TYP: Update pyright #56892
TYP: Update pyright #56892
Conversation
| @@ -69,16 +69,26 @@ def fast_multiget( | |||
| mapping: dict, | |||
| keys: np.ndarray, # object[:] | |||
| default=..., | |||
| ) -> np.ndarray: ... | |||
| ) -> ArrayLike: ... | |||
There was a problem hiding this comment.
maybe_convert_objects returns an ExtensionArray or an np.ndarray. Most functions call maybe_convert_objects only when convert=True and otherwise return an np.ndarray (see the overloads).
| def unstack(self, level: IndexLabel = -1, fill_value=None, sort: bool = True): | ||
| def unstack( | ||
| self, level: IndexLabel = -1, fill_value=None, sort: bool = True | ||
| ) -> DataFrame | Series: |
There was a problem hiding this comment.
returns a Series when a DataFrame has a non-MultiIndex index (this is also documented; a series with a non-MultiIndex raises instead). Pyright's magic was able to infer that DataFrame | Series can be returned, unfortunately, this requires quite a few ignore codes.
pandas/core/indexes/range.py
Outdated
| @@ -500,13 +500,13 @@ def _minmax(self, meth: str): | |||
|
|
|||
| return self.start + self.step * no_steps | |||
|
|
|||
| def min(self, axis=None, skipna: bool = True, *args, **kwargs) -> int: | |||
| def min(self, axis=None, skipna: bool = True, *args, **kwargs) -> float: | |||
There was a problem hiding this comment.
This is a bit sad: _minmax can return np.nan which is a float
There was a problem hiding this comment.
Could we still annotate int | float here (and for _minmax). I think we had previously discussed wanting to be a bit more explicit with float capturing int
| @@ -456,7 +456,7 @@ def _str_lstrip(self, to_strip=None): | |||
| def _str_rstrip(self, to_strip=None): | |||
| return self._str_map(lambda x: x.rstrip(to_strip)) | |||
|
|
|||
| def _str_removeprefix(self, prefix: str) -> Series: | |||
| def _str_removeprefix(self, prefix: str): | |||
There was a problem hiding this comment.
this was simply wrong
| @@ -112,7 +114,8 @@ def _plot( # type: ignore[override] | |||
| *, | |||
| bins, | |||
| **kwds, | |||
| ): | |||
| # might return a subset from the possible return types of Axes.hist(...)[2]? | |||
| ) -> BarContainer | Polygon | list[BarContainer | Polygon]: | |||
There was a problem hiding this comment.
Matplotlib has type annotations and lists the above types as possible return types. I think it only returns BarContainer in test tests. As Polygon could be returned, I opted to reflect that in the signature as it could uncover runtime issues (but I'm not familiar enough to make further decisions)
| @@ -637,7 +638,7 @@ def _str_map( | |||
| # error: Argument 1 to "dtype" has incompatible type | |||
| # "Union[ExtensionDtype, str, dtype[Any], Type[object]]"; expected | |||
| # "Type[object]" | |||
| dtype=np.dtype(dtype), # type: ignore[arg-type] | |||
| dtype=np.dtype(cast(type, dtype)), | |||
There was a problem hiding this comment.
map_infer_mask requires that dtype is np.dtype but np.dtype is a generic type: since the call to it seems invalid to mypy, it cannot infer the correct generic type of np.dtype which creates the existing type error. This PR uses overloads for map_infer_mask (convert=False returns np.ndarray): since the call to it seems invalid, mypy picks the last overload, which is not correct here - so I cast the argument for np.dtype to avoid this mess.
|
Thanks @twoertwein |
* TYP: Update pyright * isort * int | float
Newer releases of pyright infer return types of partially annotated functions, which uncovers quite a few difficult-to-fix type errors (some functions had no return types because it would create too many typing errors elsewhere).
(pyright can also try to infer return types for completely unannotated functions, but that is disabled for now in pyright_reportGeneralTypeIssues.json as it would trigger more difficult-to-fix cases.)