Hi, I have implemented a blackbox likelihood function in my model, using the framework described in Using a “black box” likelihood function (numpy) — PyMC example gallery In this code, there is a user-defined log likelihood function: def my_loglike(theta, x, data, sigma): model = my_model(the…

The example shouldn’t have data and x be passed to __init__ which make is unnecessarily restrictive and doesn’t really match the CustomDist API. I’ll update the example, but for now your logp function would look something like: def logp(value, x, theta): # `value` and `x` are ignored because they…

Explanation of logp and customDist for Blackbox likelihood functions

Questions v5

ricardoV94 January 30, 2024, 8:30am 4

Yup

Topic		Replies	Views
Pointwise log-likelihood for loo/waic with custom "black-box" probability distribution v5	9	790	January 19, 2024
Likelihood evaluation and DensityDist Questions	2	1709	February 21, 2018
Likelihood Specification and DensityDist Questions doc	3	1025	October 5, 2021
How to set up a custom likelihood function for two variables Questions	8	13306	March 9, 2018
Using pm.DensityDist and customized likelihood with a black-box function Questions	10	6853	August 29, 2018

Explanation of logp and customDist for Blackbox likelihood functions

Related topics