[Bug]: timedelta64 with pcolormesh #26525
Description
Bug summary
timedelta64 doesn't work with pcolormesh, and many other methods.
Code for reproduction
import matplotlib.pyplot as plt
import numpy as np
time = np.arange(0, 50, dtype='timedelta64[h]')
y = np.arange(10)
Z = np.random.randn(10, 50)
fig, ax = plt.subplots()
ax.pcolormesh(time, y, Z)Actual outcome
Traceback (most recent call last):
File "/Users/jklymak/matplotlib/testit.py", line 9, in <module>
ax.pcolormesh(time, y, Z)
File "/Users/jklymak/matplotlib/lib/matplotlib/__init__.py", line 1465, in inner
return func(ax, *map(sanitize_sequence, args), **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/jklymak/matplotlib/lib/matplotlib/axes/_axes.py", line 6270, in pcolormesh
X, Y, C, shading = self._pcolorargs('pcolormesh', *args,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/jklymak/matplotlib/lib/matplotlib/axes/_axes.py", line 5854, in _pcolorargs
X = _interp_grid(X)
^^^^^^^^^^^^^^^
File "/Users/jklymak/matplotlib/lib/matplotlib/axes/_axes.py", line 5833, in _interp_grid
dX = np.diff(X, axis=1)/2.
~~~~~~~~~~~~~~~~~~^~~
File "/Users/jklymak/mambaforge/envs/mpl-dev/lib/python3.11/site-packages/numpy/ma/core.py", line 4254, in __truediv__
return true_divide(self, other)
^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/jklymak/mambaforge/envs/mpl-dev/lib/python3.11/site-packages/numpy/ma/core.py", line 1171, in __call__
m |= domain(da, db)
^^^^^^^^^^^^^^
File "/Users/jklymak/mambaforge/envs/mpl-dev/lib/python3.11/site-packages/numpy/ma/core.py", line 858, in __call__
return umath.absolute(a) * self.tolerance >= umath.absolute(b)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: The DType <class 'numpy.dtype[timedelta64]'> could not be promoted by <class 'numpy.dtype[float64]'>. This means that no common DType exists for the given inputs. For example they cannot be stored in a single array unless the dtype is `object`. The full list of DTypes is: (<class 'numpy.dtype[timedelta64]'>, <class 'numpy.dtype[float64]'>)
Expected outcome
The expected outcome is debatable. We have tried to have this go through units like datetime64. However, I really think we should just convert to float. Yeah, that will give giant numbers for timedelta64 of ns, but I think it's better than failing.
Additional information
No response
Operating system
No response
Matplotlib Version
main
Matplotlib Backend
No response
Python version
No response
Jupyter version
No response
Installation
None