I have a very computationally expensive likelihood function. Using scipy’s maximum likelihood, I’m able to get a good estimate of my parameters. However, I’d like to get the distribution of those parameters and better statistics using NUTS. Is there a way to give NUTS my maximum likelihood estimates…

Feeding NUTS initialization of model parameters

evenmm April 19, 2023, 3:42pm 4

In pymc version 5.1.1 or newer, is there a way to specify the exact starting point like this?
For example to use advi with a different obj_optimizer than the default?

Topic		Replies	Views
Should find_MAP() be used for gaussian process models? Questions	1	622	May 22, 2020
Differenc between PyMC 3.1 and PyMC 3.0 on NUTS Questions	1	1297	August 28, 2017
About sampling start with find_MAP Questions	9	6327	May 24, 2024
Starting NUTS at specific value in pymc 5 v5 variational_inferenc	1	561	April 19, 2023
Feed the pm.fit results to NUTS sampler Questions	4	479	July 31, 2020

Feeding NUTS initialization of model parameters

Related topics