Closed moustakas closed 1 year ago
actually, the credits go to Christophe M. from saclay.
he noticed that in desispec.emlinefit.emlines_gaussfit()
, which I ll have to also modify, btw.
Here's a quick comparison with the Portsmouth SDSS catalog-- https://data.sdss.org/datamodel/files/BOSS_GALAXY_REDUX/GALAXY_VERSION/portsmouth_emlinekin_full.html
This definitely shows that I messed up the factor of (1+z). In particular, all measurements are done in the observed frame and there's no factor of 1+z in the integrated flux, but the continuum flux (in the denominator) needs to be multiplied by (1+z) to move it to the rest frame. So my factor
here should be
factor = 1 / ((1 + redshift) * result['{}_CONT'.format(linename)]) # --> rest frame
Edit: my plot says "delta" but it's a ratio of EWs.
@araichoor pointed out a terribly embarrassing bug in my EWs: the factor of 1+z should be in the denominator, not the numerator-- https://github.com/desihub/fastspecfit/blob/main/py/fastspecfit/emlines.py#L1554
I thought I had checked these EWs against SDSS, pPXF, and other codes, but somehow this wrong factor slipped through. I'll do some additional checks before making this change, but this will definitely be fixed in the next version of FastSpecFit.