TYP: type NDFrame.(_get_axis|_get_axis_name|_get_axis_number) #33610
Conversation
| except AttributeError as err: | ||
| raise TypeError("Index must be DatetimeIndex") from err | ||
| if not isinstance(index, DatetimeIndex): | ||
| raise TypeError("Index must be DatetimeIndex") |
| name = self._get_axis_name(axis) | ||
| return getattr(self, name) | ||
|
|
||
| @classmethod | ||
| def _get_block_manager_axis(cls, axis): | ||
| def _get_block_manager_axis(cls, axis) -> int: |
There was a problem hiding this comment.
can "axis" be typed as Axis?
There was a problem hiding this comment.
Yeah, but in some cases axis is returned, so that that would require some refactoring inside the function, which I'd like to do seperately. For example I'm pretty sure _AXIS_ALIASES can be removed from the code base.
|
LGTM; @simonjayhawkins might have an opinion about the "type: ignore" |
There was a problem hiding this comment.
Thanks @topper-123 generally lgtm. can you add comments when adding #type: ignore to help reviewers and help track.
pandas/core/generic.py
Outdated
| @@ -3490,6 +3490,7 @@ class animal locomotion | |||
| axis = self._get_axis_number(axis) | |||
| labels = self._get_axis(axis) | |||
| if level is not None: | |||
| assert isinstance(labels, MultiIndex), type(labels) | |||
There was a problem hiding this comment.
could this raise a user-facing AssertionError?
There was a problem hiding this comment.
Yeah, Changed it to raise a TypeError, which is also better than an AttributeError in master.
| @@ -590,7 +590,7 @@ def swapaxes(self: FrameOrSeries, axis1, axis2, copy=True) -> FrameOrSeries: | |||
| if copy: | |||
| new_values = new_values.copy() | |||
|
|
|||
| return self._constructor(new_values, *new_axes).__finalize__( | |||
| return self._constructor(new_values, *new_axes).__finalize__( # type: ignore | |||
There was a problem hiding this comment.
hmm, can you add the mypy error as a comment.
IIUC there were issues with deprecating this method, xref #26654.
For series, it's a no-op. maybe now that Panel is deprecated, we could have logic in DataFrame and Series instead of NDFrame and simplify.
There was a problem hiding this comment.
Yeah, I remember that. The problem was that np.swapaxes(df, 0, 1), np.swapaxes(series, 0, 0) works in master and are tested for. It's not unreasonable that it works, but users should really use transpose/T instead.
Maybe add swapaxes to _deprecations to get it out of the tab-list?
topper-123
force-pushed
the
type_ndframe._get_axis
branch
from
4dff9c3
to
fadfb0a
Compare
|
Updated. |
Gives return types to
NDFrame._get_axisetc.