m\n",
      " * Debug mode: on\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " * Running on http://127.0.0.1:5002/ (Press CTRL+C to quit)\n"
     ]
    }
   ],
   "source": [
    "import shap\n",
    "from flask import Flask, request, abort, jsonify\n",
    "import json\n",
    "\n",
    "class NumpyEncoder(json.JSONEncoder):\n",
    "    def default(self, obj):\n",
    "        if isinstance(obj, np.ndarray):\n",
    "            return obj.tolist()\n",
    "        return json.JSONEncoder.default(self, obj)\n",
    "\n",
    "flask_app = Flask(__name__)\n",
    "positive_label = 1\n",
    "shap_explainer = shap.TreeExplainer(model)\n",
    "shap_base = shap_explainer.expected_value[1]\n",
    "\n",
    "@flask_app.route('/predict', methods=['Post'])\n",
    "def prediction_server_API_call():\n",
    "    try:\n",
    "        json_str = request.get_json()\n",
    "        df = pd.read_json(json_str, orient='records')\n",
    "    except Exception as e:\n",
    "        raise e\n",
    "\n",
    "    if df.empty:\n",
    "        return(abort(400))\n",
    "    else:\n",
    "        features = transformer.transform(df)\n",
    "        shap_values = np.transpose(shap_explainer.shap_values(features)[positive_label])\n",
    "        preds = model.predict_proba(features)[:, positive_label]\n",
    "        response = {}\n",
    "        response['prediction'] = preds\n",
    "        response['shap_base'] = shap_base\n",
    "        response['shap_values'] = {k: v for k, v in zip(features.columns.tolist(), shap_values.tolist())}\n",
    "        return json.dumps(response, cls=NumpyEncoder)\n",
    "    \n",
    "flask_app.run(host='127.0.0.1', port=5002, debug=True, use_reloader=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7.7 64-bit ('venv': venv)",
   "language": "python",
   "name": "python37764bitvenvvenv081c98e452e74ebf9017b51b5bf1890c"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}