I have a question about implementing a Hierarchical model in PyMC using a Black Box loglikelihood function. Specifically, I want to make sure that PyMC is receiving correct inputs from the likelihood function. I have a model with 4 parameters, with one partially pooled across three data sets. I h…

You can convert every item to a pytensor variable first and then stack them together (if they have compatible shapes). You can also use a scan, nothing wrong with it. But when you can use single arrays and treat data in batches it’s usually more performant because of vectorization. That of course …

Hierarchical Model with Black Box Loglikelihood

Questions version agnostic

ricardoV94 February 15, 2023, 8:18am 4

Can’t you make your potential accept higher dimensionality data? Then if things have the same shapes you could just pass the whole block of parameters instead of slicing it 50 times.

Topic		Replies	Views
Building a hierarchical model using a Black Box loglikelihood function version agnostic	1	447	September 20, 2022
First attempt at a hierarchical model with custom likelihood Questions	1	709	September 18, 2018
Hierarchical modeling with a black box likelihood function Questions	2	906	April 3, 2019
pm.Potential without Theano? Questions	14	2301	March 20, 2018
Re-coding a partial pooling from pymc to pymc3 Questions	1	467	November 2, 2018

Hierarchical Model with Black Box Loglikelihood

Related topics