[ML] Avoid log spam when we only have missing values for a feature computing candidate splits for regression and classification #1500
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
…ting candidate splits
tveasey
added a commit
to tveasey/ml-cpp-1
that referenced
this pull request
Sep 25, 2020
…mputing candidate splits for regression and classification (elastic#1500) Since we downsample the rows when computing candidate splits it's possible that a feature with non-zero probability of being selected ends up with no non-missing feature values when we compute candidate splits. This is harmless and we can happily initialise the candidate splits to an empty set in this case. However, it was generating log spam when trying to compute quantiles. In particular, we'd get repeated errors and warnings during training of the form: [CBoostedTreeImpl.cc@737] Failed to compute quantile 86.7027: ignoring split [CQuantileSketch.cc@295] No values added to quantile sketch This change simply checks that there are values before trying to compute quantiles.
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Outdated suggestions cannot be applied.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.
Since we downsample the rows when computing candidate splits it's possible that a feature with non-zero probability of being selected ends up with no non-missing feature values when we compute candidate splits. This is harmless and we can happily initialise the candidate splits to an empty set in this case. However, it was generating log spam when trying to compute quantiles. In particular, we'd get repeated errors and warnings during training of the form:
This change simply checks that there are values before trying to compute quantiles.