Here is some starter code for the MS&E 338 Reinforcement Learning implementation project. In particular, we wanted to make sure that you knew how to update and maintain sufficient statistics for the posterior distributions of PSRL. All other algorithms should be relatively trivial to implement.
posterior_sampling.py: updating and sampling from posterior distributions.example_posterior.py: example use and demonstrations of these functions.