I\'ve been attempting to construct a Bayesian belief network in Python using Pomegranate, where most of the nodes are standard discrete probabilities and so are easy to mode