I want to run a weighted regression in PyMC but with a posterior distribution of weights I have precalculated on the same data. Say I have obtained a matrix W of N observations and Q draws (i.e., draws * chains). When calculating the regression likelihood, I want to multiply the normal log-pdf of t…

"Injecting" values into each MCMC iteration

ricardoV94 November 8, 2023, 11:40am 3

You need a custom step sampler if you want to change the logp artifically at every mcmc step

Topic		Replies	Views
How to perform weighted inference Questions	8	5186	August 30, 2021
How to run logistic regression with weighted samples Questions linear_model	10	3442	August 30, 2021
Using a custom step method for latent mixing variables Questions	4	1733	April 16, 2018
How to set up a custom likelihood function for two variables Questions	8	13306	March 9, 2018
Getting observations of distribution variables whilst sampling Questions from_github	2	1008	June 15, 2017