Support column-split in row partitioner #8828
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
trivialfis
reviewed
Feb 23, 2023
src/common/partition_builder.h
Outdated
| auto const& cut_values = gmat.cut.Values(); | ||
| auto const& cut_ptrs = gmat.cut.Ptrs(); | ||
|
|
||
| auto gidx_calc = [&](auto ridx) { |
There was a problem hiding this comment.
Please consider some refactoring to reuse this function with partition(), maybe with GHistIndexMatrix::GetFValue as well.
| @@ -205,6 +244,14 @@ class CommonRowPartitioner { | |||
| ctx, tree, this->Partitions(), p_out_position, | |||
| [&](size_t idx) -> bool { return gpair[idx].GetHess() - .0f == .0f; }); | |||
| } | |||
|
|
|||
| private: | |||
There was a problem hiding this comment.
Maybe we should extract these into a different class or struct with its own utility member functions? Also, the CommonRowPartitioner has private members, please place them together.
trivialfis
approved these changes
Feb 25, 2023
1 task
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Outdated suggestions cannot be applied.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.
The approach is similar to what's done in
cpu_predictor.cc: first collect decision/missing bits, then aggregate them across all the workers, and finally use them to do the partitioning.For now only support the
approxmethod.