Deterministic function of parameters coming from different models

Hello all,

My question is not limited to PyMC3, and is more general. Say I have two models characterized by two parameters \theta and \lambda, infered from two unrelated datasets. I run my sampler twice and get two traces for the variables. How do I combine these traces to estimate, say, \theta * \lambda?


If the model are nested and the number of observation is identical in both model you can multiple the two directly.