Clean up obsolete PANDAS_GE markers#12279
Conversation
| # These warnings appear only on pandas >=2.1,<3.0 | ||
| @pytest.mark.filterwarnings(r"ignore:.*(DataFrame|Series)GroupBy\.sum:FutureWarning") | ||
| @pytest.mark.filterwarnings(r"ignore:.*DataFrame\.sum with axis=None:FutureWarning") | ||
| @pytest.mark.filterwarnings(r"ignore:.*DataFrameGroupBy\.apply:DeprecationWarning") | ||
| @pytest.mark.filterwarnings(r"ignore:.*DataFrameGroupBy\.apply:FutureWarning") |
There was a problem hiding this comment.
My first intuition was that this may indicate that, somewhere in the codebase, there may be a conditional code path that incorrectly uses PANDAS_GE_300 instead of PANDAS_GE_210. However, I could not find it.
|
|
||
|
|
||
| @contextlib.contextmanager | ||
| def check_reductions_runtime_warning(): |
Unit Test ResultsSee test report for an extended history of previous test failures. This is useful for diagnosing flaky tests. 21 files ±0 21 suites ±0 5h 34m 55s ⏱️ + 6m 4s Results for commit 4d75652. ± Comparison against base commit 45610ac. ♻️ This comment has been updated with latest results. |
|
Test failure is unrelated (#12225). |
1060871 to
9fd5a9e
Compare
9fd5a9e to
4d75652
Compare
|
@jsignell Would you have time to review this? |
|
If not maybe @TomAugspurger |
jsignell
left a comment
There was a problem hiding this comment.
This seems fine to me! I do think if you are following NEP 29 you can drop any pandas versions less than Pandas 2.2.1
Is Dask following NEP29 / SPEC0 though? I didn't think it did. |
We don't proactively follow them, but we can use them to justify dropping older versions whenever there is a maintenance benefit. |
PANDAS_GE_201andPANDAS_GE_211PANDAS_GE_300test_groupby_transform_funcsin pandas >=2.1,<3.0