Hi, I have a very complicated and computationally expensive log-likelihood function from which I am attempting to generate a posterior using NUTS. To make my log-likelihood (and gradient calculations) computationally feasible, I utilize multiprocessing.pools. I’d like to use NUTS to generate a post…

NUTS and parallelized black box likelihood

zacmon October 22, 2019, 9:47pm 3

Thank you for the reply and information!

Topic		Replies	Views
NUTS uses all cores Questions	18	6542	May 21, 2020
NUT sampler stuck under windows with njobs>1 Questions	13	4376	April 27, 2020
Problem with the "Getting started with PyMC3" tutorial Questions	4	1760	December 8, 2018
Error during run sampling method Questions	5	3809	January 14, 2019
Sampling doesn't start when njobs > 1 for some models Questions	20	4831	January 17, 2020