Merge pull request #28 from zazuko/fix-ci

Improve CI

Author: ludovicm67

.github/workflows/ci.yaml

@@ -3,7 +3,7 @@ name: Build HTML version
 on:
   push:
     branches:
-      - master
+      - "**"
 
 jobs:
   docker:
@@ -15,14 +15,22 @@ jobs:
 
     steps:
       - name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@v3
 
-      - uses: actions/setup-python@v1
+      - uses: actions/setup-python@v4
         with:
           python-version: 3.9
 
+      - name: Install and configure Poetry
+        uses: snok/install-poetry@v1
+        with:
+          version: 1.5.1
+          virtualenvs-create: false
+          virtualenvs-in-project: false
+          installer-parallel: true
+
       - name: Install dependencies
-        run: pip install -r requirements.txt
+        run: poetry install
 
       - name: Install Jupyter
         run: pip install jupyterlab sparqlkernel
@@ -36,10 +44,13 @@ jobs:
       - name: Build HTML version of the notebook
         run: ./run.sh
 
+      # Only deploy on GitHub Pages if we are on the "master" branch
       - name: Configure CNAME
+        if: ${{ github.event_name == 'push' && github.ref == 'refs/heads/master' }}
         run: echo jupyter.zazuko.com > html/CNAME
 
       - name: Deploy to GitHub Pages
+        if: ${{ github.event_name == 'push' && github.ref == 'refs/heads/master' }}
         uses: JamesIves/[email protected]
         with:
           branch: gh-pages

notebooks/zefix/addresses.json

@@ -2,6 +2,35 @@
 "type": "FeatureCollection",
 "crs": { "type": "name", "properties": { "name": "urn:ogc:def:crs:OGC:1.3:CRS84" } },
 "features": [
+{ "type": "Feature", "properties": { "address": "Industriestrasse, 6055 Alpnach Dorf, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.277393652963429, 46.94092995 ] } },
+{ "type": "Feature", "properties": { "address": "Seeburgstrasse, 6006 Luzern, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.3389154, 47.0530533 ] } },
+{ "type": "Feature", "properties": { "address": "Via Giovan Battista Pioda, 6900 Lugano, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.9542748, 46.0075449 ] } },
+{ "type": "Feature", "properties": { "address": "Via Carlo Maderno, 6900 Lugano, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.9549544, 46.0093445 ] } },
+{ "type": "Feature", "properties": { "address": "Rue du Pas de l'Ours, 3963 Crans VS, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.467844, 46.3074497 ] } },
+{ "type": "Feature", "properties": { "address": "Poststrasse, 9000 St. Gallen, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 9.3740105, 47.4250466 ] } },
+{ "type": "Feature", "properties": { "address": "Rue de la Rôtisserie, 1204 Genève, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.1471323, 46.2026403 ] } },
+{ "type": "Feature", "properties": { "address": "Rue de la Corraterie, 1204 Genève, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.143915, 46.2014168 ] } },
+{ "type": "Feature", "properties": { "address": "Ueberlandstrasse, 8600 Dübendorf, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.6187628, 47.4002515 ] } },
+{ "type": "Feature", "properties": { "address": "Selnaustrasse, 8001 Zürich, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.5329915, 47.3700265 ] } },
+{ "type": "Feature", "properties": { "address": "Route de l'Aéroport, 1950 Sion, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.3477037, 46.2236772 ] } },
+{ "type": "Feature", "properties": { "address": "Brandschenkestrasse, 8001 Zürich, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.5331798, 47.369396 ] } },
+{ "type": "Feature", "properties": { "address": "Avenue de Tivoli, 1007 Lausanne, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.616784, 46.5209716 ] } },
+{ "type": "Feature", "properties": { "address": "Haselstrasse, 5400 Baden, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.310141, 47.4780467 ] } },
+{ "type": "Feature", "properties": { "address": "Rue des Pâquis, 1201 Genève, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.1480865, 46.2098051 ] } },
+{ "type": "Feature", "properties": { "address": "Laupenstrasse, 3008 Bern, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.4301286, 46.9476323 ] } },
+{ "type": "Feature", "properties": { "address": "Ankerstrasse, 8004 Zürich, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.5249697, 47.3741796 ] } },
+{ "type": "Feature", "properties": { "address": "Hauptstrasse, 4147 Aesch BL, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.5970561, 47.4689821 ] } },
+{ "type": "Feature", "properties": { "address": "Klausstrasse, 8008 Zürich, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.551488, 47.3591733 ] } },
+{ "type": "Feature", "properties": { "address": "Huobstrasse, 8808 Pfäffikon SZ, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.7809236, 47.2005643 ] } },
+{ "type": "Feature", "properties": { "address": "Luzernerstrasse, 6014 Luzern, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.2614142, 47.0491612 ] } },
+{ "type": "Feature", "properties": { "address": "Rue de Genève, 1225 Chêne-Bourg, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.1938107, 46.1947594 ] } },
+{ "type": "Feature", "properties": { "address": "Route Aloys-Fauquez, 1018 Lausanne, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.6334145, 46.5341791 ] } },
+{ "type": "Feature", "properties": { "address": "Zürcherstrasse, 8953 Dietikon, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.4029967, 47.4044452 ] } },
+{ "type": "Feature", "properties": { "address": "Wassergrabe, 6210 Sursee, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.1094776, 47.1833104 ] } },
+{ "type": "Feature", "properties": { "address": "Espace des Remparts, 1950 Sion, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.359018660370566, 46.23214475 ] } },
+{ "type": "Feature", "properties": { "address": "Avenue du Grey, 1004 Lausanne, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.6187142, 46.5327985 ] } },
+{ "type": "Feature", "properties": { "address": "Soodstrasse, 8134 Adliswil, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.5260873, 47.3134104 ] } },
+{ "type": "Feature", "properties": { "address": "Römerstrasse, 8404 Winterthur, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.7563255, 47.508088 ] } },
 { "type": "Feature", "properties": { "address": "Altgasse, 6340 Baar, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 8.521283, 47.1923375 ] } },
 { "type": "Feature", "properties": { "address": "Güterstrasse, 4053 Basel, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 7.5911933, 47.5440842 ] } },
 { "type": "Feature", "properties": { "address": "Chemin du Chêne, 1020 Renens VD, Switzerland" }, "geometry": { "type": "Point", "coordinates": [ 6.5932191, 46.5290259 ] } },

notebooks/zefix/zefix.ipynb

@@ -85,14 +85,15 @@
     "import folium\n",
     "import mapclassify\n",
     "import matplotlib as mpl\n",
-    "import matplotlib.pyplot as plt\n",
     "import pandas as pd\n",
     "import geopandas as gpd\n",
     "import plotly.express as px\n",
     "import plotly.graph_objects as go\n",
-    "from plotly.subplots import make_subplots\n",
     "\n",
     "from graphly.api_client import SparqlClient\n",
+    "from geopy.extra.rate_limiter import RateLimiter\n",
+    "from geopy.geocoders import Nominatim\n",
+    "from shapely.geometry import Point\n",
     "\n",
     "%matplotlib inline"
    ]
@@ -501,6 +502,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "geolocator = Nominatim(user_agent=\"zefix\")\n",
+    "find_location = RateLimiter(geolocator.geocode, min_delay_seconds=1)\n",
+    "\n",
     "def geocode(df: pd.DataFrame, cache_file: str=\"addresses.json\"):\n",
     "\n",
     "    if os.path.isfile(cache_file): \n",
@@ -513,8 +517,16 @@
     "    is_new_address = ~df.address.isin(cached_addresses)\n",
     "\n",
     "    if any(is_new_address):\n",
-    "        new_locations = gpd.tools.geocode(df.address[is_new_address], user_agent=\"zazuko\", provider=\"nominatim\")\n",
-    "        new_locations[\"address\"] = df.address[is_new_address]\n",
+    "        \n",
+    "        coordinates = []\n",
+    "        addresses = []\n",
+    "        for address in df.address[is_new_address]:\n",
+    "            location = find_location(address)\n",
+    "            if coordinates:\n",
+    "                coordinates.append(Point(location))\n",
+    "                addresses.append(address)\n",
+    "\n",
+    "        new_locations = gpd.GeoDataFrame.from_dict({\"address\": addresses, \"geometry\": coordinates})\n",
     "        new_locations = new_locations.set_crs(epsg=4326)\n",
     "        new_locations = new_locations[~new_locations.geometry.is_empty]\n",
     "        all_locations = gpd.GeoDataFrame(pd.concat([new_locations, cached_locations], ignore_index=True), crs=new_locations.crs)\n",
@@ -609,7 +621,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.0"
+   "version": "3.9.12"
   },
   "title": "Central Business Name Index - Zefix"
  },