DM-4956: Adapt SRD-based measurements of astrometric performance for validate_drp

[wmwv]

Addition of astrometric validation according to SRD specifications.
Refactoring to support this use.
Added documentation.

[TallJimbo]

...that we're looking at here...

[TallJimbo]

Only general comment is that I'm concerned by the amount of explicit angle-wrangling and coordinate manipulation that's going on in this code.

We introduced the Angle class here to deal with the former, but it looks like you can't really use it much since it's not NumPy-friendly. Hopefully someday AstroPy integration will give us a better option; should we maybe create a deferred ticket to deal with that when we can (maybe that's a question for @frossie - I don't know if she considers long-deferred tickets to be helpful or harmful).

As for coordinate manipulation, it'd be very nice if you could take stock of all the low-level operations on coordinates you need here and create an issue for putting that kind of code in to afw or sphgeom. Of course, once again AstroPy integration may ultimately solve that problem for us.

[wmwv]

I find it personally helpful to create long-delayed tickets (with no due dates) to represent thoughts of things that are being pushed off as out of scope.

[wmwv]

I will think about the coordinate manipulations here and think about requests for base functions to implement these. 'averageCoord' was certainly helpful and one step in this direction. However, it's becoming clear to me that I think it should have been a spherical geometry routine, not a coordinate routine.

My current overall comment is that there a number of spherical geometry calculations that are instead in lsst.afw.coord. I shouldn't have to even known about the existence of ICRS to take the difference between two positions on the sphere.

    angleRa = [afwGeom.Angle(r, afwGeom.radians) for r in ra]
    angleDec = [afwGeom.Angle(d, afwGeom.radians) for d in dec]
    coords = [afwCoord.IcrsCoord(ar, ad) for (ar, ad) in zip(angleRa, angleDec)]

    meanRa, meanDec = afwCoord.averageCoord(coords)