Describe the bug

Currently, the axis attribute of uniform data axes can be changed (after initialization of the axis), though it should be defined by offset and scale only. Instead, it should be a protected attribute that can be changed only if offset and scale are changed, but not directly by user input. An example is that users not understanding the difference between DataAxis and UniformDataAxis, might actually change an axis to a non-uniform one just by providing the axis array without converting the axes type, which can lead to problems in a later analysis.

Would be solved by #3031

To Reproduce

Steps to reproduce the behavior:

S=hs.signals.Signal1D(np.arange(100))
S.axes_manager[0].axis = np.ones(100)

S.axes_manager and S.axes_manager[0].axis are now in contradiction to each other.

Expected behavior

Throw an error when trying to manually change the axis vector for a uniform or functional data axis.

Python environement:

• HyperSpy version: 1.7.4
• Python version: 3.11