Improve parallelization: threading backend, PI batching, auto-selection#91
Merged
Merged
Conversation
- Switch run_parallel from loky to threading backend — eliminates pickling overhead for feature-level parallelism. Sklearn predict releases the GIL, so threads scale well (tested up to 10 cores). - Add auto-selection: skip parallel overhead when task count < 3 - Batch n_permute predict calls in permutation importance into single predict call (PI 10v/10p: 78.6s → 64.2s serial, 19.1s at n_jobs=4) - Optimize PI column reassembly with numpy in-place column swap instead of pd.concat column-by-column extraction - Add stress test to benchmark suite (10K samples, 30 features, 100 trees) Scaling results (10K samples, 30 features, 100-tree RF): ALE (30 feat, 1 boot): 1.67s → 0.36s at n_jobs=10 (4.7x) ALE (30 feat, 10 boot): 16.3s → 2.7s at n_jobs=10 (6.1x) PD (5 feat, 10 boot): 29.7s → 7.4s at n_jobs=5 (4.0x) PI (10 vars, 10 perm): 66.4s → 19.1s at n_jobs=4 (3.5x) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Scaling results (10K samples, 30 features, 100-tree RF):
ALE (30 feat, 1 boot): 1.67s → 0.36s at n_jobs=10 (4.7x)
ALE (30 feat, 10 boot): 16.3s → 2.7s at n_jobs=10 (6.1x)
PD (5 feat, 10 boot): 29.7s → 7.4s at n_jobs=5 (4.0x)
PI (10 vars, 10 perm): 66.4s → 19.1s at n_jobs=4 (3.5x)