[SPARK-59162][ML] Replace BLAS calls with while loops in GBT models - #58461
Closed
zhengruifeng wants to merge 1 commit into
Closed
[SPARK-59162][ML] Replace BLAS calls with while loops in GBT models#58461zhengruifeng wants to merge 1 commit into
zhengruifeng wants to merge 1 commit into
Conversation
HyukjinKwon
approved these changes
Sep 2, 2026
zhengruifeng
added a commit
that referenced
this pull request
Sep 2, 2026
### What changes were proposed in this pull request? Replace the BLAS dot products in `GBTClassificationModel.margin` and `GBTRegressionModel.predict` with indexed `while` loops that accumulate each tree prediction and weight directly. Remove the now-unused BLAS imports. ### Why are the changes needed? The BLAS calls first allocate an array containing every tree prediction and then invoke native BLAS for a small dot product on every model prediction. Direct accumulation avoids the temporary array and native-call overhead, reducing prediction overhead and Spark Connect server pressure. ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? The existing GBT classifier and regressor suites passed: ``` build/sbt 'mllib/testOnly org.apache.spark.ml.classification.GBTClassifierSuite' build/sbt 'mllib/testOnly org.apache.spark.ml.regression.GBTRegressorSuite' ``` ### Was this patch authored or co-authored using generative AI tooling? Generated-by: OpenAI Codex (GPT-5) Closes #58461 from zhengruifeng/ml-gbt-while-loop-dev-6. Authored-by: Ruifeng Zheng <ruifengz@apache.org> Signed-off-by: Ruifeng Zheng <ruifengz@apache.org> (cherry picked from commit 9b81792) Signed-off-by: Ruifeng Zheng <ruifengz@apache.org>
Contributor
Author
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
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.
What changes were proposed in this pull request?
Replace the BLAS dot products in
GBTClassificationModel.marginandGBTRegressionModel.predictwith indexedwhileloops that accumulate each tree prediction andweight directly. Remove the now-unused BLAS imports.
Why are the changes needed?
The BLAS calls first allocate an array containing every tree prediction and then invoke native BLAS
for a small dot product on every model prediction. Direct accumulation avoids the temporary array
and native-call overhead, reducing prediction overhead and Spark Connect server pressure.
Does this PR introduce any user-facing change?
No.
How was this patch tested?
The existing GBT classifier and regressor suites passed:
Was this patch authored or co-authored using generative AI tooling?
Generated-by: OpenAI Codex (GPT-5)