Implement astropy unit conversion to rescale_path_length()

[TranHuuNhatHuy]

Description

This pull request is to address #444. Now we can use astropy units for the input arguments of rescale_path_length().

Beside, currently on the RADIS documentation, rescale_path_length() has 2 references:

1. radis.Spectrum.rescale_path_length
2. radis.Spectrum.rescale.rescale_path_length

For [1]:

• @erwanp 's syntax is added in the Examples section.
• I also notice that the gallery example for [1] does not show up on documentation. Currently the referring syntax is .. minigallery:: radis.Spectrum.rescale_path_length.
• Link for the source is currently not working. You can see it here.

For [2], it's just kinda empty here.

[erwanp]

Nice ! I think it should be radis.spectrum.Spectrum.rescale_path_length

there is a script in doc/gendocs.sh that allows you to generate docs locally; helps the debugging. You may need to install the doc specific requirements

[erwanp]

What we'll need before merging is to add tests.

There is a test_rescale.py file. You can add a new function there, testing the units.

You can use a precomputed spectrum such as

s = load_spec(getTestFile("CO_Tgas1500K_mole_fraction0.01.spec"), binary=True)

and you could, for instance, compare the absorbance (in absorption spectroscopy, the absorbance is A = k * L where k is the absorption coefficient, L the length ; so absorbance should have changed when you rescaled the spectra ) of a a spectrum rescaled with astropy units (e.g 1 * u.m ) and with default units (100 # cm )

[codecov-commenter]

Codecov Report

Merging #445 (77f71b2) into develop (778bdbf) will increase coverage by 0.02%.
The diff coverage is 91.66%.

@@             Coverage Diff             @@
##           develop     #445      +/-   ##
===========================================
+ Coverage    77.06%   77.09%   +0.02%     
===========================================
  Files          161      161              
  Lines        18641    18664      +23     
===========================================
+ Hits         14366    14389      +23     
  Misses        4275     4275              

[TranHuuNhatHuy]

@erwanp thank you for the reviews. Last commit features two changes according to your suggestions:

1. Fetch absorbance directly by using s.get("absorbance", wunit='cm-1'). I believe this is better than fetching k and L then calculate the absorbance A = k * L, in terms of computational performance and the sake of clean code.
2. Add third assertion of comparing the two path_length of 100000 cm and 1 u.km to make sure the astropy unit works fine. Although I doubt it is necessary if the two absorbances are already equal, I think with this we can check if something is wrong with the astropy units in the future (extremely unlikely thou).

[erwanp]

Excellent, merging!

[anandxkumar]

Fixes #444