search

usnistgov / MATS

Multi-spectrum Analysis Tool for Spectroscopy
https://pages.nist.gov/MATS/
Other
9 stars 7 forks source link

LM implementation #55

Closed cosselk closed 2 years ago

cosselk commented 2 years ago

I think that LM is not being calculated even if included in the linelist file. I think lines 187-188 of Fit_Dataset.py should be: lineshape_vals_real, lineshape_vals_imag = pcqsdhc(line['nu'],line['GammaD'],line['Gamma0'],line['Gamma2'],line['Shift0'],line['Shift2'], line['NuVC'],line['Eta'],wavenumbers[BoundIndexLower:BoundIndexUpper], line['Y'])

Similarly in HTP_wBeta_from_DF_select()

eadkins333 commented 2 years ago

I think we are looking at different version numbers, however in the current master version, that section of code is at line 197.

The LM implementation is included a few lines lower in lines 199 - 201.

Xsect[BoundIndexLower:BoundIndexUpper] += mol_dens * \ molefraction[line['molec_id']] * line['abun_ratio'] * \ line['LineIntensity'] * (lineshape_vals_real + line['Y']*lineshape_vals_imag)

You are correct that the an alternative approach would be to include the LM directly from the psqsdhc definition.

If you are having issues with including LM in your fits, it could be that in the Generate_FitParam_File instantiation the linemixing parameter is set to False.

cosselk commented 2 years ago

Weird. I had tried calculating a spectrum with LM and then with the LM parameters in the linelist set to 0 and saw no difference in the spectrum, which is why it seemed like something was missing. I'm not sure what changed, but it's working now, and I see a difference in the spectra. Thanks!