I’m finally doing some work which requires some genuine Bayesian analysis, and so have returned to playing with emcee. I’ve also been looking at PyMC3 which is an impressive piece of work, but also requires a bit of change of thinking from emcee.

Some notebooks can be found on GitHub.