[fitsbits] Question about CROTA

[Mark Calabretta]

On Tue 2005/11/08 10:41:27 CDT, William Thompson wrote
in a message to: Mark Calabretta <mcalabre at atnf.csiro.au>,
      FITSBITS <fitsbits at donar.cv.nrao.edu>

>The question is, what happens in AIPS and other software packages, if the 
CROTA
>value in the latitude axis is replicated in the longitude axis?  Is it simply 
>ignored, or does it lead to unintended consequences?

In AIPS I expect it would simply be ignored, Eric can give an
authoritative answer.

In WCSLIB the header parser (wcspih) loads CROTAn from all axes into an
array, but the interpreter (wcsset) only considers the one on the
latitude axis as being valid for constructing the transformation matrix
(if PCi_j and CDi_j are not present in the header).  It ignores all the
others without issuing an error or warning.

I've no idea what aips++, IRAF, MIDAS, GIPSY, or anything else does.

Mark Calabretta
ATNF