Hierarchical Bayesian Inference (Extended)

In this tutorial we show how to perform hierarchical Bayesian inference.

Hierarchical Bayesian Inference is set up in 3 phases:
  • sample the posterior distributions conditioned on each data set

  • sample the hyper parameter

  • sample the posterior given hyperparameter and one (a) data set or (b) data sets combined.

For each phase we provided an individual python file.

In phase 0 we generate synthetic data.


All 3 phases can be run with the shell script


The results are saved in sub dirs of the folder /setup.