Repository navigation
Avoid repeated sorting and observation copies in TPE model construction - #281
Merged
Merged
Conversation
…struction TPE already prepares observation columns in sorted parameter order. Building the good and poor models sorts the same names again and creates temporary observation and distribution maps. Categorical values are also copied into temporary float buffers before conversion to category indices. Pass one sorted view of names, distributions, and observation columns to both model builders. Construct category-index buffers directly from the selected rows and retain the parameter order through mixture construction. This removes repeated sorting, temporary maps, and intermediate copies.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Reuse the sorted observation view when building TPE models, avoiding repeated parameter-name sorting and temporary observation/distribution maps. Build categorical index buffers directly from selected rows while sharing model construction with the named-observation API and retaining borrowing for numerical slices.
Closes #270.
Original implementation by @tkng, cherry-picked from 0ec3628 without code changes, preserving the original commit author.