How to use a Deterministic in a mixture?

[narendramukherjee]

I think the issue is that pm.DensityDist needs an observed value. Look at this discussion:

How to set up a custom likelihood function for two variables

In my case, likelihood function was defined as f(theta_1, theta_2), the expression is already known not a standard distribution. Prior of theta_1 and theta_2 are following lognormal distribution. How can I get the posterior distribution with certain sampling method? My code with pm.Model() as model: theta_1 = pm.Lognormal('theta_1', mu=mu_1, sd=sigma_1) theta_2 = pm.Lognormal('theta_2', mu=mu_2, sd=sigma_2) like = pm.DensityDist('like', likelihood(theta_1,theta_2),shape=(…