Repository navigation
FEAT: discard dynesty mcmc chains that are too short - #1154
Open
ColmTalbot wants to merge 6 commits into
Open
ColmTalbot wants to merge 6 commits into
ColmTalbot wants to merge 6 commits into
Conversation
ColmTalbot
added
<10 lines
sampling
Collaborator
|
I can't see a button to approve, but I am happy with it going in as a "debug" statement for now. |
Also, fix string formatting for log messages
ColmTalbot
force-pushed
the
dynesty-block-too-short-mcmc
branch
from
October 21, 2026 16:18
2e73449 to
2a65016
Compare
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Following from #1153 at the beginning of the MCMC stage if
walksis set to a low value, the first MCMC chains are too short and can be unreliable (see the little spike near ~180).We can check after the MCMC finishes but before the point is returned if the actual number of accepted steps follows the correct distribution (binomial, N=walks, p=naccept/walks). This shouldn't increase the cost significantly (at least not beyond a working case), since by definition we need to run a longer MCMC in this case.
With this change, that goes away.
I'm not 100% about whether this is too strict and can introduce some stealth bias. If this is the case, we might want a flag to only do this during a warm up phase and not later in the run.