More flexible specification of sky locations?

[kyleaoman]

Currently the source and data cube accept (or set internally) an RA and a Dec. It would be possible to leverage astropy's coordinate module to offer more flexibility:

• Would be good to be able to define the source directly in Galactocentric coordinates, such as from Sibelius (x, y, z) coordinates with the Galaxy at the origin. In this case initializing with RA, Dec and distance doesn't make sense, these become derived quantities, so need to re-think the __init__ or offer a classmethod.
• Instead of RA and Dec one might want to specify l and b in Galactic coordinates, or generally other systems. The __init__ arguments could be restructured to take a SkyCoord?

Vaguely related and may as well be done at the same time:

• Systemic velocity could have an option to be a vector.