I am trying to use PYMC3 for a Bayesian model where I would like to repeatedly train my model on new unseen data. I am thinking I would need to update the priors with the po
