Multinomial hierarchical regression with multiple observations per group ("Bad energy issue")

[junpenglao]

Have you try the solution in FAQ?

Frequently Asked Questions

Getting NaN or Inf during NUTS or VI (e.g., 'Bad initial energy: inf. The model might be misspecified., NaN occurred in optimization.)

Getting 'Bad initial energy: inf' when trying to sample simple model

Hi @Stargrazer82301, this is actually one of our FAQs. You can have a look at the answer here: For example, in your case, running: for RV in model.basic_RVs: print(RV.name, RV.logp(model.test_point)) gives: Intercept 1.1370228148748596 Gradient -inf Psi_log__ -0.7257913526447277 z_likelihood 73.49976264100106 And looking closer at gradient: gradient = pm.Normal('Gradient', mu=leastsq_fit[0], sd=leastsq_fit[0], testval=leastsq_fit[0]) You can see that sd is defined as leastsq_fit[0], …

Also, some times there is problem with the init method, so try trace = pm.sample(..., init='adapt_diag') as well.