DEPR: ArrayManager #55044
DEPR: ArrayManager #55044
Conversation
…_array_manager # Conflicts: # doc/source/whatsnew/v2.2.0.rst
…_array_manager
|
I'm seeing warnings related to the dateutil package upon import of pandas; is this expected? https://github.com/pandas-dev/pandas/actions/runs/6177889227/job/16769987130?pr=55044#step:5:92
|
…ndas into depr_array_manager # Conflicts: # pandas/tests/internals/test_managers.py
…_array_manager
I'm not surprised by the warning itself, but I'd expect it to be showing up on all builds, not specific to this PR |
Thanks - maybe it's only incurred by using subprocess like the test added here. And even then, it only seems to happen in certain jobs. |
| env=test_env, | ||
| check=True, | ||
| ) | ||
| msg = "<string>:1: FutureWarning: Using ArrayManger through the environment" |
There was a problem hiding this comment.
typo ArrayManger -> ArrayManager
| msg = "<string>:1: FutureWarning: Using ArrayManger through the environment" | ||
| stderr_msg = response.stderr.decode("utf-8") | ||
| if manager == "block": | ||
| assert msg not in stderr_msg, stderr_msg |
There was a problem hiding this comment.
do we want to tell them that the environment variable is deprecated even if they are using "block"?
There was a problem hiding this comment.
I could go either way here - I anticipate it would be quite rare people are using the env variable, especially to specify block. Any preference?
There was a problem hiding this comment.
very slight preference for warning but NBD
|
handful of comments, generally looks good. the complexity of this deprecation (as opposed to just warning inside |
…_array_manager � Conflicts: � doc/source/whatsnew/v2.2.0.rst
|
Thanks @rhshadrach |
* DEPR: ArrayManager * Fixup * Test fixup * debug CI * Test fixup * warn if PANDAS_DATA_MANAGER is set * single_cpu
doc/source/whatsnew/vX.X.X.rstfile if fixing a bug or adding a new feature.