A Bayesian Optimisation package for ROS developed by the Intelligent Control Systems (ICS) group at ETH Zurich.

We provide a official and up-to-date documentation here.

For a short tutorial on how to get started with the BayesOpt4ROS package, please see the official documentation.

In case you find our package helpful and want to contribute, please either raise an issue or directly make a pull request.

• We follow the NumPy convention for docstrings
• We use the Black formatter with the --line-length option set to 88

To facilitate easier contribution, we have a continuous integration pipeline set up using Github Actions. To run the tests locally you can use the following commands:

pytest src/bayesopt4ros/test/unit/

To run all integration tests (this command is executed in the CI pipeline):

catkin_make_isolated -j1 --catkin-make-args run_tests && catkin_test_results

Or if you want to just run a specific integration test:

rostest bayesopt4ros test_client_python 

If you found our package helpful in your research, we would appreciate if you cite it as follows:

@misc {Froehlich2021BayesOpt4ROS, 
  author       = {Fr\"ohlich, Lukas P. and Carron, Andrea},
  title        = {BayesOpt4ROS: A {B}ayesian Optimization package for the Robot Operating System},
  howpublished = {\url{ https://github.com/IntelligentControlSystems/bayesopt4ros}},
  year         = {2021},
  note         = {DOI 10.5281/zenodo.5560966},
}