# Clone via ssh
git clone [email protected]:vroland/obstsalat-odc19.git
# Create virtual environment
python -m venv env
# Activate environment
source env/bin/activate
# Install dependencies
pip install -r requirements.txt
- clone the openrouteservice repository
- To disable openrouteservice rate limiting, apply
./openrouteservice_config.patchto the sample config of the openrouteservice docker container. - Download an appropriate open streetmap dataset (e.g. saxony)
- delete everything in
docker/graphs! This will force openrouteservice to rebuild the map data. Otherwise, you might get unexpectedError Code 2099errors, because the requested coordinates are off map. Running docker compose without the-dflag (in foreground mode) helps catch any out of memory exceptions. - make sure the container has enough RAM available. Increase RAM limit as described in https://github.com/GIScience/openrouteservice/blob/master/docker/README.md. Using a VM with 3G of ram did not work for me, while 6G worked just fine.
- Make sure you have openrouteservice running under localhost (follow https://github.com/GIScience/openrouteservice/blob/master/docker/README.md and load the dataset for saxony)
- Start the backend service:
cd obst/tsopt/ && python3 optimize.py. It should load the pre-computed dataset graphs. - Open the front-end file in
obst/ui/index.html