Merge plotspecutils

[armengau]

Not fully ready yet for merge, but ready for review.

viewer.py is the result of a manual merge between plotframes.py and plotspecutils.py. I kept the most up-to-date code from both, with minimal changes.

Input spectra in viewer.plotspectra() can be Spectra or list[Frame] or Spectrum1D or SpectrumList. Key changes are at the beginning of plotspectra(), and in make_cds_spectra().

As of now I made tests using DESI spectra only:

• from a notebook.
• reprocessing a sample of Blanc VI dataset.
• Still TODO: more tests with DESI spectra (test all options in viewer.plotspectra()).

@weaverba137 Could you in particular:

• look a the code (diff viewer.py plotspecutils.py)
• test on SDSS (Spectrum1D and SpectrumList): I hope it works, but definitely has to be tested.
• Do you have a better idea than viewer.py for the name of the file? From the user point of view, from prospect import viewer and viewer.plotspectra() look ~ok.
• After all tests and before merging, if you agree I can remove plotframes.py and plotspecutils.py

[coveralls]

Pull Request Test Coverage Report for Build 513356796

• 0 of 787 (0.0%) changed or added relevant lines in 3 files are covered.
• 2 unchanged lines in 2 files lost coverage.
• Overall coverage decreased (-0.4%) to 1.497%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
py/prospect/plotspecutils.py	0	1	0.0%
py/prospect/plotframes.py	0	2	0.0%
py/prospect/viewer.py	0	784	0.0%
<!--	Total:	0	787	0.0%	-->

Files with Coverage Reduction	New Missed Lines	%
py/prospect/plotframes.py	1	0%
py/prospect/plotspecutils.py	1	0%
<!--	Total:	2		-->

Totals
Change from base Build 496910645:	-0.4%
Covered Lines:	59
Relevant Lines:	3942

---

💛 - Coveralls

[weaverba137]

I won't have time to review until next week, but thank you for getting this out.

[weaverba137]

With a minor typo fix, the viewer.py code is working in a Data Lab notebook. However there is still an issue with the units of the model spectrum which is preventing plotting SDSS models. That won't be hard to fix, I've seen it before.

I still want to do a line-by-line code comparison to make sure nothing subtle has been missed, but overall, I think things are working.

[weaverba137]

The problem with units on the model is fixed on the Data Lab side, it doesn't need a fix in this code.

[armengau]

About flux units: I suppose that flux units will never change in DESI. I had already noticed that, while the unit is present in the fits file: eg fits.getheader('/global/cfs/cdirs/desi/spectro/redux/daily/tiles/68000/20200314/coadd-0-68000-20200314.fits','brz_flux')['bunit'] we apparently cannot recover it from the Spectra objects. Well, that would be a request to desispec: in desispec.io.read_spectra() put flux units somewhere in meta or extra?

[weaverba137]

In this particular case, it's not a case of whether the flux array has (Astropy) units, but what scaling is applied to those units. Prospect expects 1e-17 erg cm^-2 s^-1 A^-1, but the Data Lab spectrum service returns erg cm^-2 s^-1 A^-1, so it's just a matter of multiplying or dividing by 1e-17 as needed.