Load all data into CSV files with polar

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b784745d",
   "metadata": {
    "papermill": {
     "duration": 0.00625,
     "end_time": "2024-09-16T08:51:19.889852",
     "exception": false,
     "start_time": "2024-09-16T08:51:19.883602",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Overview\n",
    "\n",
    "In this notebook, we will process and represent time-series data. There are two important parameters\n",
    "\n",
    "## The length of time-series data\n",
    "\n",
    "Default is 72 hours\n",
    "\n",
    "## The bucket size\n",
    "\n",
    "The bucket size which tells in what size time windows we want to divide our time-series data. For example, if we select 2 bucket size, it will aggregate data for every 2 hours and a time-series of length 24 hours will be represented as time-series with 12 time-windows where data for every 2 hours is aggregated from original raw tim-series.\n",
    "\n",
    "We will save the time-series data in data dictionaries in the format that can be directly used for the next step(deep learning analysis.)\n",
    "\n",
    "\n",
    "# Data Imputation\n",
    "\n",
    "The data imputation is optional feature. It will be done by forward fill and mean or median imputation. Values will be forward fill first and if no value exists for that admission we will use mean or median value for the patient."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "58f18a1e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:19.903301Z",
     "iopub.status.busy": "2024-09-16T08:51:19.902834Z",
     "iopub.status.idle": "2024-09-16T08:51:20.234818Z",
     "shell.execute_reply": "2024-09-16T08:51:20.233564Z"
    },
    "papermill": {
     "duration": 0.342175,
     "end_time": "2024-09-16T08:51:20.237885",
     "exception": false,
     "start_time": "2024-09-16T08:51:19.895710",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import polars as pl\n",
    "from tqdm import tqdm\n",
    "import pickle\n",
    "import os\n",
    "import numpy as np\n",
    "import warnings\n",
    "\n",
    "warnings.simplefilter(action='ignore', category=FutureWarning)\n",
    "\n",
    "if_mort=True\n",
    "if_admn=False\n",
    "if_los=False\n",
    "feat_cond=True # diagnosis\n",
    "feat_proc=True # procedures\n",
    "feat_out=True # out events\n",
    "feat_chart=True # chart events\n",
    "feat_med=True # medications\n",
    "impute=False\n",
    "include_time=72\n",
    "bucket=1\n",
    "predW=2"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e08cf662",
   "metadata": {
    "papermill": {
     "duration": 0.005371,
     "end_time": "2024-09-16T08:51:20.249071",
     "exception": false,
     "start_time": "2024-09-16T08:51:20.243700",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Load the base data(patient cohort)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3c1270c0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:20.262229Z",
     "iopub.status.busy": "2024-09-16T08:51:20.261757Z",
     "iopub.status.idle": "2024-09-16T08:51:20.749285Z",
     "shell.execute_reply": "2024-09-16T08:51:20.747906Z"
    },
    "papermill": {
     "duration": 0.497386,
     "end_time": "2024-09-16T08:51:20.752095",
     "exception": false,
     "start_time": "2024-09-16T08:51:20.254709",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "data = pl.read_csv('/kaggle/input/mimic-iv-classification-regression-tasks/mimic_iv_classification_regression_tasks.csv')\n",
    "data = data.with_columns([\n",
    "    pl.col('intime').str.strptime(pl.Datetime, format='%Y-%m-%d %H:%M:%S'),\n",
    "    pl.col('outtime').str.strptime(pl.Datetime, format='%Y-%m-%d %H:%M:%S')\n",
    "])\n",
    "\n",
    "data = data.with_columns([\n",
    "    (pl.col('outtime') - pl.col('intime')).alias('los')\n",
    "])\n",
    "data = data.with_columns([\n",
    "    pl.col('los').cast(pl.Float64) / 3600000000.0  # Convert seconds to hours\n",
    "])\n",
    "data = data.filter(pl.col('los') > 0)\n",
    "data = data.with_columns([\n",
    "    pl.col('Age').cast(pl.Int32),\n",
    "    pl.col('los').floor().cast(pl.Int32).alias('los') # Take the integer part of 'los'\n",
    "])\n",
    "data = data.with_columns([\n",
    "    pl.col('intime').dt.strftime('%Y-%m-%d %H:%M:%S').alias('intime'),\n",
    "    pl.col('outtime').dt.strftime('%Y-%m-%d %H:%M:%S').alias('outtime')\n",
    "])\n",
    "data.write_csv('mimic_iv_classification_regression_tasks.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38d4f6cc",
   "metadata": {
    "papermill": {
     "duration": 0.005752,
     "end_time": "2024-09-16T08:51:20.763528",
     "exception": false,
     "start_time": "2024-09-16T08:51:20.757776",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Load all the features\n",
    "\n",
    "We will load all the datasets in here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "854324e2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:20.777005Z",
     "iopub.status.busy": "2024-09-16T08:51:20.776577Z",
     "iopub.status.idle": "2024-09-16T08:51:21.273952Z",
     "shell.execute_reply": "2024-09-16T08:51:21.272636Z"
    },
    "papermill": {
     "duration": 0.507613,
     "end_time": "2024-09-16T08:51:21.277147",
     "exception": false,
     "start_time": "2024-09-16T08:51:20.769534",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ ======READING DIAGNOSIS ]\n"
     ]
    }
   ],
   "source": [
    "print(\"[ ======READING DIAGNOSIS ]\")\n",
    "cond = pl.read_csv(\"/kaggle/input/reduce-dimension-mimic-iv-v2-0-icu-diagnosis-icd10/dimension_reduce_diag_icu.csv\")\n",
    "cond = cond.filter(pl.col('stay_id').is_in(data['stay_id']))\n",
    "\n",
    "# Group by 'stay_id' and count the occurrences, then find the maximum count\n",
    "cond_per_adm = cond.group_by('stay_id').agg(pl.count('stay_id').alias('count')).select(pl.col('count').max()).to_numpy()[0][0]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "9d16e95c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:21.292028Z",
     "iopub.status.busy": "2024-09-16T08:51:21.291350Z",
     "iopub.status.idle": "2024-09-16T08:51:21.298548Z",
     "shell.execute_reply": "2024-09-16T08:51:21.297126Z"
    },
    "papermill": {
     "duration": 0.017821,
     "end_time": "2024-09-16T08:51:21.301230",
     "exception": false,
     "start_time": "2024-09-16T08:51:21.283409",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "39\n"
     ]
    }
   ],
   "source": [
    "print(cond_per_adm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ee8d465c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:21.315595Z",
     "iopub.status.busy": "2024-09-16T08:51:21.314456Z",
     "iopub.status.idle": "2024-09-16T08:51:21.469992Z",
     "shell.execute_reply": "2024-09-16T08:51:21.468630Z"
    },
    "papermill": {
     "duration": 0.165533,
     "end_time": "2024-09-16T08:51:21.472753",
     "exception": false,
     "start_time": "2024-09-16T08:51:21.307220",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "cond.write_csv('dimension_reduce_diag_icu.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "15c300f8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:21.487346Z",
     "iopub.status.busy": "2024-09-16T08:51:21.485984Z",
     "iopub.status.idle": "2024-09-16T08:51:22.592657Z",
     "shell.execute_reply": "2024-09-16T08:51:22.591123Z"
    },
    "papermill": {
     "duration": 1.116775,
     "end_time": "2024-09-16T08:51:22.595357",
     "exception": false,
     "start_time": "2024-09-16T08:51:21.478582",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ ======READING PROCEDURES ]\n",
      "shape: (1, 9)\n",
      "┌────────────┬──────────┬──────────┬────────┬───┬────────────┬──────────────────┬─────┬────────────┐\n",
      "│ subject_id ┆ hadm_id  ┆ stay_id  ┆ itemid ┆ … ┆ intime     ┆ event_time_from_ ┆ los ┆ start_time │\n",
      "│ ---        ┆ ---      ┆ ---      ┆ ---    ┆   ┆ ---        ┆ admit            ┆ --- ┆ ---        │\n",
      "│ i64        ┆ i64      ┆ i64      ┆ i64    ┆   ┆ str        ┆ ---              ┆ i32 ┆ i8         │\n",
      "│            ┆          ┆          ┆        ┆   ┆            ┆ str              ┆     ┆            │\n",
      "╞════════════╪══════════╪══════════╪════════╪═══╪════════════╪══════════════════╪═════╪════════════╡\n",
      "│ 10000032   ┆ 29079034 ┆ 39553978 ┆ 224275 ┆ … ┆ 2180-07-23 ┆ 0 days 00:24:00  ┆ 9   ┆ 0          │\n",
      "│            ┆          ┆          ┆        ┆   ┆ 14:00:00   ┆                  ┆     ┆            │\n",
      "└────────────┴──────────┴──────────┴────────┴───┴────────────┴──────────────────┴─────┴────────────┘\n"
     ]
    }
   ],
   "source": [
    "print(\"[ ======READING PROCEDURES ]\")\n",
    "proc = pl.read_csv(\"/kaggle/input/mimic-iv-v2-0-icu-procedures/preproc_proc_icu.csv\")\n",
    "\n",
    "\n",
    "# Filter rows based on stay_id\n",
    "proc = proc.join(\n",
    "    data.select([\"stay_id\", \"los\"]),\n",
    "    on=\"stay_id\",\n",
    "    how=\"inner\",\n",
    ")\n",
    "\n",
    "# # Split event_time_from_admit and extract start_time\n",
    "proc = proc.with_columns(\n",
    "    pl.col(\"event_time_from_admit\").str.split(\" \").alias(\"split_event_time\"),\n",
    ")\n",
    "proc = proc.with_columns([\n",
    "    pl.col(\"split_event_time\").list.last().str.split(\":\").alias('event_time'),\n",
    "    pl.col(\"split_event_time\").list.first().alias('event_days'),\n",
    "])\n",
    "proc = proc.with_columns(\n",
    "    pl.col(\"event_time\").list.get(0).alias('event_hours')\n",
    ")\n",
    "proc = proc.with_columns([\n",
    "    pl.col(\"event_days\").cast(pl.Int8).alias(\"event_days\"),\n",
    "    pl.col(\"event_hours\").cast(pl.Int8).alias(\"event_hours\"),\n",
    "])\n",
    "proc = proc.with_columns([\n",
    "    (pl.col('event_days') * 24 + pl.col('event_hours')).alias('start_time')\n",
    "])\n",
    "\n",
    "proc = proc.drop([\"split_event_time\", \"event_days\", \"event_hours\",\"event_time\"])\n",
    "\n",
    "# Filter rows with start_time >= 0 and sanity > 0\n",
    "proc = proc.filter(pl.col(\"start_time\") >= 0)\n",
    "proc = proc.with_columns((pl.col(\"los\") - pl.col(\"start_time\")).alias(\"sanity\"))\n",
    "proc = proc.filter(pl.col(\"sanity\") > 0)\n",
    "proc = proc.drop([\"sanity\"])\n",
    "first_row = proc.head(1)\n",
    "\n",
    "print(first_row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "47d09f69",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:22.610820Z",
     "iopub.status.busy": "2024-09-16T08:51:22.609702Z",
     "iopub.status.idle": "2024-09-16T08:51:22.770329Z",
     "shell.execute_reply": "2024-09-16T08:51:22.769113Z"
    },
    "papermill": {
     "duration": 0.171763,
     "end_time": "2024-09-16T08:51:22.773404",
     "exception": false,
     "start_time": "2024-09-16T08:51:22.601641",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "proc.write_csv('preproc_proc_icu.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ea34bce2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:22.788237Z",
     "iopub.status.busy": "2024-09-16T08:51:22.787736Z",
     "iopub.status.idle": "2024-09-16T08:51:27.234168Z",
     "shell.execute_reply": "2024-09-16T08:51:27.232284Z"
    },
    "papermill": {
     "duration": 4.457294,
     "end_time": "2024-09-16T08:51:27.237049",
     "exception": false,
     "start_time": "2024-09-16T08:51:22.779755",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ ======READING OUT EVENTS ]\n",
      "shape: (1, 8)\n",
      "┌────────────┬──────────┬──────────┬────────┬─────────────┬─────────────┬─────────────┬────────────┐\n",
      "│ subject_id ┆ hadm_id  ┆ stay_id  ┆ itemid ┆ charttime   ┆ intime      ┆ event_time_ ┆ start_time │\n",
      "│ ---        ┆ ---      ┆ ---      ┆ ---    ┆ ---         ┆ ---         ┆ from_admit  ┆ ---        │\n",
      "│ i64        ┆ i64      ┆ i64      ┆ i64    ┆ str         ┆ str         ┆ ---         ┆ i16        │\n",
      "│            ┆          ┆          ┆        ┆             ┆             ┆ str         ┆            │\n",
      "╞════════════╪══════════╪══════════╪════════╪═════════════╪═════════════╪═════════════╪════════════╡\n",
      "│ 10000032   ┆ 29079034 ┆ 39553978 ┆ 226560 ┆ 2180-07-23  ┆ 2180-07-23  ┆ 0 days      ┆ 1          │\n",
      "│            ┆          ┆          ┆        ┆ 15:00:00    ┆ 14:00:00    ┆ 01:00:00    ┆            │\n",
      "└────────────┴──────────┴──────────┴────────┴─────────────┴─────────────┴─────────────┴────────────┘\n"
     ]
    }
   ],
   "source": [
    "print(\"[ ======READING OUT EVENTS ]\")\n",
    "out = pl.read_csv(\"/kaggle/input/mimic-iv-v2-0-icu-output-events-data/preproc_out_icu.csv\")\n",
    "out = out.filter(pl.col('stay_id').is_in(data['stay_id']))\n",
    "out = out.with_columns(\n",
    "    pl.col(\"event_time_from_admit\").str.split(\" \").alias(\"split_event_time\"),\n",
    ")\n",
    "out = out.with_columns([\n",
    "    pl.col(\"split_event_time\").list.last().str.split(\":\").alias('event_time'),\n",
    "    pl.col(\"split_event_time\").list.first().alias('event_days'),\n",
    "])\n",
    "out = out.with_columns(\n",
    "    pl.col(\"event_time\").list.get(0).alias('event_hours')\n",
    ")\n",
    "out = out.with_columns([\n",
    "    pl.col(\"event_days\").cast(pl.Int16).alias(\"event_days\"),\n",
    "    pl.col(\"event_hours\").cast(pl.Int16).alias(\"event_hours\"),\n",
    "])\n",
    "out = out.with_columns([\n",
    "    (pl.col('event_days') * 24 + pl.col('event_hours')).alias('start_time')\n",
    "])\n",
    "\n",
    "out = out.drop([\"split_event_time\", \"event_days\", \"event_hours\",\"event_time\"])\n",
    "\n",
    "# Filter rows with start_time >= 0 and sanity > 0\n",
    "# out = out.filter(pl.col(\"start_time\") >= 0)\n",
    "# out = out.with_columns((pl.col(\"los\") - pl.col(\"start_time\")).alias(\"sanity\"))\n",
    "# out = out.filter(pl.col(\"sanity\") > 0)\n",
    "# out = out.drop([\"sanity\"])\n",
    "first_row = out.head(1)\n",
    "\n",
    "print(first_row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "032f0a20",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:27.253629Z",
     "iopub.status.busy": "2024-09-16T08:51:27.252483Z",
     "iopub.status.idle": "2024-09-16T08:51:28.499380Z",
     "shell.execute_reply": "2024-09-16T08:51:28.498002Z"
    },
    "papermill": {
     "duration": 1.258553,
     "end_time": "2024-09-16T08:51:28.502337",
     "exception": false,
     "start_time": "2024-09-16T08:51:27.243784",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "out.write_csv('preproc_out_icu.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e1171fc9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:28.517588Z",
     "iopub.status.busy": "2024-09-16T08:51:28.517115Z",
     "iopub.status.idle": "2024-09-16T08:51:34.048370Z",
     "shell.execute_reply": "2024-09-16T08:51:34.046673Z"
    },
    "papermill": {
     "duration": 5.542512,
     "end_time": "2024-09-16T08:51:34.051516",
     "exception": false,
     "start_time": "2024-09-16T08:51:28.509004",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ ======READING MEDICATIONS ]\n",
      "shape: (1, 13)\n",
      "┌────────────┬──────────┬──────────┬────────┬───┬───────────┬─────────┬────────────┬───────────┐\n",
      "│ subject_id ┆ hadm_id  ┆ stay_id  ┆ itemid ┆ … ┆ amount    ┆ orderid ┆ start_time ┆ stop_time │\n",
      "│ ---        ┆ ---      ┆ ---      ┆ ---    ┆   ┆ ---       ┆ ---     ┆ ---        ┆ ---       │\n",
      "│ i64        ┆ i64      ┆ i64      ┆ i64    ┆   ┆ f64       ┆ i64     ┆ i16        ┆ i16       │\n",
      "╞════════════╪══════════╪══════════╪════════╪═══╪═══════════╪═════════╪════════════╪═══════════╡\n",
      "│ 10000032   ┆ 29079034 ┆ 39553978 ┆ 220862 ┆ … ┆ 49.999999 ┆ 427849  ┆ 3          ┆ 4         │\n",
      "└────────────┴──────────┴──────────┴────────┴───┴───────────┴─────────┴────────────┴───────────┘\n"
     ]
    }
   ],
   "source": [
    "print(\"[ ======READING MEDICATIONS ]\")\n",
    "meds = pl.read_csv(\"/kaggle/input/mimic-iv-v2-0-icu-medications-data/preproc_med_icu.csv\")\n",
    "meds = meds.filter(pl.col('stay_id').is_in(data['stay_id']))\n",
    "meds = meds.with_columns(\n",
    "    pl.col(\"start_hours_from_admit\").str.split(\" \").alias(\"split_start_time\"),\n",
    ")\n",
    "meds = meds.with_columns([\n",
    "    pl.col(\"split_start_time\").list.last().str.split(\":\").alias('start_time_admit'),\n",
    "    pl.col(\"split_start_time\").list.first().alias('start_days'),\n",
    "])\n",
    "meds = meds.with_columns(\n",
    "    pl.col(\"start_time_admit\").list.get(0).alias('start_hours')\n",
    ")\n",
    "meds = meds.with_columns([\n",
    "    pl.col(\"start_days\").cast(pl.Int16).alias(\"start_days\"),\n",
    "    pl.col(\"start_hours\").cast(pl.Int16).alias(\"start_hours\"),\n",
    "])\n",
    "meds = meds.with_columns([\n",
    "    (pl.col('start_days') * 24 + pl.col('start_hours')).alias('start_time')\n",
    "])\n",
    "\n",
    "# find out stop time\n",
    "meds = meds.with_columns(\n",
    "    pl.col(\"stop_hours_from_admit\").str.split(\" \").alias(\"split_stop_time\"),\n",
    ")\n",
    "meds = meds.with_columns([\n",
    "    pl.col(\"split_stop_time\").list.last().str.split(\":\").alias('stop_time_admit'),\n",
    "    pl.col(\"split_stop_time\").list.first().alias('stop_days'),\n",
    "])\n",
    "meds = meds.with_columns(\n",
    "    pl.col(\"stop_time_admit\").list.get(0).alias('stop_hours')\n",
    ")\n",
    "meds = meds.with_columns([\n",
    "    pl.col(\"stop_days\").cast(pl.Int16).alias(\"stop_days\"),\n",
    "    pl.col(\"stop_hours\").cast(pl.Int16).alias(\"stop_hours\"),\n",
    "])\n",
    "meds = meds.with_columns([\n",
    "    (pl.col('stop_days') * 24 + pl.col('stop_hours')).alias('stop_time')\n",
    "])\n",
    "\n",
    "meds = meds.drop([\"split_start_time\", \"start_days\", \"start_hours\",\"start_time_admit\",\"split_stop_time\", \"stop_days\", \"stop_hours\",\"stop_time_admit\"])\n",
    "\n",
    "meds = meds.with_columns((pl.col(\"stop_time\") - pl.col(\"start_time\")).alias(\"sanity\"))\n",
    "meds = meds.filter(pl.col(\"sanity\") > 0)\n",
    "meds = meds.drop([\"sanity\"])\n",
    "\n",
    "# meds = meds.with_columns((pl.col(\"los\") - pl.col(\"start_time\")).alias(\"sanity\"))\n",
    "# meds = meds.filter(pl.col(\"sanity\") > 0)\n",
    "# meds = meds.drop([\"sanity\"])\n",
    "meds = meds.with_columns([\n",
    "    pl.col('rate').cast(pl.Float64).alias('rate'),\n",
    "    pl.col('amount').cast(pl.Float64).alias('amount')\n",
    "])\n",
    "first_row = meds.head(1)\n",
    "print(first_row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "92ea5174",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2024-09-16T08:51:34.067298Z",
     "iopub.status.busy": "2024-09-16T08:51:34.066772Z",
     "iopub.status.idle": "2024-09-16T08:51:35.917712Z",
     "shell.execute_reply": "2024-09-16T08:51:35.916563Z"
    },
    "papermill": {
     "duration": 1.862444,
     "end_time": "2024-09-16T08:51:35.920680",
     "exception": false,
     "start_time": "2024-09-16T08:51:34.058236",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "meds.write_csv('preproc_med_icu.csv')"
   ]
  }
 ],
 "metadata": {
  "kaggle": {
   "accelerator": "none",
   "dataSources": [
    {
     "datasetId": 5648283,
     "sourceId": 9323836,
     "sourceType": "datasetVersion"
    },
    {
     "datasetId": 5658228,
     "sourceId": 9337355,
     "sourceType": "datasetVersion"
    },
    {
     "datasetId": 5658603,
     "sourceId": 9337863,
     "sourceType": "datasetVersion"
    },
    {
     "datasetId": 5658995,
     "sourceId": 9338379,
     "sourceType": "datasetVersion"
    },
    {
     "datasetId": 5659160,
     "sourceId": 9338614,
     "sourceType": "datasetVersion"
    }
   ],
   "dockerImageVersionId": 30761,
   "isGpuEnabled": false,
   "isInternetEnabled": true,
   "language": "python",
   "sourceType": "notebook"
  },
  "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": "3.10.14"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 19.825984,
   "end_time": "2024-09-16T08:51:36.551030",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2024-09-16T08:51:16.725046",
   "version": "2.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}