dax spectral models?

[kglotfelty]

I noticed in the dax regression test the blurb about xswabs being deprecated. We haven't update the list of spectral models available with dax in a while. Are there any we should consider adding/removing?

[DougBurke]

Well, xswabs is I think going to stay around, it is just not something that it's worth using as it's outdated.

I can't think of any others in terms of deprecation.

What is the current list of models that dax knows about?

[kglotfelty]

For spectral models:

model menu {Model: mdl1} powlaw1d|bbody|bpl1d|gauss1d|normgauss1d|lorentz1d|logparabola|xszpowerlw|xszbbody|xsbknpower|xscutoffpl|xszgauss|xsdiskline|xsapec|xsbapec|xsmekal|xsvmekal|xsraymond|xszbremss|xscemekl {Spectral Model - model editor will open}
add_model menu {Additional component: mdl2} none|powlaw1d|bbody|bpl1d|gauss1d|normgauss1d|lorentz1d|logparabola|xszpowerlw|xszbbody|xsbknpower|xscutoffpl|xszgauss|xsdiskline|xsapec|xsbapec|xsmekal|xsvmekal|xsraymond|xszbremss|xscemekl {Added to mdl1 - model editor will open}

For absorptions models:

specfit_absmodel menu {Absorption model: abs1} xsphabs|xswabs|xstbabs|xszwabs|xszphabs|xsztbabs {Absorption model - model editor will open}
specfit_absxtra menu {2nd Absorption model: abs2} none|xsphabs|xswabs|xstbabs|xszwabs|xszphabs|xsztbabs {Absorption model - model editor will open}

Just for completeness

For spatial models

img_mdl1 menu {Model 1} gauss2d|beta2d|devaucouleurs2d|lorentz2d|const2d|polynom2d|sigmagauss2d|sersic2d|scale2d|normgauss2d {First model component}
img_mdl2 menu {Model 2} const2d|none|polynom2d|gauss2d|beta2d|devaucouleurs2d|lorentz2d|sigmagauss2d|sersic2d|scale2d|normgauss2d {Second model component}

and for radial profiles

rprof_srcA menu {Model1} gauss1d|beta1d|const1d|lorentz1d|polynom1d|powlaw1d {First model component}
rprof_srcB menu {Model2} none|gauss1d|beta1d|const1d|lorentz1d|polynom1d|powlaw1d {2nd additive model component}

[anetasie]

I agree with Doug. We can keep the xswabs for completeness. The default dax setting shows xsphabs which is fine.

The list of models looks good. I'm wondering about the statistics setting for 2d image fitting - preferentially for X-ray images in counts we should have a default stat set to cstat.

[kglotfelty]

We don't know if user is fitting X-ray images or other images (radio, optical, etc). I presume the same reason why sherpa's default statistic is also chi2gehrels.