{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!curl --location --remote-name --remote-header-name \\\n",
    "    http://ftp.eea.europa.eu/www/eprtr/v17/E-PRTR_database_v17_xls.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!unzip E-PRTR_database_v17_xls.zip 'Pollutant releases.xlsx'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas\n",
    "\n",
    "pollutant_releases = pandas.read_excel('Pollutant releases.xlsx')\n",
    "pollutant_releases.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import geopandas\n",
    "\n",
    "# http://geopandas.org/gallery/create_geopandas_from_pandas.html\n",
    "pollutant_releases_geo = geopandas.GeoDataFrame(\n",
    "    pollutant_releases[[\n",
    "        'FacilityID', 'Lat', 'Long', 'ReportingYear', 'PollutantName', \n",
    "        'PollutantGroupName', 'TotalQuantity', 'TotalQuantity',\n",
    "    ]],\n",
    "    geometry=geopandas.points_from_xy(\n",
    "        pollutant_releases['Long'],\n",
    "        pollutant_releases['Lat'],\n",
    "    ),\n",
    ")\n",
    "for column_name in pollutant_releases_geo.select_dtypes('object'):\n",
    "    pollutant_releases_geo[column_name] = pollutant_releases_geo[column_name].astype('category')\n",
    "pollutant_releases_geo.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'geopandas.geodataframe.GeoDataFrame'>\n",
      "RangeIndex: 652351 entries, 0 to 652350\n",
      "Data columns (total 9 columns):\n",
      "FacilityID            652351 non-null int64\n",
      "Lat                   652351 non-null float64\n",
      "Long                  652351 non-null float64\n",
      "ReportingYear         652351 non-null int64\n",
      "PollutantName         652351 non-null category\n",
      "PollutantGroupName    652351 non-null category\n",
      "TotalQuantity         652351 non-null float64\n",
      "TotalQuantity         652351 non-null float64\n",
      "geometry              652351 non-null geometry\n",
      "dtypes: category(2), float64(4), geometry(1), int64(2)\n",
      "memory usage: 36.1 MB\n"
     ]
    }
   ],
   "source": [
    "pollutant_releases_geo.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "world = geopandas.read_file(geopandas.datasets.get_path('naturalearth_lowres'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%config InlineBackend.figure_format = 'retina'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "european_facilities_2011 = geopandas.GeoDataFrame(pollutant_releases_geo[\n",
    "    (pollutant_releases_geo.Long > -28)\n",
    "    & (pollutant_releases_geo.Long < 32)\n",
    "    & (pollutant_releases_geo.Lat > 32)\n",
    "    & (pollutant_releases_geo.Lat < 75)\n",
    "    & (pollutant_releases_geo.ReportingYear == 2011)\n",
    "].set_index('FacilityID').groupby('FacilityID').first())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import geoplot\n",
    "\n",
    "ax = geoplot.kdeplot(\n",
    "    european_facilities_2011,\n",
    "    clip=world.geometry,\n",
    "    shade=True,\n",
    "    cmap='Reds',\n",
    "    projection=geoplot.crs.AlbersEqualArea(),\n",
    "    figsize=(20, 12),\n",
    "    # extent=(-28, 32, 32, 75),\n",
    ")\n",
    "geoplot.polyplot(\n",
    "    world.geometry,\n",
    "    ax=ax,\n",
    "    linewidth=1,\n",
    ")\n",
    "geoplot.pointplot(\n",
    "    european_facilities_2011,\n",
    "    color='orange',\n",
    "    s=0.4,\n",
    "    ax=ax,\n",
    ");\n",
    "ax.set_title('E-PRTR_database_v17_xls.zip / Pollutant releases.xlsx / 2011 / facilities');"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}