The src folder contains a bunch of scripts to develop, benchmark,
debug or help in using VROOM.
- plot generates a simple svg visualisation for a solution file.
- overpass_to_json generates a problem from all OSM nodes with a
specific
key=valuetag in a chosen bounding box or city. - random_problem generates a ready-to-solve random problem based on a bounding box.
- global_indicators collects indicators on all solutions contained in a folder.
- vehicle_stats collects indicators on vehicle usage in a solution.
- json_to_csv dumps all locations coordinates from a json input instance to a csv file.
- tsplib_to_json converts a TSPLIB file to json.
- cvrplib_to_json converts a CVRPLIB file to json.
- vrptw_to_json converts a VRPTW file to json.
- pdptw_to_json converts a PDPTW file to json.
- hvrp_to_json converts a HVRP file to json.
- add_osrm_matrix creates a "standalone" version of a json input
instance by adding a
matriceskey using OSRM. - add_ors_matrix creates a "standalone" version of a json input
instance by adding a
matriceskey using Openrouteservice. - asap reach out to a solution minimizing completion time or a set of trade-offs between completion time and cost.
The benchmarks folder contains everything required to reproduce
results on literature benchmarks with a single command. Detailed
instructions for:
Contributions on existing (or new) scripts and benchmark classes are highly appreciated! Feel free to use the bugtracker for any question, especially if you're unsure about potential changes or additions.
We use:
-
blackfor python code formatting in order to forget about how the code looks like and focus on what it does. -
flake8for style guide enforcement and PEP8 compliance.
A convenient way to automate the whole workflow is to setup a
pre-commit hook. pre-commit can do that
for you based on our .pre-commit-config.yaml config file. Simply
install and run once:
pip install pre-commit
pre-commit install