Repository navigation
[Bug]: Type annotation for add_subplots has incorrect return type for projection="3d" #29334
Description
Bug summary
The return type for add_subplots with projection="3d" should be Axes3D.
The current implementation is:
@overload
def add_subplot(
self, nrows: int, ncols: int, index: int | tuple[int, int], **kwargs
) -> Axes: ...
@overload
def add_subplot(self, pos: int, **kwargs) -> Axes: ...
@overload
def add_subplot(self, ax: Axes, **kwargs) -> Axes: ...
@overload
def add_subplot(self, ax: SubplotSpec, **kwargs) -> Axes: ...
@overload
def add_subplot(self, **kwargs) -> Axes: ...
@overload
def subplots(Code for reproduction
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d.axes3d import Axes3D
ax: Axes3D = plt.figure().add_subplot(projection="3d")Actual outcome
Pyright complains:
error: Type "Axes" is not assignable to declared type "Axes3D"
"Axes" is not assignable to "Axes3D" (reportAssignmentType)
Expected outcome
No error.
Additional information
No response
Operating system
No response
Matplotlib Version
3.10.0
Matplotlib Backend
No response
Python version
No response
Jupyter version
No response
Installation
None
Activity
rcomer commentedon Dec 17, 2024
Duplicate of #27455 I think.
stefanv commentedon Dec 18, 2024
I'd need some feedback on #29341 from y'all who know more than I do about type annotation, but the PoC PR at least fixes the problem for me.
tacaswell commentedon Dec 24, 2024
To be slightly picky, it is a correct but less specific base type.
QuLogic commentedon Dec 24, 2024
The wider problem is that there are no type hints for any of the
mpl_toolkits, so even if we annotate anAxes3Dhere, it won't help you any...scottshambaugh commentedon Jan 24, 2025
This is a wider problem with typing in Axes3D, but this does seem like low hanging fruit
stefanv commentedon Jan 24, 2025
This did solve my specific problem, FWIW: the matplotlib examples in the SP lecture notes.