MIMIC-III_hospitalization_time_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting hospitalization times\n",
    "## Nádia Soares, Rafael Santos, Tiago Ribeiro\n",
    "\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The goal of this project is to predict the hospitalization times given the MIMIC-III dataset on hospitalization events. The data includes information about the patient, ICU, medical exams that were performed and their results, among other variables.\n",
    "\n",
    "This notebook presents an exploratory data analysis using Seaborn, model creation and evaluation using Keras with a TensorFlow backend. It was created to run on a PySpark cluster, which, in our case, was deployed on Google Cloud."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The following commands need to be executed outside the python interpreter:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* gsutil cp gs:/.../*.GZ .\n",
    "* gsutil cp gs:/.../*.CSV .\n",
    "\n",
    "* gunzip *.GZ\n",
    "\n",
    "* sudo hadoop fs -put EVENTS.csv\n",
    "* sudo hadoop fs -put ADMISSIONS.csv\n",
    "* sudo hadoop fs -put DIAGNOSES_ICD.csv\n",
    "* sudo hadoop fs -put D_ICD_DIAGNOSES.csv\n",
    "* sudo hadoop fs -put D_ITEMS.csv\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (1) Definition of basic functions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "if __name__ == \"__main__\" :\n",
    "    # This block is required to run the program from the command line\n",
    "    # in interface with a single Spark instance\n",
    "    from pyspark import SparkContext\n",
    "    from pyspark.sql import SparkSession\n",
    "    \n",
    "    spark = SparkSession\\\n",
    "        .builder\\\n",
    "        .appName(\"BDCCp2\")\\\n",
    "        .master(\"local[*]\")\\\n",
    "        .getOrCreate()\n",
    "        \n",
    "    \n",
    "    sc = spark.sparkContext\n",
    "    sc.setLogLevel(\"WARN\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "from pyspark.sql import functions as F\n",
    "from pyspark.sql.types import ArrayType,IntegerType\n",
    "import numpy as np\n",
    "import matplotlib.mlab as mlab\n",
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "\n",
    "import pathlib\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "from sklearn import preprocessing\n",
    "\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def readCSV(file, debug=False):\n",
    "    if debug:\n",
    "        print('Reading ' + file)\n",
    "    return spark.read.csv(file, inferSchema=True, header=True)\n",
    "\n",
    "def readParquet(file, debug=False): \n",
    "    if debug:\n",
    "        print('Reading ' + file)\n",
    "    return spark.read.parquet(file)\n",
    "\n",
    "def writeCSV(df, path): \n",
    "    df.write.csv(path, header=True, mode='overwrite')\n",
    "\n",
    "def writeParquet(df,path):\n",
    "    df.write.mode(\"overwrite\").parquet(path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define F.array_intersect if not defined (Spark version < 2.4)\n",
    "if not hasattr(F,'array_intersect'):\n",
    "    F.array_intersect = spark.udf\\\n",
    "        .register('array_intersect', \n",
    "           lambda x,y: list(set(x) & set(y)), ArrayType(IntegerType()))\n",
    "\n",
    "# Define F.array_union if not defined (Spark version < 2.4)\n",
    "if not hasattr(F,'array_union'):\n",
    "    F.array_union = spark.udf\\\n",
    "        .register('array_union', \n",
    "            lambda x,y: list(set(x) | set(y)), ArrayType(IntegerType()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:red\"> Note: </span> If the parquet files already exist, jump to **7!** "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (2) Load the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "database = readCSV('EVENTS.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(database,\"database.parquet\")\n",
    "#22:32 - 23:14"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "admissions = readCSV('ADMISSIONS.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(admissions,\"admissions.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "diagnoses_icd = readCSV('DIAGNOSES_ICD.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(diagnoses_icd,\"diagnoses_icd.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_icd_diagnoses = readCSV('D_ICD_DIAGNOSES.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(d_icd_diagnoses,\"d_icd_diagnoses.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_icd_procedures = readCSV('D_ICD_PROCEDURES.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(d_icd_procedures,\"d_icd_procedures.parquet\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_items = readCSV('D_ITEMS.csv', False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(d_items,\"d_items.parquet\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (3) Generate Tempos dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tempos = database\\\n",
    "    .groupby('SUBJECT_ID', 'HADM_ID', 'ICUSTAY_ID')\\\n",
    "    .agg(F.min('CHARTTIME').alias('first_timestamp'),F.max('CHARTTIME').alias('last_timestamp'))\n",
    "\n",
    "\n",
    "tempos = tempos.withColumn('duration', \n",
    "                           (F.unix_timestamp(F.col('last_timestamp')) -\n",
    "                            F.unix_timestamp(F.col('first_timestamp')))/60)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(tempos,\"tempos.parquet\")\n",
    "#22:12 - 22:31"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (4) Generate 24h dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "join_condition = [database.SUBJECT_ID == tempos.SUBJECT_ID , \\\n",
    "                  database.HADM_ID == tempos.HADM_ID , \\\n",
    "                  database.ICUSTAY_ID == tempos.ICUSTAY_ID]\n",
    "\n",
    "database24h = database\\\n",
    "    .join(tempos,join_condition)\\\n",
    "    .filter(F.unix_timestamp(database.CHARTTIME) < F.unix_timestamp(tempos.first_timestamp) + (86400))\\\n",
    "    .select(database.SUBJECT_ID,\\\n",
    "            database.HADM_ID,\\\n",
    "            database.ICUSTAY_ID,\\\n",
    "            database.ITEMID,\\\n",
    "            database.CHARTTIME,\\\n",
    "            database.CGID,\\\n",
    "            database.VALUE,\\\n",
    "            database.VALUENUM,\\\n",
    "            database.VALUEUOM)\n",
    "\n",
    "    \n",
    "# 86400 = 60s * 60min * 24h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(database24h,\"database24h.parquet\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (5) Generate aggregation of the first 24 hours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "agg24h = database24h\\\n",
    "    .drop('CHARTTIME')\\\n",
    "    .drop('CGID')\\\n",
    "    .drop('VALUE')\\\n",
    "    .drop('VALUEUOM')\\\n",
    "    .groupby('SUBJECT_ID', 'HADM_ID', 'ICUSTAY_ID','ITEMID')\\\n",
    "    .agg(F.min('VALUENUM').alias('min_value'),\\\n",
    "         F.max('VALUENUM').alias('max_value'),\\\n",
    "         F.avg('VALUENUM').alias('avg_value'),\\\n",
    "         F.count('VALUENUM').alias('count'),\\\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "writeParquet(agg24h,\"agg24h.parquet\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (6) Generate final Pandas dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "\n",
    "icustay_df = agg24h.join(tempos,tempos.ICUSTAY_ID == agg24h.ICUSTAY_ID)\\\n",
    "                   .select(agg24h.SUBJECT_ID, agg24h.HADM_ID, agg24h.ICUSTAY_ID, agg24h.ITEMID, \n",
    "                           agg24h.min_value, agg24h.max_value, agg24h.avg_value, tempos.duration)\n",
    "\n",
    "# Convert spark dataframe to pandas dataframe\n",
    "icustay_df = icustay_df.toPandas()\n",
    "\n",
    "# Replace NaN values with 1\n",
    "icustay_df.fillna(1, inplace=True)\n",
    "\n",
    "# Pivot table\n",
    "icustay_df = icustay_df.pivot_table('avg_value', ['SUBJECT_ID', 'HADM_ID', 'ICUSTAY_ID', 'duration'], 'ITEMID')\n",
    "icustay_df.reset_index(inplace=True)\n",
    "icustay_df.columns = icustay_df.columns.astype(str)\n",
    "\n",
    "# Replace NaN values with 0\n",
    "icustay_df.fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Serialize (pickle) to binary python object\n",
    "icustay_df.to_pickle('cached_dataframe.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (7) Read parquet files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "database = readParquet(\"database.parquet\")\n",
    "database.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "admissions = readParquet(\"admissions.parquet\")\n",
    "admissions.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "diagnoses_icd = readParquet(\"diagnoses_icd.parquet\")\n",
    "diagnoses_icd.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_icd_diagnoses = readParquet(\"d_icd_diagnoses.parquet\")\n",
    "d_icd_diagnoses.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_icd_procedures = readParquet(\"d_icd_procedures.parquet\")\n",
    "d_icd_procedures.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataFrame[ROW_ID: int, ITEMID: int, LABEL: string, ABBREVIATION: string, DBSOURCE: string, LINKSTO: string, CATEGORY: string, UNITNAME: string, PARAM_TYPE: string, CONCEPTID: string]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_items = readParquet(\"d_items.parquet\")\n",
    "d_items.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataFrame[SUBJECT_ID: int, HADM_ID: int, ICUSTAY_ID: int, first_timestamp: timestamp, last_timestamp: timestamp, duration: double]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tempos = readParquet(\"tempos.parquet\")\n",
    "tempos.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataFrame[SUBJECT_ID: int, HADM_ID: int, ICUSTAY_ID: int, ITEMID: int, CHARTTIME: timestamp, CGID: int, VALUE: string, VALUENUM: double, VALUEUOM: string]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "database24h = readParquet(\"database24h.parquet\")\n",
    "database24h.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataFrame[SUBJECT_ID: int, HADM_ID: int, ICUSTAY_ID: int, ITEMID: int, min_value: double, max_value: double, avg_value: double, count: bigint]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg24h = readParquet(\"agg24h.parquet\")\n",
    "agg24h.cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>ITEMID</th>\n",
       "      <th>SUBJECT_ID</th>\n",
       "      <th>HADM_ID</th>\n",
       "      <th>ICUSTAY_ID</th>\n",
       "      <th>duration</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>...</th>\n",
       "      <th>228442</th>\n",
       "      <th>228443</th>\n",
       "      <th>228444</th>\n",
       "      <th>228445</th>\n",
       "      <th>228446</th>\n",
       "      <th>228447</th>\n",
       "      <th>228448</th>\n",
       "      <th>228449</th>\n",
       "      <th>228450</th>\n",
       "      <th>228451</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>163353</td>\n",
       "      <td>243653</td>\n",
       "      <td>100.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>145834</td>\n",
       "      <td>211552</td>\n",
       "      <td>8580.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>185777</td>\n",
       "      <td>294638</td>\n",
       "      <td>2251.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>178980</td>\n",
       "      <td>214757</td>\n",
       "      <td>15.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>107064</td>\n",
       "      <td>228232</td>\n",
       "      <td>5250.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4391 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ITEMID  SUBJECT_ID  HADM_ID  ICUSTAY_ID  duration    1    2    3    4    5  \\\n",
       "0                2   163353      243653     100.0  0.0  0.0  0.0  0.0  0.0   \n",
       "1                3   145834      211552    8580.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2                4   185777      294638    2251.0  0.0  0.0  0.0  0.0  0.0   \n",
       "3                5   178980      214757      15.0  0.0  0.0  0.0  0.0  0.0   \n",
       "4                6   107064      228232    5250.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "ITEMID    6  ...  228442  228443  228444  228445  228446  228447  228448  \\\n",
       "0       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "1       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "2       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "3       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "4       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "\n",
       "ITEMID  228449  228450  228451  \n",
       "0          0.0     0.0     0.0  \n",
       "1          0.0     0.0     0.0  \n",
       "2          0.0     0.0     0.0  \n",
       "3          0.0     0.0     0.0  \n",
       "4          0.0     0.0     0.0  \n",
       "\n",
       "[5 rows x 4391 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "icustay_df = pd.read_pickle('cached_dataframe.pkl')\n",
    "icustay_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# (8) Exploratory data analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>ITEMID</th>\n",
       "      <th>SUBJECT_ID</th>\n",
       "      <th>HADM_ID</th>\n",
       "      <th>ICUSTAY_ID</th>\n",
       "      <th>duration</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>...</th>\n",
       "      <th>228442</th>\n",
       "      <th>228443</th>\n",
       "      <th>228444</th>\n",
       "      <th>228445</th>\n",
       "      <th>228446</th>\n",
       "      <th>228447</th>\n",
       "      <th>228448</th>\n",
       "      <th>228449</th>\n",
       "      <th>228450</th>\n",
       "      <th>228451</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>163353</td>\n",
       "      <td>243653</td>\n",
       "      <td>100.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>145834</td>\n",
       "      <td>211552</td>\n",
       "      <td>8580.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>185777</td>\n",
       "      <td>294638</td>\n",
       "      <td>2251.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>178980</td>\n",
       "      <td>214757</td>\n",
       "      <td>15.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>107064</td>\n",
       "      <td>228232</td>\n",
       "      <td>5250.0</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>...</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4391 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ITEMID  SUBJECT_ID  HADM_ID  ICUSTAY_ID  duration    1    2    3    4    5  \\\n",
       "0                2   163353      243653     100.0  0.0  0.0  0.0  0.0  0.0   \n",
       "1                3   145834      211552    8580.0  0.0  0.0  0.0  0.0  0.0   \n",
       "2                4   185777      294638    2251.0  0.0  0.0  0.0  0.0  0.0   \n",
       "3                5   178980      214757      15.0  0.0  0.0  0.0  0.0  0.0   \n",
       "4                6   107064      228232    5250.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "ITEMID    6  ...  228442  228443  228444  228445  228446  228447  228448  \\\n",
       "0       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "1       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "2       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "3       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "4       0.0  ...     0.0     0.0     0.0     0.0     0.0     0.0     0.0   \n",
       "\n",
       "ITEMID  228449  228450  228451  \n",
       "0          0.0     0.0     0.0  \n",
       "1          0.0     0.0     0.0  \n",
       "2          0.0     0.0     0.0  \n",
       "3          0.0     0.0     0.0  \n",
       "4          0.0     0.0     0.0  \n",
       "\n",
       "[5 rows x 4391 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "icustay_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(61301, 4391)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "icustay_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 64 ms, sys: 0 ns, total: 64 ms\n",
      "Wall time: 63 ms\n"
     ]
    },
    {
     "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": [
    "%%time\n",
    "sns.distplot(icustay_df['duration'], kde=False)\n",
    "\n",
    "# Add labels\n",
    "plt.title('Histogram of ICU stay duration (min)')\n",
    "plt.xlabel('duration (min)')\n",
    "plt.ylabel('frequency')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 436 ms, sys: 20 ms, total: 456 ms\n",
      "Wall time: 57 s\n"
     ]
    },
    {
     "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": [
    "%%time\n",
    "sns.distplot(agg24h.groupBy(\"ICUSTAY_ID\").agg(F.count(\"ITEMID\").alias('count')).toPandas()['count'], kde=False)\n",
    "\n",
    "# Add labels\n",
    "plt.title('Histogram of the number of items per ICU stay')\n",
    "plt.xlabel('number of items')\n",
    "plt.ylabel('frequency')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+----------------------+\n",
      "|count(DISTINCT ITEMID)|\n",
      "+----------------------+\n",
      "|                  4387|\n",
      "+----------------------+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "tmp = agg24h.groupBy(\"ITEMID\").agg(F.count(\"ITEMID\").alias('count')).orderBy('count', ascending=False)\n",
    "tmp.select(F.countDistinct(\"ITEMID\")).show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+------+-------------------+-----+\n",
      "|ITEMID|              LABEL|count|\n",
      "+------+-------------------+-----+\n",
      "|   211|         Heart Rate|36770|\n",
      "|    31|           Activity|35567|\n",
      "|   926|           Religion|35222|\n",
      "|    80|       Bowel Sounds|32304|\n",
      "|   829|Potassium (3.5-5.3)|29087|\n",
      "+------+-------------------+-----+\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Top 10 items\n",
    "tmp = tmp.limit(5).alias('tmp')\n",
    "tmp = d_items.join(tmp, d_items.ITEMID == tmp.ITEMID).select('tmp.ITEMID', 'LABEL', 'tmp.count')\\\n",
    "             .orderBy('count', ascending=False)\n",
    "tmp.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 76 ms, sys: 4 ms, total: 80 ms\n",
      "Wall time: 82.9 ms\n"
     ]
    },
    {
     "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": [
    "%%time\n",
    "sns.distplot(icustay_df['211'], kde=False)\n",
    "\n",
    "# Add labels\n",
    "plt.title('Histogram of the Heart Rate measurements')\n",
    "plt.xlabel('heart rate')\n",
    "plt.ylabel('frequency')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 72 ms, sys: 8 ms, total: 80 ms\n",
      "Wall time: 79 ms\n"
     ]
    },
    {
     "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": [
    "%%time\n",
    "\n",
    "corr = icustay_df[['duration', '211', '31', '926', '80', '829']].corr()\n",
    "sns.heatmap(corr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# (9) Prediction model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "#!pip install keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.optimizers import Adam, Nadam\n",
    "from keras.wrappers.scikit_learn import KerasRegressor\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "# create scaler\n",
    "scaler = MinMaxScaler(feature_range=(-1,1))\n",
    "# fit scaler on data\n",
    "scaler.fit(icustay_df)\n",
    "# apply transform\n",
    "normalized = scaler.transform(icustay_df)\n",
    "final_df = pd.DataFrame(normalized, columns=icustay_df.columns, index=icustay_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_column = ['duration'] \n",
    "unused_columns = ['SUBJECT_ID', 'HADM_ID', 'ICUSTAY_ID']\n",
    "predictors = list(set(list(final_df.columns))-set(target_column)-set(unused_columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(42910, 4387)\n",
      "(18391, 4387)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "X = final_df[predictors].values\n",
    "y = final_df[target_column].values\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=40)\n",
    "print(X_train.shape); print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_11 (Dense)             (None, 128)               561664    \n",
      "_________________________________________________________________\n",
      "dense_12 (Dense)             (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 561,793\n",
      "Trainable params: 561,793\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Define model\n",
    "model = Sequential()\n",
    "\n",
    "# The Hidden Layer :\n",
    "model.add(Dense(128, input_dim = X_train.shape[1], activation='relu'))\n",
    "\n",
    "# The Output Layer :\n",
    "model.add(Dense(1, activation='linear'))\n",
    "\n",
    "# Compile the network :\n",
    "model.compile(loss='mean_squared_error', optimizer=Adam(), metrics=['mean_squared_error'])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 42910 samples, validate on 18391 samples\n",
      "Epoch 1/15\n",
      "42910/42910 [==============================] - 11s 248us/step - loss: 0.2839 - mean_squared_error: 0.2839 - val_loss: 0.0095 - val_mean_squared_error: 0.0095\n",
      "Epoch 2/15\n",
      "42910/42910 [==============================] - 10s 230us/step - loss: 0.0097 - mean_squared_error: 0.0097 - val_loss: 0.0077 - val_mean_squared_error: 0.0077\n",
      "Epoch 3/15\n",
      "42910/42910 [==============================] - 10s 228us/step - loss: 0.0089 - mean_squared_error: 0.0089 - val_loss: 0.0090 - val_mean_squared_error: 0.0090\n",
      "Epoch 4/15\n",
      "42910/42910 [==============================] - 10s 228us/step - loss: 0.0088 - mean_squared_error: 0.0088 - val_loss: 0.0079 - val_mean_squared_error: 0.0079\n",
      "Epoch 5/15\n",
      "42910/42910 [==============================] - 10s 235us/step - loss: 0.0085 - mean_squared_error: 0.0085 - val_loss: 0.0076 - val_mean_squared_error: 0.0076\n",
      "Epoch 6/15\n",
      "42910/42910 [==============================] - 10s 228us/step - loss: 0.0079 - mean_squared_error: 0.0079 - val_loss: 0.0090 - val_mean_squared_error: 0.0090\n",
      "Epoch 7/15\n",
      "42910/42910 [==============================] - 10s 233us/step - loss: 0.0079 - mean_squared_error: 0.0079 - val_loss: 0.0068 - val_mean_squared_error: 0.0068\n",
      "Epoch 8/15\n",
      "42910/42910 [==============================] - 10s 231us/step - loss: 0.0077 - mean_squared_error: 0.0077 - val_loss: 0.0070 - val_mean_squared_error: 0.0070\n",
      "Epoch 9/15\n",
      "42910/42910 [==============================] - 10s 230us/step - loss: 0.0077 - mean_squared_error: 0.0077 - val_loss: 0.0070 - val_mean_squared_error: 0.0070\n",
      "Epoch 10/15\n",
      "42910/42910 [==============================] - 10s 232us/step - loss: 0.0077 - mean_squared_error: 0.0077 - val_loss: 0.0067 - val_mean_squared_error: 0.0067\n",
      "Epoch 11/15\n",
      "42910/42910 [==============================] - 10s 231us/step - loss: 0.0076 - mean_squared_error: 0.0076 - val_loss: 0.0074 - val_mean_squared_error: 0.0074\n",
      "Epoch 12/15\n",
      "42910/42910 [==============================] - 10s 240us/step - loss: 0.0076 - mean_squared_error: 0.0076 - val_loss: 0.0084 - val_mean_squared_error: 0.0084\n",
      "Epoch 13/15\n",
      "42910/42910 [==============================] - 10s 230us/step - loss: 0.0074 - mean_squared_error: 0.0074 - val_loss: 0.0067 - val_mean_squared_error: 0.0067\n",
      "Epoch 14/15\n",
      "42910/42910 [==============================] - 10s 229us/step - loss: 0.0073 - mean_squared_error: 0.0073 - val_loss: 0.0066 - val_mean_squared_error: 0.0066\n",
      "Epoch 15/15\n",
      "42910/42910 [==============================] - 10s 232us/step - loss: 0.0073 - mean_squared_error: 0.0073 - val_loss: 0.0066 - val_mean_squared_error: 0.0066\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train,\n",
    "                    batch_size=32,\n",
    "                    epochs=15,\n",
    "                    verbose=1,\n",
    "                    validation_data=(X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy: 0.0066115863224317575\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate(X_test, y_test, verbose=0)\n",
    "print('Test accuracy:', scores[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f765e519ef0>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.plot(history.history['mean_squared_error'])\n",
    "plt.plot(history.history['val_mean_squared_error'])\n",
    "\n",
    "plt.title('Train and test performances')\n",
    "plt.ylabel('MSE')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['Train MSE', 'Test MSE'], loc='upper right')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_predictions = model.predict(X_test)\n",
    "\n",
    "plt.scatter(y_test, test_predictions)\n",
    "plt.xlabel('True Values')\n",
    "plt.ylabel('Predictions')\n",
    "_ = plt.plot([-1, 1], [-1, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}