Line Fitting

[eteq]

From @keflavich: "should use the astropy.modeling machinery (including choices for the user of model and fitter) but wrap it in a useful/easy way"

[eteq]

To expand more: this is a very broad topic - this "meta issue" is to try to wrap a lot of idas together in one.

Key guideline: there is not "one true way" to do line fitting. Rather specutils should provide a toolbox of pieces that make it straightforward to do standard things, but also doesn't get in the way of more advanced/experimental methods.

A general guide (in progress at the top of this writing) for how this might work is here: https://github.com/spacetelescope/dat_pyinthesky/blob/master/pyinthesky_specutils_fitting.ipynb
with a start at implementation in #210.

None of this is set in stone, and all is open for discussion here... But we are trying to get some initial implementation done in a unified way consistent with Spectrum1D so that there's something to build from.

[SaOgaz]

related or duplicate of #195

[PatrickOgle]

Basic line-fitting functionality is available in specutils. Leave open for further discussion whether this requires a wrapper or dedicated notebook to explore line-fitting methods.

[nmearl]

We currently have documentation that covers the use case of arbitrary models and line fitting. Closing this as a new issue should be open in the case where work needs to be done on adding a dedicated notebook.