Make the bounding box for biomass data smaller #23
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
Collaborator
|
@treigerm I think this will mostly help with JPEG in the end since PCO and ZFP are upstream issues that I'm working on getting fixed. But if JPEG is really just an OOM problem then a small reduction would help. |
Collaborator
|
I’ll investigate a bit more and then we’ll see if we need this change, thanks for putting it up! |
Member
Author
|
Sounds good, of course if we can fix this otherwise that will be better! I just also needed a short term fix for now. |
Collaborator
|
Since we now have fixes for all compressors to work on the original size (ClimateBenchPress/compressor#44, ClimateBenchPress/compressor#45, ClimateBenchPress/compressor#46), I close this PR |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.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.
This slightly reduces the bounding box that determines what data is included in the benchmark. The uncompressed size of the biomass data is now 382MB. This makes all compressors pass instead of the PCO ones. This seems acceptable to me now and I'm hesitant to make the bounding box any smaller because we want to make sure we capture at least some variability even for the tiny benchmark.
Also given that you identified the issue with PCO (pcodec/pcodec#298) @juntyr , it is not necessarily an issue on our end.
Visualization of the bounding box:
