Support multidimensional boolean set/inc_subtensor in Numba via rewrite #1108
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
Setting to false can lead to slower code on C/Numba backend which don't support np.add.at natively.
ricardoV94
force-pushed
the
numba_bool_adv_set_subtensor
branch
from
f94a44c to
a982af9
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #1108 +/- ##
==========================================
- Coverage 82.14% 82.12% -0.03%
==========================================
Files 183 183
Lines 48104 48111 +7
Branches 8665 8667 +2
==========================================
- Hits 39517 39510 -7
- Misses 6419 6435 +16
+ Partials 2168 2166 -2
|
…'t support np.add.at natively. Support multidimensional boolean set/inc_subtensor in Numba via rewrite
ricardoV94
force-pushed
the
numba_bool_adv_set_subtensor
branch
from
a982af9 to
8dc6727
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Outdated suggestions cannot be applied.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.
Follow up to #818
It's simple to support multidimensional boolean set_subtensor. The corresponding integer indexing would be slightly trickier because it requires reshaping y as well.
Although there are still cases that are not covered outside of object mode (np.newaxis, non-contiguous adv index, broadcasted integer indexing, mixed boolean/integer indexing) I think it's fine to close #772. Efforts can be directed on a per-need basis. Numpy indexing is just too wild :)
📚 Documentation preview 📚: https://pytensor--1108.org.readthedocs.build/en/1108/