{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9de5907f-18f5-4cb1-903e-26028ff1fa03",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_rows', 100)\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad0e6044",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_patients_22 = pd.read_csv(\"VAERS/2021VAERSData/2021VAERSDATA.csv\", index_col='VAERS_ID', encoding='latin1', low_memory=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4472e902",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_patients_22.index\n",
    "df_patients_22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fb54ec47",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_vax_22 = pd.read_csv(\"VAERS/2021VAERSData/2021VAERSVAX.csv\", index_col='VAERS_ID', encoding='latin1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8942466",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_vax_22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d31c5225",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_COVID19_22 = df_vax_22[df_vax_22[\"VAX_TYPE\"] == \"COVID19\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "be2b57c9",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_COVID19_22"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8c432ab2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_patients_COVID19 = pd.merge(df_patients_22, df_COVID19_22, left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fef882ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_patients_22.shape, df_COVID19_22.shape, df_patients_COVID19.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6ce15322",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# pd.set_option('display.max_rows', 100)\n",
    "df_patients_COVID19"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "233bc590",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_PFIZER_BIONTECH = df_patients_COVID19[df_patients_COVID19[\"VAX_MANU\"] == \"MODERNA\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e41885a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_PFIZER_BIONTECH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86e0e4f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_PFIZER_BIONTECH.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b3cb943",
   "metadata": {},
   "outputs": [],
   "source": [
    "# table = pd.pivot_table(df_PFIZER_BIONTECH, values='DIED', columns=['DIED'], aggfunc=np.sum)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "284bf448",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_rows', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "349da946",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_DIED = df_PFIZER_BIONTECH[df_PFIZER_BIONTECH['DIED']=='Y'][['VAX_LOT']]\n",
    "df_DIED"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9c5ff2fd",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# df = df.groupby(['VAX_LOT']).count()\n",
    "#df2 = df.sort_values(by=['DIED'], ascending=False)\n",
    "#pd.set_option('display.max_rows', None)\n",
    "#print(df2)\n",
    "df_DIED.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2f660707",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_DISABLE = df_PFIZER_BIONTECH[df_PFIZER_BIONTECH['DISABLE']=='Y'][['VAX_LOT']]\n",
    "df_DISABLE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d782392",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_DISABLE.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9f537710",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_L_THREAT = df_PFIZER_BIONTECH[df_PFIZER_BIONTECH['L_THREAT']=='Y'][['VAX_LOT']]\n",
    "df_L_THREAT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a629007",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_L_THREAT.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f675fcfe",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_ADR = df_PFIZER_BIONTECH[['VAX_LOT']]\n",
    "df_ADR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7b3005db",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_ADR_valueCounts = df_ADR.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "07c93e74",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df_ADR_valueCounts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1244c494",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}