Tune-up chunking #47
Merged
Tune-up chunking #47
Conversation
This file contains 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
also set dtype to int16 (in case it is related to the black bars bug)
using the fsstore was giving us all zeroed niftis. not sure why
To hopefully address performance issues in the bigstitcher step when dealing with large datasets. Also sets dtype to int16 instead of uint16 for this only -- some imagesc posts describing the black bars had this as a fix..
and moves rechunking from zarr2bdv to apply_flatfield_corr should hopefully improve performance
since that likely wasn't correcting the problem
akhanf
added a commit
that referenced
this pull request
Sep 14, 2024
Needs to be 128x128x128 or 256x256x256 instead of 1x4096x4096. The latter causes memory issues as described in #47
akhanf
added a commit
that referenced
this pull request
Sep 14, 2024
* chunk size (pre bigstitcher) wasn't set correctly Needs to be 128x128x128 or 256x256x256 instead of 1x4096x4096. The latter causes memory issues as described in #47
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Outdated suggestions cannot be applied.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 makes changes to the default chunking, to fix issues happening with bigstitcher and fusion where too many threads and too large chunks were causing memory and garbage collection issues that were stalling the run..
Problem is likely because we had large-ish 2d chunks, but the processing was requiring 3d context, so ultimately too much memory was being allocated when tryibng to load the 3d data..
This also makes some fixes/updates to the tif stack to zarr