sumo-tasks-pipeline 3.0

Simple Python interface for a traffic simulator: SUMO

sumo-tasks-pipeline

---

Run SUMO simulators as easy as possible!

The package sumo-docker-pipeline enables you to run a traffic simulator SUMO efficiently and to interact with Python easily.

Example

Just three lines to run a SUMO simulation.

from sumo_tasks_pipeline import LocalSumoController, SumoConfigObject
from pathlib import Path

path_config = Path().cwd().parent.joinpath('tests/resources/config_complete')
sumo_controller = LocalSumoController(sumo_command='/usr/local/bin/sumo')
sumo_config = SumoConfigObject(scenario_name='example', path_config_dir=path_config, config_name='grid.sumo.cfg')
sumo_result_obj = sumo_controller.start_job(sumo_config)

See examples directory to know more.

Features

• Possible to resume your tasks. The feature is useful when you run simulators on Google Colab.
• Possible to save SUMO simulation result to Google Cloud Storage (GCS). No worries even when your local storage is small.
• Possible to run SUMO simulations with multiple machines if you use GCS as the storage backend.

Requirement

• python > 3.5
• docker
• docker-compose

Install

Pull the image (or build of a docker image with SUMO)

The existing image is on the Dockerhub.

docker pull kensukemi/sumo-ubuntu18

If you prefer to build with yourself, you run the following command.

docker-compose build 

Install a python package

make install

For developer

pytest tests

license and credit

The source code is licensed MIT. The website content is licensed CC BY 4.0.

@misc{sumo-docker-pipeline,
  author = {Kensuke Mitsuzawa},
  title = {sumo-docker-pipeline},
  year = {2021},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/Kensuke-Mitsuzawa/sumo_docker_pipeline}},
}