Following discussion with @emhuff , it would be good to allow flexibility in the type of observables that are modeled by the fitting code. This is currently handled in fit.vmapFit where arguments vobs and imObs can be optionally set to observed values or to None to be ignored (i.e. to fit only the spectroscopic data, or only the imaging data).
We want to expand the data vector to a full spectral datacube, but preserve the ability to use the "collapsed" data vector of a few of velocity measurements and imaging observables.
Implementing this may be easiest with a Data class akin to the Model class, which holds the observables and a description of what data to use in the fit.
Following discussion with @emhuff , it would be good to allow flexibility in the type of observables that are modeled by the fitting code. This is currently handled in fit.vmapFit where arguments vobs and imObs can be optionally set to observed values or to None to be ignored (i.e. to fit only the spectroscopic data, or only the imaging data).
We want to expand the data vector to a full spectral datacube, but preserve the ability to use the "collapsed" data vector of a few of velocity measurements and imaging observables.
Implementing this may be easiest with a Data class akin to the Model class, which holds the observables and a description of what data to use in the fit.