Diffmap enhancements

[kif]

Start using WorkerConfig
Update the way polarization works
close #2396
close #2390
close #2383
close #2382

[EdgarGF93]

Continuing with the conversation from the other PR, I am thinking that the guess_binning should be done in the worker (maybe in the process method: like self.ai.detector.guess_binning(data=data)`) . Because at some point we will face the same issue, the data is binned and this information is missing. But then, the automatic mask for Dectris has to change also (this is not implemented yet)

[kif]

To me it is a nonsense to bin pixel-detector data ... so for Dectris I consider it makes little sense.
I processed some files from Dectris and the mask is apparently properly binned.

[loichuder]

If we can check for the size, what about going through all axes elements instead of these specific indices ?

[kif]

I intent to make the default view (in the Nexus sense) on the map array, not on the intensity array. This is to try to ensure the different axes have the proper shape while keeping the compatibility with former files.

[kif]

Let's check how pilx reacts with files produced with former version of pyFAI ...

[kif]

close #2396

[kif]

TODO: check pilx for the opening of diffmap files from v 2024.X

[EdgarGF93]

pyfai-diffmap-view from the current branch can handle diffmap files generated with config files v3, v4 and v5.
I tried with these files:

dmap_pyFAI202405_configv3_gui.h5    dmap_pyFAI202409_configv4_gui.h5    dmap_pyFAIdev_configv5_gui.h5
dmap_pyFAI202405_configv3_nogui.h5  dmap_pyFAI202409_configv4_nogui.h5  dmap_pyFAIdev_configv5_nogui.h5

pyFAI 2024.05 uses a configuration file version 3
pyFAI2024.09 uses a configuration file version 4
pyFAIdev (this branch) uses a configuration file version 5

[kif]

Yes, but this branch provides the motor position handling in addition.

[kif]

look at the map, its axes are in motor position, which eases the identification of the sample position.

[EdgarGF93]

But 2024.x version of pyFAI cannot open the new diffmap files

pyFAI-diffmap-view pyFAI.app.pilx:main
########### IMPORTLIB ###########
Traceback (most recent call last):
  File "/gpfs/jazzy/data/scisoft/edgar/work/pyFAI/bootstrap.py", line 223, in <module>
    run_entry_point(res[1], res[2], argv)
  File "/gpfs/jazzy/data/scisoft/edgar/work/pyFAI/bootstrap.py", line 163, in run_entry_point
    func()
  File "/gpfs/jazzy/data/scisoft/edgar/work/pyFAI/build/lib/python3.12/site-packages/pyFAI/app/pilx.py", line 59, in main
    window.initData(file_name=options.filename,
  File "/gpfs/jazzy/data/scisoft/edgar/work/pyFAI/build/lib/python3.12/site-packages/pyFAI/gui/pilx/MainWindow.py", line 152, in initData
    self.displayPatternAtIndices(initial_indices, legend="INTEGRATE")
  File "/gpfs/jazzy/data/scisoft/edgar/work/pyFAI/build/lib/python3.12/site-packages/pyFAI/gui/pilx/MainWindow.py", line 180, in displayPatternAtIndices
    self._integrated_plot_widget.addCurve(
  File "/gpfs/jazzy/data/scisoft/edgar/work/silx/src/silx/gui/plot/PlotWidget.py", line 1258, in addCurve
    curve.setData(x, y, xerror, yerror, baseline=baseline, copy=copy)
  File "/gpfs/jazzy/data/scisoft/edgar/work/silx/src/silx/gui/plot/items/curve.py", line 347, in setData
    PointsBase.setData(self, x=x, y=y, xerror=xerror, yerror=yerror, copy=copy)
  File "/gpfs/jazzy/data/scisoft/edgar/work/silx/src/silx/gui/plot/items/core.py", line 1756, in setData
    assert len(x) == len(y)
AssertionError

[kif]

This is an expected bug ... It would be extremely complex to make code "forward-compatible" without a lot of planning and design ...