[fitswcs] Status of WCS negotiations

[Doug Mink]

In response to Don Wells:
> Four questions seem to me to block the path toward agreement:
> 
> (1) SPLITTING-G&C:

I think that we have a ways to go before having the FITS conventions
standardized, but that those FITS issues related to the actual
projections
are pretty well settled, so I agree with Don's splitting proposal.

> (2) MULTIPLE-WCS: It appears to me that there is a requirement to
> support multiple WCS descriptions of the same data object. I recommend
> that we do this by appending a character code [A-Z] to the WCS
> keywords for secondary descriptions; the keywords without the
> character would be the default operational ones.

In thinking about implementing this proposal, I wonder whether
prepending
the character rather than appending it might be easier, though I don't
have a strong feeling against appending.

> (3) CD-MATRIX:
This would be defined in the first paper, I guess, which would include
information from the Hanisch and Wells 1988 memorandum, which does
not correctly define how to get CROTA2 from the CD matrix in all cases.
Since I have yet to see any images with anything but CROTA2 or CDn_n,
this makes sense to me.  My WCSTools software does support the PC
matrix,
since I use WCSLIB for the actual projections and I wanted to be able
to pass it through if it were present.

> (4) PIXEL-CORRECTION-FUNCTION: 
I have been substituting a polynomial for the CD matrix, which is after
all, a first-order polynomial with cross terms with CRVALn as the
constant.  I fit a CD matrix and CRVALn and use the CD matrix as initial
values for a polynomial with more terms, in what is basically a simple
plate solution.  Second order fits most CCD images I have seen to the
accuracy of the reference catalogs, though sometimes adding a
third-order
improves the fit.  This seems to be the simplest thing to do, although
I am open to suggestions which will improve the accuracy of the method.

My code is available in the platepos.c subroutine in

http://tdc-www.harvard.edu/software/wcstools/wcssubs-2.3.tar.gz

with the full package in

http://tdc-www.harvard.edu/software/wcstools/wcstools-2.3.tar.gz

(our ftp server is down)

-Doug Mink