Plotting updates

[shanosborne]

These plotting updates come from problems found while investigating issues #1 and #71.

A summary of the changes made:

• Made NIRSpec MSA detector channel plotting possible by plotting the channels from the NRS_FULL apertures
• Removed NotImplementedError for siaf.py plot_detector_channels()
• Add ax argument to aperture.py and siaf.py plot_detector_channels()
• Renamed plot_all_apertures() showchannels argument to detector_channels in order to match plot_main_apertures()
• Updated labeling arguments to be more logical between the siaf.py and aperture.py plot() methods
• Updated siaf.py plot() docstring to have frame examples in lowercase since that's what the argument expects.

[shanosborne]

I also found an issue with plot_all_apertures(), where it tries to apply the **kwargs from the function definition in 2 places: aps = Siaf(instr, **kwargs) and aps.plot(clear=False, subarrays=subarrays, **kwargs).

I can either change it so there’s only 1 function that gets the **kwargs, or I can sort the kwargs based on the expected arguments of the 2 functions which would look something like:

siaf_kwargs_list = inspect.signature(Siaf).parameters.keys()
siaf_kwargs = dict((key, value) for key, value in kwargs.items() if key in siaf_kwargs_list)

However, this second method does assume that there is no overlap in the argument names for the 2 functions (which appears true for now, may not be in the future). Which method would you prefer to use to fix this issue? Do you think the kwargs will more likely be used for one of the functions and thus we can just remove the kwargs call from the other one?

[Johannes-Sahlmann]

Thanks for spotting this. I think **kwargs should be passed to aps.plot only (definitely not to two different methods).

[Johannes-Sahlmann]

PR looks good to me, I just would like the tests to pass before approving.

[shanosborne]

It looks like this warning doesn't have anything to do with my changes since it's also present in my master branch. But I'll look into what's causing it.

[Johannes-Sahlmann]

Thanks for fixing the pytest warning more elegantly than I would have done.
I am not sure I actually understand how that fixture works now, but I'll look into this at a later time.