Issue with observed data in joint distribution

[narendramukherjee]

Skimming through your posts, I think you are trying to pull the NormalMixture log likelihood into your DensityDist logp. In that case, you shouldn’t be using pm.NormalMixture, but instead should use pm.NormalMixture.dist and use its logp method. That lets you make a pymc3 distribution object and use it to get the log likelihood of that distribution. Something like this:

What is pm.Normal.dist.logp?

or

Declare an internal Poisson distribution inside a custom distribution failes

It’s a bit difficult to give you advice without knowing what you are trying to achieve. But at least if you are trying to make the above code running without error you can rewrite the logp as: def logp(value): ipoisson = pm.Poisson.dist(mu = mu) return ipoisson.logp(value)