Notebooks illustrating estimating the posterior in a Bayesian logistic regression problem using the Markov chain Monte Carlo methods implemented in Mici and PyStan. ArviZ is used to compute chain summary statistics and Corner to plot pairwise posterior marginals.
The notebooks should be runnable from a Python 3.8+ environment with the packages listed in the requirements.txt file installed.
To install the requirements in the current environment using pip run
pip install -r requirements.txt
from the root directory of a clone of the repository.
A Jupyter Lab instance can then be launched to allow viewing and running the notebooks locally by running
jupyter lab