MAINT: Adapt to NumPy 2 promotion changes #16141
Conversation
| try: | ||
| is_safe = source_dtype.type(other) == other | ||
| except OverflowError: | ||
| is_safe = False |
There was a problem hiding this comment.
This just adds the OverflowError check, because NumPy is now strict, otherwise the error type would change here.
| # Go via NumPy to get the value | ||
| other = np.array(other) | ||
| if other.dtype.kind in "ifc": | ||
| other = other.item() |
There was a problem hiding this comment.
The biggest change, on NumPy 2, we need to pass Python scalars to align with Pandas (not NumPy, which pandas does not align with here).
On NumPy 1, this doesn't really make a difference.
To get those Python scalars, .item() is one reasonable way.
| @@ -341,7 +341,6 @@ def test_dtype(in_dtype, expect): | |||
| np.complex128, | |||
| complex, | |||
| "S", | |||
| "a", | |||
There was a problem hiding this comment.
Just avoids an error "a" has been removed, and seems not particularly important here.
mroeschke
added
improvement
|
/okay to test |
1 similar comment
|
/okay to test |
| def prinoptions(cls): | ||
| # TODO: NumPy now prints scalars as `np.int8(1)`, etc. this should | ||
| # be adapted evantually. | ||
| if np.lib.NumpyVersion(np.__version__) >= "2.0.0rc1": |
There was a problem hiding this comment.
Another suggestion to use packaging.version.parse if you recommend against np.lib usage
There was a problem hiding this comment.
Since it is used in many tests, adopted now (and rebased, but all except last commits are unchanged).
There was a problem hiding this comment.
An optional fix https://github.com/rapidsai/cudf/pull/16141/files#r1671422229 but overall looks good
Note that pandas `where` seems to promote the Series based on the value even with NumPy 2. This was never copied by cudf (i.e. an outstanding issue)
Pandas keeps using weak promotion even for strongly typed "scalars" (i.e. 0-d objects). This tries to (mostly) match that, but there may be better ways to do it. I am having difficulty to think of the best way though.
seberg
force-pushed
the
numpy2-compat-not-full
branch
from
66efccf
to
927904c
Compare
seberg
force-pushed
the
numpy2-compat-not-full
branch
from
927904c
to
6af0dec
Compare
|
/okay to test |
|
Friendly ping: I think this should be pretty safe to push through. There is the follow-up needed to ensure integer comparisons don't misbehave (the test suite currently doesn't catch this problem). |
|
/merge |
Splitting out the non API changes from gh-15897, the Scalar API change is required for the tests to pass with NumPy 2, but almost all changes should be relatively straight forward here on their own.
(I will add inline comments.)
This PR does not fix integer comparisons, there are currently no tests that run into these.
xref: rapidsai/build-planning#38