kernel-1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<H2> Preparing the data first... </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load in \n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "import os\n",
    "from keras import regularizers\n",
    "import tensorflow as tf\n",
    "from sklearn import preprocessing\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras.optimizers import SGD, Adam\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint\n",
    "from keras.callbacks import Callback, EarlyStopping\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "# roc curve and auc score\n",
    "from sklearn import preprocessing\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/kaggle/input/widsdatathon2020/unlabeled.csv\n",
      "/kaggle/input/widsdatathon2020/training_v2.csv\n",
      "/kaggle/input/widsdatathon2020/solution_template.csv\n",
      "/kaggle/input/widsdatathon2020/WiDS Datathon 2020 Dictionary.csv\n",
      "/kaggle/input/widsdatathon2020/samplesubmission.csv\n"
     ]
    }
   ],
   "source": [
    "# Input data files are available in the \"../input/\" directory.\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n",
    "\n",
    "for dirname, _, filenames in os.walk('/kaggle/input'):\n",
    "    for filename in filenames:\n",
    "        print(os.path.join(dirname, filename))\n",
    "\n",
    "trainData = pd.read_csv(\"../input/widsdatathon2020/training_v2.csv\")\n",
    "testData = pd.read_csv(\"../input/widsdatathon2020/unlabeled.csv\")\n",
    "\n",
    "# Any results you write to the current directory are saved as output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
    "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a"
   },
   "outputs": [],
   "source": [
    "categorical_cols = [c for c in trainData.columns if (trainData[c].dtype != np.number)& (trainData[c].dtype != int) ]\n",
    "Categorical_df= trainData[categorical_cols]\n",
    "# for col in categorical_cols:\n",
    "#     print(col, \"*****************\")\n",
    "#     print(Categorical_df[col].value_counts(), Categorical_df[col].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>encounter_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>hospital_id</th>\n",
       "      <th>hospital_death</th>\n",
       "      <th>age</th>\n",
       "      <th>bmi</th>\n",
       "      <th>elective_surgery</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>height</th>\n",
       "      <th>...</th>\n",
       "      <th>aids</th>\n",
       "      <th>cirrhosis</th>\n",
       "      <th>diabetes_mellitus</th>\n",
       "      <th>hepatic_failure</th>\n",
       "      <th>immunosuppression</th>\n",
       "      <th>leukemia</th>\n",
       "      <th>lymphoma</th>\n",
       "      <th>solid_tumor_with_metastasis</th>\n",
       "      <th>apache_3j_bodysystem</th>\n",
       "      <th>apache_2_bodysystem</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>66154</td>\n",
       "      <td>25312</td>\n",
       "      <td>118</td>\n",
       "      <td>0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>22.73</td>\n",
       "      <td>0</td>\n",
       "      <td>0.770709</td>\n",
       "      <td>M</td>\n",
       "      <td>180.3</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.128008</td>\n",
       "      <td>0.423233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>114252</td>\n",
       "      <td>59342</td>\n",
       "      <td>81</td>\n",
       "      <td>0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>27.42</td>\n",
       "      <td>0</td>\n",
       "      <td>0.770709</td>\n",
       "      <td>F</td>\n",
       "      <td>160.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.126580</td>\n",
       "      <td>0.126580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>119783</td>\n",
       "      <td>50777</td>\n",
       "      <td>118</td>\n",
       "      <td>0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>31.95</td>\n",
       "      <td>0</td>\n",
       "      <td>0.770709</td>\n",
       "      <td>F</td>\n",
       "      <td>172.7</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.083412</td>\n",
       "      <td>0.083412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>79267</td>\n",
       "      <td>46918</td>\n",
       "      <td>118</td>\n",
       "      <td>0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>22.64</td>\n",
       "      <td>1</td>\n",
       "      <td>0.770709</td>\n",
       "      <td>F</td>\n",
       "      <td>165.1</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.327096</td>\n",
       "      <td>0.423233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>92056</td>\n",
       "      <td>34377</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.770709</td>\n",
       "      <td>M</td>\n",
       "      <td>188.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.041892</td>\n",
       "      <td>0.041892</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 186 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   encounter_id  patient_id  hospital_id  hospital_death   age    bmi  \\\n",
       "0         66154       25312          118               0  68.0  22.73   \n",
       "1        114252       59342           81               0  77.0  27.42   \n",
       "2        119783       50777          118               0  25.0  31.95   \n",
       "3         79267       46918          118               0  81.0  22.64   \n",
       "4         92056       34377           33               0  19.0    NaN   \n",
       "\n",
       "   elective_surgery  ethnicity gender  height  ...  aids  cirrhosis  \\\n",
       "0                 0   0.770709      M   180.3  ...   0.0        0.0   \n",
       "1                 0   0.770709      F   160.0  ...   0.0        0.0   \n",
       "2                 0   0.770709      F   172.7  ...   0.0        0.0   \n",
       "3                 1   0.770709      F   165.1  ...   0.0        0.0   \n",
       "4                 0   0.770709      M   188.0  ...   0.0        0.0   \n",
       "\n",
       "   diabetes_mellitus hepatic_failure  immunosuppression  leukemia  lymphoma  \\\n",
       "0                1.0             0.0                0.0       0.0       0.0   \n",
       "1                1.0             0.0                0.0       0.0       0.0   \n",
       "2                0.0             0.0                0.0       0.0       0.0   \n",
       "3                0.0             0.0                0.0       0.0       0.0   \n",
       "4                0.0             0.0                0.0       0.0       0.0   \n",
       "\n",
       "   solid_tumor_with_metastasis  apache_3j_bodysystem  apache_2_bodysystem  \n",
       "0                          0.0              0.128008             0.423233  \n",
       "1                          0.0              0.126580             0.126580  \n",
       "2                          0.0              0.083412             0.083412  \n",
       "3                          0.0              0.327096             0.423233  \n",
       "4                          0.0              0.041892             0.041892  \n",
       "\n",
       "[5 rows x 186 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# frequency encoder\n",
    "for col in ['ethnicity', 'hospital_admit_source', 'icu_admit_source', 'apache_3j_bodysystem', 'apache_2_bodysystem', 'icu_type']:\n",
    "    trainData[col] = trainData[col].astype('str')\n",
    "    freq = trainData.groupby(col).size()/len(trainData[col])\n",
    "    trainData[col] = trainData[col].map(freq)\n",
    "    \n",
    "    testData[col] = testData[col].astype('str')\n",
    "    freq = testData.groupby(col).size()/len(testData[col])\n",
    "    testData[col] = testData[col].map(freq)\n",
    "\n",
    "trainData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ordinal encoder\n",
    "icu_st_dict={'admit':0,'readmit':1,'transfer':2}\n",
    "\n",
    "trainData['icu_stay_type'] = trainData['icu_stay_type'].astype('str')\n",
    "trainData['icu_stay_type'] = trainData['icu_stay_type'].map(icu_st_dict)\n",
    "\n",
    "testData['icu_stay_type'] = testData['icu_stay_type'].astype('str')\n",
    "testData['icu_stay_type'] = testData['icu_stay_type'].map(icu_st_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "91703    0.551568\n",
       "91704    0.052075\n",
       "91705    0.551568\n",
       "91706    0.551568\n",
       "91707    0.551568\n",
       "91708    0.052075\n",
       "91709    0.551568\n",
       "91710    0.551568\n",
       "91711    0.551568\n",
       "91712    0.551568\n",
       "Name: icu_type, dtype: float64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainData['icu_stay_type'].tail(10)\n",
    "trainData['icu_type'].tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>encounter_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>hospital_id</th>\n",
       "      <th>hospital_death</th>\n",
       "      <th>age</th>\n",
       "      <th>bmi</th>\n",
       "      <th>elective_surgery</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>height</th>\n",
       "      <th>...</th>\n",
       "      <th>aids</th>\n",
       "      <th>cirrhosis</th>\n",
       "      <th>diabetes_mellitus</th>\n",
       "      <th>hepatic_failure</th>\n",
       "      <th>immunosuppression</th>\n",
       "      <th>leukemia</th>\n",
       "      <th>lymphoma</th>\n",
       "      <th>solid_tumor_with_metastasis</th>\n",
       "      <th>apache_3j_bodysystem</th>\n",
       "      <th>apache_2_bodysystem</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>127112</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>56.0</td>\n",
       "      <td>21.102439</td>\n",
       "      <td>0</td>\n",
       "      <td>0.769309</td>\n",
       "      <td>2</td>\n",
       "      <td>180.3</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.152361</td>\n",
       "      <td>0.152361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>130737</td>\n",
       "      <td>178</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19.500595</td>\n",
       "      <td>1</td>\n",
       "      <td>0.769309</td>\n",
       "      <td>1</td>\n",
       "      <td>145.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.152361</td>\n",
       "      <td>0.152361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>101458</td>\n",
       "      <td>86</td>\n",
       "      <td>NaN</td>\n",
       "      <td>66.0</td>\n",
       "      <td>32.518596</td>\n",
       "      <td>1</td>\n",
       "      <td>0.113743</td>\n",
       "      <td>2</td>\n",
       "      <td>170.2</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.289839</td>\n",
       "      <td>0.377251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8</td>\n",
       "      <td>48067</td>\n",
       "      <td>163</td>\n",
       "      <td>NaN</td>\n",
       "      <td>64.0</td>\n",
       "      <td>20.798520</td>\n",
       "      <td>0</td>\n",
       "      <td>0.769309</td>\n",
       "      <td>2</td>\n",
       "      <td>182.8</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.134426</td>\n",
       "      <td>0.134426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>25865</td>\n",
       "      <td>59</td>\n",
       "      <td>NaN</td>\n",
       "      <td>56.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.769309</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.152361</td>\n",
       "      <td>0.152361</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 186 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   encounter_id  patient_id  hospital_id  hospital_death   age        bmi  \\\n",
       "0             2      127112            7             NaN  56.0  21.102439   \n",
       "1             5      130737          178             NaN   NaN  19.500595   \n",
       "2             7      101458           86             NaN  66.0  32.518596   \n",
       "3             8       48067          163             NaN  64.0  20.798520   \n",
       "4            10       25865           59             NaN  56.0        NaN   \n",
       "\n",
       "   elective_surgery  ethnicity  gender  height  ...  aids  cirrhosis  \\\n",
       "0                 0   0.769309       2   180.3  ...   0.0        0.0   \n",
       "1                 1   0.769309       1   145.0  ...   0.0        0.0   \n",
       "2                 1   0.113743       2   170.2  ...   0.0        0.0   \n",
       "3                 0   0.769309       2   182.8  ...   0.0        0.0   \n",
       "4                 0   0.769309       2     NaN  ...   0.0        0.0   \n",
       "\n",
       "   diabetes_mellitus  hepatic_failure  immunosuppression  leukemia  lymphoma  \\\n",
       "0                0.0              0.0                0.0       0.0       0.0   \n",
       "1                0.0              0.0                0.0       0.0       0.0   \n",
       "2                0.0              0.0                0.0       0.0       0.0   \n",
       "3                0.0              0.0                1.0       0.0       0.0   \n",
       "4                0.0              0.0                0.0       0.0       0.0   \n",
       "\n",
       "   solid_tumor_with_metastasis  apache_3j_bodysystem  apache_2_bodysystem  \n",
       "0                          0.0              0.152361             0.152361  \n",
       "1                          0.0              0.152361             0.152361  \n",
       "2                          0.0              0.289839             0.377251  \n",
       "3                          0.0              0.134426             0.134426  \n",
       "4                          0.0              0.152361             0.152361  \n",
       "\n",
       "[5 rows x 186 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# label encoding the data gender\n",
    "le = LabelEncoder()\n",
    "for col in ['gender']:\n",
    "    trainData[col] = trainData[col].astype('str')\n",
    "\n",
    "    #Fit LabelEncoder\n",
    "    le.fit(np.unique(trainData[col].unique()))\n",
    "\n",
    "    #At the end 0 will be used for null values so we start at 1 \n",
    "    trainData[col] = le.transform(trainData[col])+1\n",
    "    trainData[col] = trainData[col].replace(np.nan, 0).astype('int')\n",
    "    \n",
    "    testData[col] = testData[col].astype('str')\n",
    "\n",
    "    #Fit LabelEncoder\n",
    "    le.fit(np.unique(testData[col].unique()))\n",
    "\n",
    "    #At the end 0 will be used for null values so we start at 1 \n",
    "    testData[col] = le.transform(testData[col])+1\n",
    "    testData[col] = testData[col].replace(np.nan, 0).astype('int')\n",
    "\n",
    "testData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# replace na with mean for following categories\n",
    "\n",
    "for col in ['age', 'bmi', 'weight', 'height']:\n",
    "    mean = trainData[col].mean()\n",
    "    trainData[col] = trainData[col].replace(np.nan, mean).astype('int')\n",
    "    \n",
    "    mean = testData[col].mean()\n",
    "    testData[col] = testData[col].replace(np.nan, mean).astype('int')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "encounter_id                       0\n",
       "patient_id                         0\n",
       "hospital_id                        0\n",
       "hospital_death                 39308\n",
       "age                                0\n",
       "                               ...  \n",
       "leukemia                         179\n",
       "lymphoma                         179\n",
       "solid_tumor_with_metastasis      179\n",
       "apache_3j_bodysystem               0\n",
       "apache_2_bodysystem                0\n",
       "Length: 186, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = testData.isnull().sum(axis=0)\n",
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>encounter_id</th>\n",
       "      <th>patient_id</th>\n",
       "      <th>hospital_id</th>\n",
       "      <th>hospital_death</th>\n",
       "      <th>age</th>\n",
       "      <th>bmi</th>\n",
       "      <th>elective_surgery</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>gender</th>\n",
       "      <th>height</th>\n",
       "      <th>...</th>\n",
       "      <th>aids</th>\n",
       "      <th>cirrhosis</th>\n",
       "      <th>diabetes_mellitus</th>\n",
       "      <th>hepatic_failure</th>\n",
       "      <th>immunosuppression</th>\n",
       "      <th>leukemia</th>\n",
       "      <th>lymphoma</th>\n",
       "      <th>solid_tumor_with_metastasis</th>\n",
       "      <th>apache_3j_bodysystem</th>\n",
       "      <th>apache_2_bodysystem</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>127112</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>56</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>180</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>130737</td>\n",
       "      <td>178</td>\n",
       "      <td>0</td>\n",
       "      <td>61</td>\n",
       "      <td>19</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>145</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>101458</td>\n",
       "      <td>86</td>\n",
       "      <td>0</td>\n",
       "      <td>66</td>\n",
       "      <td>32</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>170</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8</td>\n",
       "      <td>48067</td>\n",
       "      <td>163</td>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>20</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>182</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>25865</td>\n",
       "      <td>59</td>\n",
       "      <td>0</td>\n",
       "      <td>56</td>\n",
       "      <td>28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>169</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 186 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   encounter_id  patient_id  hospital_id  hospital_death  age  bmi  \\\n",
       "0             2      127112            7               0   56   21   \n",
       "1             5      130737          178               0   61   19   \n",
       "2             7      101458           86               0   66   32   \n",
       "3             8       48067          163               0   64   20   \n",
       "4            10       25865           59               0   56   28   \n",
       "\n",
       "   elective_surgery  ethnicity  gender  height  ...  aids  cirrhosis  \\\n",
       "0                 0          0       2     180  ...     0          0   \n",
       "1                 1          0       1     145  ...     0          0   \n",
       "2                 1          0       2     170  ...     0          0   \n",
       "3                 0          0       2     182  ...     0          0   \n",
       "4                 0          0       2     169  ...     0          0   \n",
       "\n",
       "   diabetes_mellitus  hepatic_failure  immunosuppression  leukemia  lymphoma  \\\n",
       "0                  0                0                  0         0         0   \n",
       "1                  0                0                  0         0         0   \n",
       "2                  0                0                  0         0         0   \n",
       "3                  0                0                  1         0         0   \n",
       "4                  0                0                  0         0         0   \n",
       "\n",
       "   solid_tumor_with_metastasis  apache_3j_bodysystem  apache_2_bodysystem  \n",
       "0                            0                     0                    0  \n",
       "1                            0                     0                    0  \n",
       "2                            0                     0                    0  \n",
       "3                            0                     0                    0  \n",
       "4                            0                     0                    0  \n",
       "\n",
       "[5 rows x 186 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainData = trainData.replace(np.nan, 0).astype('int')\n",
    "testData = testData.replace(np.nan, 0).astype('int')\n",
    "testData.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2> Training the data now... </h2>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>age</th>\n",
       "      <th>bmi</th>\n",
       "      <th>elective_surgery</th>\n",
       "      <th>height</th>\n",
       "      <th>hospital_admit_source</th>\n",
       "      <th>icu_admit_source</th>\n",
       "      <th>icu_id</th>\n",
       "      <th>icu_stay_type</th>\n",
       "      <th>icu_type</th>\n",
       "      <th>pre_icu_los_days</th>\n",
       "      <th>...</th>\n",
       "      <th>aids</th>\n",
       "      <th>cirrhosis</th>\n",
       "      <th>diabetes_mellitus</th>\n",
       "      <th>hepatic_failure</th>\n",
       "      <th>immunosuppression</th>\n",
       "      <th>leukemia</th>\n",
       "      <th>lymphoma</th>\n",
       "      <th>solid_tumor_with_metastasis</th>\n",
       "      <th>apache_3j_bodysystem</th>\n",
       "      <th>apache_2_bodysystem</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>68</td>\n",
       "      <td>22</td>\n",
       "      <td>0</td>\n",
       "      <td>180</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>92</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>77</td>\n",
       "      <td>27</td>\n",
       "      <td>0</td>\n",
       "      <td>160</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>90</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>25</td>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "      <td>172</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>93</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>81</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>165</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>92</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>188</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>91</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 180 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   age  bmi  elective_surgery  height  hospital_admit_source  \\\n",
       "0   68   22                 0     180                      0   \n",
       "1   77   27                 0     160                      0   \n",
       "2   25   31                 0     172                      0   \n",
       "3   81   22                 1     165                      0   \n",
       "4   19   29                 0     188                      0   \n",
       "\n",
       "   icu_admit_source  icu_id  icu_stay_type  icu_type  pre_icu_los_days  ...  \\\n",
       "0                 0      92              0         0                 0  ...   \n",
       "1                 0      90              0         0                 0  ...   \n",
       "2                 0      93              0         0                 0  ...   \n",
       "3                 0      92              0         0                 0  ...   \n",
       "4                 0      91              0         0                 0  ...   \n",
       "\n",
       "   aids  cirrhosis  diabetes_mellitus  hepatic_failure  immunosuppression  \\\n",
       "0     0          0                  1                0                  0   \n",
       "1     0          0                  1                0                  0   \n",
       "2     0          0                  0                0                  0   \n",
       "3     0          0                  0                0                  0   \n",
       "4     0          0                  0                0                  0   \n",
       "\n",
       "   leukemia  lymphoma  solid_tumor_with_metastasis  apache_3j_bodysystem  \\\n",
       "0         0         0                            0                     0   \n",
       "1         0         0                            0                     0   \n",
       "2         0         0                            0                     0   \n",
       "3         0         0                            0                     0   \n",
       "4         0         0                            0                     0   \n",
       "\n",
       "   apache_2_bodysystem  \n",
       "0                    0  \n",
       "1                    0  \n",
       "2                    0  \n",
       "3                    0  \n",
       "4                    0  \n",
       "\n",
       "[5 rows x 180 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "to_drop = ['gender','ethnicity' ,'encounter_id', 'patient_id',  'hospital_death', 'hospital_id']\n",
    "testDataOld = testData\n",
    "trainLabel = trainData['hospital_death']\n",
    "for col in to_drop:\n",
    "    trainData = trainData.drop(col, axis = 1)\n",
    "    testData = testData.drop(col, axis = 1)\n",
    "\n",
    "trainData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'y_test' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-12-157200a9a21d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mcols_with_missing\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mcol\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mcol\u001b[0m \u001b[0;32min\u001b[0m \u001b[0my_test\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0my_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcol\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misnull\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0many\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mcol\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mcols_with_missing\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m     \u001b[0my_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcol\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m'_was_missing'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0my_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcol\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misnull\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m     \u001b[0my_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcol\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m'_was_missing'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0my_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcol\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misnull\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'y_test' is not defined"
     ]
    }
   ],
   "source": [
    "cols_with_missing = (col for col in y_test.columns if y_test[col].isnull().any())\n",
    "for col in cols_with_missing:\n",
    "    y_test[col + '_was_missing'] = y_test[col].isnull()\n",
    "    y_test[col + '_was_missing'] = y_test[col].isnull()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64199, 180)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train, y_test ,x_label, y_label = train_test_split(trainData, trainLabel, test_size=0.3, random_state=1)\n",
    "std_scale = preprocessing.StandardScaler().fit(x_train)\n",
    "x_train = std_scale.transform(x_train)\n",
    "y_test  = std_scale.transform(y_test)\n",
    "testData = std_scale.transform(testData)\n",
    "x_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.linear_model import LinearRegression\n",
    "# from sklearn.feature_selection import RFE\n",
    "# from sklearn.feature_selection import SelectKBest, SelectFpr, f_classif\n",
    "\n",
    "# reg = LinearRegression()\n",
    "# x_train_new = SelectFpr(f_classif, alpha=0.01).fit_transform(x_train, x_label)\n",
    "# fit = reg.fit(x_train_new, x_label)\n",
    "# pred = fit.predict(y_test)\n",
    "# print(testData.shape, pred.shape)\n",
    "\n",
    "# auc = roc_auc_score(y_label, pred)\n",
    "# fpr, tpr, thresholds = roc_curve(y_label, pred)\n",
    "# plot_roc_curve(fpr, tpr)\n",
    "# print(auc)\n",
    "# pred = fit.predict(testData)\n",
    "# testDataOld[\"hospital_death\"] = pred\n",
    "# testDataOld[[\"encounter_id\",\"hospital_death\"]].to_csv(\"submission.csv\",index=False)\n",
    "# testDataOld[[\"encounter_id\",\"hospital_death\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_roc_curve(fpr, tpr):\n",
    "    plt.plot(fpr, tpr, color='orange', label='ROC')\n",
    "    plt.plot([0, 1], [0, 1], color='darkblue', linestyle='--')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('Receiver Operating Characteristic (ROC) Curve')\n",
    "    plt.legend()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(27514, 180)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "checkpoint_callback = ModelCheckpoint(\"model.h5\", monitor='accuracy', save_best_only=True, save_freq=2)\n",
    "y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Dense(64, input_shape=(180,),kernel_initializer='normal', activation='sigmoid', name='fc1'))\n",
    "model.add(Dense(32, activation='sigmoid',kernel_initializer='normal', name='fc3'))\n",
    "model.add(Dense(1, name='output'))\n",
    "optimizer = tf.keras.optimizers.RMSprop(0.0001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer, loss='mse', metrics=['accuracy', 'mse'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 64199 samples\n",
      "Epoch 1/50\n",
      "64199/64199 [==============================] - 3s 43us/sample - loss: 0.0663 - accuracy: 0.9165 - mse: 0.0663\n",
      "Epoch 2/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0617 - accuracy: 0.9215 - mse: 0.0617\n",
      "Epoch 3/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0612 - accuracy: 0.9225 - mse: 0.0612\n",
      "Epoch 4/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0609 - accuracy: 0.9232 - mse: 0.0609\n",
      "Epoch 5/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0607 - accuracy: 0.9235 - mse: 0.0607\n",
      "Epoch 6/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0606 - accuracy: 0.9234 - mse: 0.0606\n",
      "Epoch 7/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0605 - accuracy: 0.9235 - mse: 0.0605\n",
      "Epoch 8/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0604 - accuracy: 0.9238 - mse: 0.0604\n",
      "Epoch 9/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0603 - accuracy: 0.9237 - mse: 0.0603\n",
      "Epoch 10/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0601 - accuracy: 0.9240 - mse: 0.0601\n",
      "Epoch 11/50\n",
      "64199/64199 [==============================] - 2s 34us/sample - loss: 0.0601 - accuracy: 0.9237 - mse: 0.0601\n",
      "Epoch 12/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0600 - accuracy: 0.9239 - mse: 0.0600\n",
      "Epoch 13/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0599 - accuracy: 0.9243 - mse: 0.0599\n",
      "Epoch 14/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0598 - accuracy: 0.9240 - mse: 0.0598\n",
      "Epoch 15/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0597 - accuracy: 0.9241 - mse: 0.0597\n",
      "Epoch 16/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0597 - accuracy: 0.9240 - mse: 0.0597\n",
      "Epoch 17/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0596 - accuracy: 0.9244 - mse: 0.0596\n",
      "Epoch 18/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0595 - accuracy: 0.9243 - mse: 0.0595\n",
      "Epoch 19/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0595 - accuracy: 0.9245 - mse: 0.0595\n",
      "Epoch 20/50\n",
      "64199/64199 [==============================] - 2s 34us/sample - loss: 0.0593 - accuracy: 0.9245 - mse: 0.0593\n",
      "Epoch 21/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0593 - accuracy: 0.9245 - mse: 0.0593\n",
      "Epoch 22/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0592 - accuracy: 0.9248 - mse: 0.0592\n",
      "Epoch 23/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0591 - accuracy: 0.9249 - mse: 0.0591\n",
      "Epoch 24/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0590 - accuracy: 0.9249 - mse: 0.0590\n",
      "Epoch 25/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0589 - accuracy: 0.9247 - mse: 0.0589\n",
      "Epoch 26/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0588 - accuracy: 0.9249 - mse: 0.0588\n",
      "Epoch 27/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0588 - accuracy: 0.9249 - mse: 0.0588\n",
      "Epoch 28/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0586 - accuracy: 0.9251 - mse: 0.0586\n",
      "Epoch 29/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0585 - accuracy: 0.9255 - mse: 0.0585\n",
      "Epoch 30/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0584 - accuracy: 0.9256 - mse: 0.0584\n",
      "Epoch 31/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0583 - accuracy: 0.9256 - mse: 0.0583\n",
      "Epoch 32/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0581 - accuracy: 0.9260 - mse: 0.0581\n",
      "Epoch 33/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0580 - accuracy: 0.9257 - mse: 0.0580\n",
      "Epoch 34/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0578 - accuracy: 0.9263 - mse: 0.0578\n",
      "Epoch 35/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0576 - accuracy: 0.9261 - mse: 0.0576\n",
      "Epoch 36/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0575 - accuracy: 0.9263 - mse: 0.0575\n",
      "Epoch 37/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0573 - accuracy: 0.9267 - mse: 0.0573\n",
      "Epoch 38/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0572 - accuracy: 0.9263 - mse: 0.0572\n",
      "Epoch 39/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0570 - accuracy: 0.9269 - mse: 0.0570\n",
      "Epoch 40/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0569 - accuracy: 0.9267 - mse: 0.0569\n",
      "Epoch 41/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0568 - accuracy: 0.9271 - mse: 0.0568\n",
      "Epoch 42/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0567 - accuracy: 0.9274 - mse: 0.0567\n",
      "Epoch 43/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0566 - accuracy: 0.9272 - mse: 0.0566\n",
      "Epoch 44/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0565 - accuracy: 0.9273 - mse: 0.0565\n",
      "Epoch 45/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0564 - accuracy: 0.9273 - mse: 0.0564\n",
      "Epoch 46/50\n",
      "64199/64199 [==============================] - 2s 32us/sample - loss: 0.0563 - accuracy: 0.9278 - mse: 0.0563\n",
      "Epoch 47/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0562 - accuracy: 0.9278 - mse: 0.0562\n",
      "Epoch 48/50\n",
      "64199/64199 [==============================] - 2s 31us/sample - loss: 0.0560 - accuracy: 0.9282 - mse: 0.0560\n",
      "Epoch 49/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0559 - accuracy: 0.9278 - mse: 0.0559\n",
      "Epoch 50/50\n",
      "64199/64199 [==============================] - 2s 33us/sample - loss: 0.0559 - accuracy: 0.9280 - mse: 0.0559\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7f7d582637f0>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x_train, x_label, batch_size=60, epochs=50, callbacks=[checkpoint_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC-ROC : 0.8721801432424632\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0.01159199, 0.07404302, 0.05196293, ..., 0.06427459, 0.0302742 ,\n",
       "       0.05582491], dtype=float32)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "probs = model.predict_proba(y_test).flatten()\n",
    "auc = roc_auc_score(y_label, probs)\n",
    "fpr, tpr, thresholds = roc_curve(y_label, probs)\n",
    "plot_roc_curve(fpr, tpr)\n",
    "print(\"AUC-ROC :\",auc)\n",
    "probs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.05391884]\n",
      " [-0.0076026 ]\n",
      " [ 0.05280583]\n",
      " ...\n",
      " [ 0.11069826]\n",
      " [ 0.0492777 ]\n",
      " [ 0.12105015]]\n"
     ]
    },
    {
     "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>encounter_id</th>\n",
       "      <th>hospital_death</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>0.053919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>-0.007603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>0.052806</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8</td>\n",
       "      <td>0.229574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>0.164922</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   encounter_id  hospital_death\n",
       "0             2        0.053919\n",
       "1             5       -0.007603\n",
       "2             7        0.052806\n",
       "3             8        0.229574\n",
       "4            10        0.164922"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "probstest = model.predict_proba(testData)\n",
    "probstest = probstest[:]\n",
    "print(probstest)\n",
    "testDataOld[\"hospital_death\"] = probstest\n",
    "testDataOld[[\"encounter_id\",\"hospital_death\"]].to_csv(\"submission3.csv\",index=False)\n",
    "testDataOld[[\"encounter_id\",\"hospital_death\"]].head()\n",
    "testDataOld[[\"encounter_id\",\"hospital_death\"]].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='submission3.csv' target='_blank'>submission3.csv</a><br>"
      ],
      "text/plain": [
       "/kaggle/working/submission3.csv"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import FileLink\n",
    "import os\n",
    "os.chdir(r'/kaggle/working')\n",
    "FileLink(r'submission3.csv')"
   ]
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}