This project provides some scripts to compare stops from official data sources (e.g. from DELFI e.V.) with stops extracted from OpenStreetMap.
Currently supported are the following stop sources:
- stops list Nahverkehrsgesellschaft Baden-Württemberg (NVBW)
- DELFI's zentrales Haltestellenverzeichnis (zHV)
- a GTFS file (Beta, developed for bwgesamt.zip, currently code makes various assumptions regarding the content )
To compare transit stops with osm data, you need
- an OSM pbf file covering the transit area, e.g. http://download.geofabrik.de/europe/germany/baden-wuerttemberg-latest.osm.pbf
- a stops file, e.g. curl --output zhv.zip -L https://scraped.data.public-transport.earth/de/zhv.zip
- a gtfs file e.g. curl --output gtfs-germany.zip -L https://data.public-transport.earth/gtfs/de
These can be downloaded e.g. with the provided scripts download_delfi_and_osm_data.sh
or download_nvbw_and_osm_data.sh
.
see e.g. https://medium.com/@carusot42/installing-and-loading-spatialite-on-macos-28bf677f0436
brew install sqlite3
find /usr/local -path "*sqlite3" | grep -e "sqlite3$" | grep "/bin"
/usr/local/Cellar/sqlite/3.39.2/bin/sqlite3 out/stops-bw.db
Optionally, you might want to create a python environment to install needed libraries in an isolated environment:
makevirtualenv nosc
Then, you should install the required python libraries:
pip install -r requirements.txt
The GTFS/NVBW Stops comparison supports three official stop sources for comparison against OSM:
- DELFI's stop register (zHV)
- NVBW's stop register (zHV)
- GTFS feeds
python3 compare_stops.py -o data/germany-latest.osm.pbf -g data/gtfs-germany.zip -s data/zhv.csv -p DELFI -d out/stops.db
or, to just re-run the matching without reload OSM:
python3 compare_stops.py -g data/gtfs-germany.zip -s data/zhv.csv -p DELFI -d out/stops.db
or, to just re-run the matching without reloading any files, specify mode ( -m ) match
:
python3 compare_stops.py -m match -p DELFI -d out/stops.db
python3 compare_stops.py -o data/baden-wuerttemberg-latest.osm.pbf -g data/gtfs-bw.zip -s data/zhv-bw.csv -d out/stops-bw.db -p NVBW -l out/matching.nvbw.log
The comparison option GTFS
compares stops provide by a GTFS-Feed to OpenStreetMap.
Note, that we currently apply some very NVBW-feed specific processing to these stops, which should be configurable and extendable for other feeds.
python3 compare_stops.py -o data/baden-wuerttemberg-latest.osm.pbf -g data/gtfs-bw.zip -d out/stops-bw.db -p GTFS
python3 compare_stops.py -o data/baden-wuerttemberg-latest.osm.pbf -g data/gtfs-bw.zip -s data/zhv-bw.csv -d out/stops-bw.db -p NVBW -d out/stops-bw.db
This script will create a spatialite database, import the official stops and the osm stops, calculate probable match candidates and finally pick the most probable matches.
If you want to use this project via docker, proceed as follows:
Build the docker image via
docker build -t mfdz/gtfs-osm-stops-matcher .
Run the comparison via
docker run --rm -v $(PWD)/data:/usr/src/app/data -v $(PWD)/out:/usr/src/app/out mfdz/gtfs-osm-stops-matcher python compare_stops.py -o data/germany-latest.osm.pbf -s data/zhv.csv -g data/gtfs-germany.zip -p DELFI
docker run --rm -v $(PWD)/data:/usr/src/app/data -v $(PWD)/out:/usr/src/app/out mfdz/gtfs-osm-stops-matcher python compare_stops.py -g data/gtfs-germany.zip -p DELFI
Run the tests by executing one of the commands respectively:
python3 -m unittest tests/test_stop_matcher.py
python3 -m unittest tests/test_match_picker.py