Fixes for binned non_uniform_axes

[jlaehne]

Work on key issues from #2398

Description of the change

• Fix plotting for binned spectrum images with non uniform axis obtained by convert_to_non_uniform_axis (index_in_array was lost)
• Fix curve fitting for binned spectrum images with non-uniform axis
• Handle occurences where 'scale' is used for binned signals also in the non-uniform case (e.g. in signals used for background removal)
• Stacking along non-uniform axes
• Raise warnings where still necessary, e.g. for EELS and EDX; add to docstrings
• Test for warnings
• Increase coverage of axes.py and other files affected by the non-uniform axis branch

Progress of the PR

• fix plotting,
• fix curve fitting,
• handle 'scale' for non-uniform, binned signals,
• stacking for non-uniform axes,
• warnings for EELS and EDS models,
• check for additional functions that need to raise warnings,
• increase coverage,
• add tests,
• ready for review.

[codecov]

Codecov Report

Merging #2728 (967fe74) into non_uniform_axes (3ebb886) will increase coverage by 0.31%.
The diff coverage is 93.84%.

@@                 Coverage Diff                  @@
##           non_uniform_axes    #2728      +/-   ##
====================================================
+ Coverage             77.50%   77.81%   +0.31%     
====================================================
  Files                   203      203              
  Lines                 31066    31103      +37     
  Branches               6782     6791       +9     
====================================================
+ Hits                  24077    24204     +127     
+ Misses                 5159     5096      -63     
+ Partials               1830     1803      -27     

Impacted Files	Coverage Δ
hyperspy/_signals/eds_sem.py	61.38% <ø> (+1.98%)	⬆️
hyperspy/_signals/hologram_image.py	88.61% <ø> (+3.96%)	⬆️
hyperspy/_signals/signal2d.py	80.47% <ø> (+1.20%)	⬆️
hyperspy/signal_tools.py	67.62% <0.00%> (-0.11%)	⬇️
hyperspy/_signals/eels.py	76.75% <66.66%> (+1.64%)	⬆️
hyperspy/models/model1d.py	87.96% <75.00%> (-0.34%)	⬇️
hyperspy/misc/utils.py	85.54% <90.90%> (+0.02%)	⬆️
hyperspy/_components/doniach.py	94.73% <100.00%> (+0.14%)	⬆️
hyperspy/_components/exponential.py	82.60% <100.00%> (+2.89%)	⬆️
hyperspy/_components/gaussian.py	97.33% <100.00%> (+2.59%)	⬆️
... and 26 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3ebb886...967fe74. Read the comment docs.

[jlaehne]

The failing test in test_mrcz.py is unrelated (fails also in other branches).

For the remaining 3 Codecov warnings I am not sure which test to amend to catch these lines.

[thomasaarholt]

Aside from a few small comments, LGTM! Excellent choice of axis name and units in the tests - "lovely plumage" :)

[jlaehne]

Aside from a few small comments, LGTM! Excellent choice of axis name and units in the tests - "lovely plumage" :)

Yes, we definitely need more Monty (in HyperSpy's) Python :)

[ericpre]

A first iteration of comments, I may have some more later.

[ericpre]

Looks good to me. There is one docstring left to update and two comments I will leave it to you, if you want to take them into account or not.

[jlaehne]

I took into account all the comments, thanks for all the feedback @ericpre and @thomasaarholt.

[thomasaarholt]

Really nice @jlaehne!