Cleaned up province segmentation

notebooks/Segment Provinces in Wijken and Gemeentes.ipynb

@@ -0,0 +1,139 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Segment provinces\n",
+    "\n",
+    "\n",
+    "Create wijk and gemeente level segmentations for all Dutch provinces and save as geojson and Gabmap KML.\n",
+    "\n",
+    "All is based on CBS data.\n",
+    "\n",
+    "For Friesland, several wijken are merged.\n",
+    "\n",
+    "Note: only applied to Groningen and Friesland, because other provinces give gemetry errors."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%load_ext autoreload\n",
+    "\n",
+    "%autoreload 2"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import sys\n",
+    "sys.path.append('../')\n",
+    "\n",
+    "from stimmen.latitude_longitude import reverse_latitude_longitude\n",
+    "from stimmen.shapefile import shapefiles_to_geojson\n",
+    "from stimmen.cbs import gwb_in_province, get_available_provinces\n",
+    "from stimmen.geojson import merge_features\n",
+    "from gabmap import as_gabmap_kml\n",
+    "\n",
+    "from shapely.geometry import shape, box, mapping\n",
+    "\n",
+    "import json\n",
+    "import folium\n",
+    "import pickle\n",
+    "\n",
+    "from collections import defaultdict"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for province in ['Groningen', 'Friesland']:\n",
+    "    wijken_geojson = gwb_in_province(province, 'wijk', 2018)\n",
+    "    gemeente_geojson = gwb_in_province(province, 'gem', 2018)\n",
+    "\n",
+    "    if province == 'Friesland':\n",
+    "        for gemeente in {'Ameland', 'Harlingen', 'Schiermonnikoog', 'Terschelling', 'Vlieland'}:\n",
+    "            merged_geojson = merge_features(\n",
+    "                wijken_geojson.copy(),\n",
+    "                condition=lambda feature: feature['properties']['GM_NAAM'] == gemeente,\n",
+    "                aggregate={'WK_NAAM': ' '.join}\n",
+    "            )\n",
+    "\n",
+    "        merge_leeuwarden_only_above_latitude = 53.167\n",
+    "        wijken_geojson = merge_features(\n",
+    "            wijken_geojson,\n",
+    "            condition=lambda feature: (\n",
+    "                feature['properties']['GM_NAAM'] == 'Leeuwarden' and\n",
+    "                shape(feature['geometry']).centroid.y > merge_leeuwarden_only_above_latitude\n",
+    "            ),\n",
+    "            aggregate={'WK_NAAM': ' '.join}\n",
+    "        )\n",
+    "    \n",
+    "    #Some gemeentes appear twice in the cbs data.\n",
+    "    for gemeente in [feature['properties']['GM_NAAM'] for feature in gemeente_geojson['features']]:\n",
+    "        gemeente_geojson = merge_features(\n",
+    "            gemeente_geojson, condition=lambda feature: feature['properties']['GM_NAAM'] == gemeente)\n",
+    "        \n",
+    "    for feature in wijken_geojson['features']:\n",
+    "        feature['properties']['gemeente_en_wijk_naam'] = (\n",
+    "            feature['properties']['GM_NAAM'] +\n",
+    "            ', ' +\n",
+    "            feature['properties'].get('WK_NAAM', '')\n",
+    "        ).replace('&', 'en').replace('/', ' ').replace('\"', ' ').replace(\"'\", ' ')\n",
+    "        \n",
+    "    for feature in gemeente_geojson['features']:\n",
+    "        feature['properties']['gemeente_naam'] = (\n",
+    "            feature['properties']['GM_NAAM']\n",
+    "        ).replace('&', 'en').replace('/', ' ').replace('\"', ' ').replace(\"'\", ' ')\n",
+    "    \n",
+    "    with open('../data/{}_wijken.geojson'.format(province), 'w') as f:\n",
+    "        json.dump(wijken_geojson, f)\n",
+    "    with open('../data/{}_gemeentes.geojson'.format(province), 'w') as f:\n",
+    "        json.dump(gemeente_geojson, f)\n",
+    "    \n",
+    "    with open('../data/{}_wijken.kml'.format(province), 'w') as f:\n",
+    "        f.write(as_gabmap_kml(wijken_geojson, name_property='gemeente_en_wijk_naam'))\n",
+    "    with open('../data/{}_gemeentes.kml'.format(province), 'w') as f:\n",
+    "        f.write(as_gabmap_kml(gemeente_geojson, name_property='gemeente_naam'))"
+   ]
+  },
+  {
+   "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": "3.6.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}