Bug summary

The XAxis / YAxis objects attached to an Axes have get_transform and set_transform methods, but unlike other getters/setters, you can't get the value you set.

Code for reproduction

import matplotlib as mpl
import matplotlib.pyplot as plt
_, ax = plt.subplots()
xfm = mpl.scale.LogTransform(10)
ax.xaxis.set_transform(xfm)
print(ax.xaxis.get_transform())

Actual outcome

IdentityTransform()

Expected outcome

LogTransform(base=10, nonpositive='clip')

Additional information

I'm not sure if this is really a bug or if it's just a confusing / rough edge. The scale/transform API is fairly cumbersome, but improving that situation is a big project. Is the transform you set with axis.set_transform ever actually used? If not, then maybe set_transform should not exist (or it should raise if called)?

Matplotlib Version

3.5.1