ENH: speed up _harmonize_tensor #603
Merged
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
Common case is just a single element, thus that the overhead of NumPy is very large.
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## develop #603 +/- ##
========================================
Coverage 76.62% 76.62%
========================================
Files 72 72
Lines 9672 9672
========================================
Hits 7411 7411
Misses 2261 2261
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
Thanks @DerWeh -- I was wondering why merging EBMs had become so slow in more recent versions. |
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Outdated suggestions cannot be applied.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.
As far as I can see, cell_map is typically a list with a single tuple. This function call incurs tremendous function overhead for doing nothing (multiplying a single number with 1): first we generate a list, then go through all the NumPy dispatching.
If this indeed is the common case, it's much faster to avoid creating a list (using a generator) and using the simple Python math.prod.
Equivalently, we could use
This is mostly a matter of taste (the performance is identical).
For the simple case of merging a model with itself, this yielded a speed-up of the factor 4 for me. Can you please comment @paulbkoch if my assumptions are correct, or if there are use cases I am overlooking?
This speed up is part of PR #578. It doesn't seem that I'll find the time soon to fix the remaining bug in
_harmonize_tensor. So I thought it's a good idea to include this performance improvement right away.