I know the general advice is to initialize Stan (or any MCMC algorithm) from widely dispersed values relative to the posterior. My understanding is this helps catch issues like multimodality or chains not mixing well, which might not occur if started from the same point. But if the idea is to find …

[image] monnahc: actually totally clears up what I should be looking for after I run my model wrt initial values (i.e., not as dispersed as I thought). So that’s helpful. Thanks. Great! If you come up with a solution you like, please post also here.

Overdispersed initial values -- general questions

Algorithms

avehtari May 26, 2018, 5:48pm 21

Great! If you come up with a solution you like, please post also here.

Topic		Replies	Views
Undesirable(?) behavior in initial values of hierarchical model General	9	1005	November 13, 2020
Is it fair/legitimate to use least square or optimization estimated parameter values to initialize the chain? Modeling techniques	6	828	April 8, 2021
Initilization with dispersion around a good guess Modeling initialization , diagnostics	1	617	January 9, 2023
Using output from optimization algorithms to initialize sampler Algorithms optimization , mcmc	6	1346	April 25, 2019
One MCMC chain not moving Modeling cognitive-science	7	2773	November 11, 2020

Overdispersed initial values -- general questions

Related topics