Faster implementation of Join in C backend #1340
Draft
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
ricardoV94
force-pushed
the
faster_join_impl_c_backend
branch
from
fa3b2cf to
d8efc2f
Compare
Reuse outputs when possible, and avoid numpy overhead on public facing function
ricardoV94
force-pushed
the
faster_join_impl_c_backend
branch
from
d8efc2f to
9b81d41
Compare
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Outdated suggestions cannot be applied.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 numpy C-API function has a lot of overhead, as it is made to work with arbitrary types of sequences from Python, plus arguments we don't care about.
Also importantly it doesn't allow one to manage their own buffer, which is useful when we have GC disabled in the C-backend.
I've saw 2x local speedup for small joins (where the copy overhead is smaller and we are bounded by the overhead).
Also simplified the implementation by removing the exotic
view_flag, which closes #753📚 Documentation preview 📚: https://pytensor--1340.org.readthedocs.build/en/1340/