For validation of hsoy, this lets you produce various plots doing global
comparisons between hsoy and ppmxl (and potentially gaia).

Targets:
	
	<column-name>.systematic.png -- differences between the healpix-averages
	of column-name in hsoy and ppmxl.

	<column-name>.average.png -- averages by healpix for hsoy columns

	plus several more, literally given in the Makefile



Missing Objects
===============

It turned out that almost all high-proper-motion objects are missing in Hsoy.
Among the ones found missing were ipixes

266846627338158592, 491806603301896448, 466294291405221504, 540136495894234496,
1094235724309932160, 2379962599871451264, 4857520566244401664

It turns out that all of them are already missing in cleandppmxl.  The table
hsoy_missing_ppmxl_vs_gaia.vot contains a hand-selected set of
high-proper-motion objects where ppmxl proper motions match the gaia
observations just fine but that are still not in hsoy.

Crossmatching this with supercosmos.sources does indeed not yield
many matches.  ipix 3397063242071739738, however, has a nice match
in supercosmos but is missing in withgaia.  It's in cleanedppmxl, though.
Hm.  Perhaps that's a fluke.