Code for the Causal Bayesian Optimization algorithm (http://proceedings.mlr.press/v108/aglietti20a/aglietti20a.pdf)

The code to run the CBO algorthm is in runCBO.py. The file runBO.py can be used to run the non causal version of BO which is used as a benchmark in the paper. It is possible to change the causal graph that by changing the value of the variables 'experiment'. The results are saved in the folder Data/ when running the experiments. The folder Data/ contains the observational and intervational data used to produce the results in the paper.

Feel free to contact the first author of the paper (Virginia Aglietti) for questions