Closed
Description
Bug report
Bug summary
For people who use numpy.datetime64() arrays with pcolormesh, a DeprecationWarning has been appearing prompting users to specify a shading kwarg 'nearest', 'auto' or 'gouraud' since shading='flat' is being deprecated. However, passing 'nearest' results in a UFuncTypeError when using 'nearest' due to a casting issue in _axes.py from _interp_grid()
Code for reproduction
import numpy as np
import matplotlib.pyplot as plt
times = np.arange(np.datetime64('2020-12-01'), np.datetime64('2020-12-31'))
yaxis = np.arange(0, 20)
caxis = np.random.normal(0.0, 1.0, (times.shape[0], yaxis.shape[0]))
plt.pcolormesh(times, yaxis, caxis.T, shading='nearest')
plt.show()Actual outcome
Traceback (most recent call last):
File "..\test.py", line 15, in <module>
plt.pcolormesh(times, yaxis, caxis.T, shading='nearest')
File "C:\Users\iarey\anaconda3\lib\site-packages\matplotlib\pyplot.py", line 2801, in pcolormesh
__ret = gca().pcolormesh(
File "C:\Users\iarey\anaconda3\lib\site-packages\matplotlib\__init__.py", line 1438, in inner
return func(ax, *map(sanitize_sequence, args), **kwargs)
File "C:\Users\iarey\anaconda3\lib\site-packages\matplotlib\axes\_axes.py", line 6093, in pcolormesh
X, Y, C, shading = self._pcolorargs('pcolormesh', *args,
File "C:\Users\iarey\anaconda3\lib\site-packages\matplotlib\axes\_axes.py", line 5651, in _pcolorargs
X = _interp_grid(X)
File "C:\Users\iarey\anaconda3\lib\site-packages\matplotlib\axes\_axes.py", line 5633, in _interp_grid
if not (np.all(dX >= 0) or np.all(dX <= 0)):
UFuncTypeError: Cannot cast ufunc 'greater_equal' input 0 from dtype('<m8[D]') to dtype('<m8') with casting rule 'same_kind'
Expected outcome
A pcolormesh timeseries plot with random data and a np.datetime64() x-axis.
Matplotlib version
- Operating system: Windows 10
- Matplotlib version: 3.3.2 via standard conda
- Matplotlib backend: Qt5Agg
- Python version: 3.8.5
- Jupyter version (if applicable):
- Other libraries: numpy-1.19.2