Add support for specifying multiple isotopogues

[suzil]

This is the isotope parameter of calc_spectrum: https://radis.readthedocs.io/en/latest/source/radis.lbl.calc.html#radis.lbl.calc.calc_spectrum

[minouHub]

Hello,

This is a major issue that prevents the app to be used as much as spectraplot.com Typically, we use this app to quickly check where are the lines we want to probe. I was surprised first, but isotopes lines present some important features in the spectra. For instance, CO around 2008 cm-1. For simplicity of application, would it just be easier to include all isotopologue during each calculation?

[suzil]

Sorry about that. It's an easy fix to just use all: https://github.com/suzil/radis-app/commit/aad6935e509b24e09e52a9ec612736949f9b8da8 Hoping that it works but I'm unable to test it right now because the HITRAN API currently isn't letting me connect. Will try testing it later.

It doesn't seem like spectraplot.com supports specifying particular isotopes right?

[minouHub]

It doesn't seem like spectraplot.com supports specifying particular isotopes right?

Yes you're right. And it makes sense. As far as I know, people simulate molecular spectra as they occur "in nature", which means including the relative abundance of isotopologue . Those who work with special isotopologue are probably advanced enough to use radis (or other) on python.

For the HITRAN API, it seems to me that the website is just down (11.30 pm, 26th)

[suzil]

I'm having some issues using this API: Please define isotope explicitely (cannot use 'all' with fetch_databank('hitran')

I've tried isotope="all" and isotope={"CO": "all"}. I'm going to revert the last commit to fix the app.

[minouHub]

So it seems that this option prevented for HITRAN, but not for HITEMP. Is there a reason @erwanp ?

[erwanp]

The option to have all isotopes for HITRAN been request in RADIS (https://github.com/radis/radis/issues/254) but it's not implemented yet The reason it's not obvious is that in HITRAN you have to query each isotope individually (but you don't know how many there are ... CO2 has 12 for instance) ; in HITEMP you just download one file for a given molecule, with all lines, including all available isotopes. To implement it would not be very hard, because we have the list of all isotopes in molparams, read in MolParams , but no-one did! )

For RADIS-App :

• I think having all isotopes will be a lot of querying ; and waste a lot of time (until we have all the HITRAN database pre-downloaded on disk )

I'd use :

• by default, isotopes 1 to 3. For most use-cases it's more than enough (see the isotopic abundances here , and download will already be 3x longer!). So isotope='1,2,3'

If, for some molecule, isotopes don't exist, the backend will send a KeyError. See by yourselves :

radis.test_spectrum(molecule='CH3Cl', isotope='1,2,3')
>>> KeyError: '(24, 3) <<w this error occured in Astroquery. Maybe these molecule (CH3Cl) and isotope (3) are not supported'

In which case we could :

• default to isotope="1"
• (nice-to-have) send a Warning to the user (the same way the Errors appear as pop-up ? )