DPPL 0.39 #2715
Draft
DPPL 0.39 #2715
Conversation
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
Contributor
|
Turing.jl documentation for PR #2715 is available at: |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
This PR uses fast evaluation where it can. MCMC sampling mostly works. There are a few outstanding problems:
Chains no longer store an overall logevidence. I think that can be patched upstream in DynamicPPL, basically check if all the stats have a
.logevidencefield and if so, set the chain's logevidence to be the mean of that. This does feel a bit lame because the user should just be able to do the mean themselves, but I guess not doing this would be gratuitous breakage. The issue is that this works fine for SMC and PG (I think), but IS. Not sure how important IS is.Chains aren't consistent between
:lpand:logjoint, because I messed this up: ParamsWithStats:lporlogjoint? DynamicPPL.jl#1140 So some samplers will have one and other samplers the other. This needs to be fixed upstream.I didn't update the optimisation interface. I think we should merge or reject Remove Optim.jl interface + minor tidying up of src/optimisation/Optimisation #2708 first.