> but although kappa gives no errors, nothing seems to happen. If I look at the
> header in gaia, the naxis etc keywords are still there.
The NAXIS FITS headers relate to the original FITS file and not to the
structure of the NDF. They are part of the so-called FITS Airlock.
Components of the NDF have general but defined meanings, whereas most
FITS keywords are specialised without fixed meaning in the FITS
standard. (You would need a data dictionary from the host institution
to decide whether a keyword maps to a NDF component). The Airlock
connects the defined semantics of the NDF with the unknown of the
outside world.
If you export the NDF to FITS, FITS2NDF will replace certain headers,
including the NAXIS keywords, with values corresponding to the NDF's
properties and not use the original values. You can edit the FITS
headers too. Examples and more about the so-called FITS Airlock are in
SUN/95.
% findme "FITS Airlock" sun95
You need to run KAPPA's NDFTRACE as Tim suggested to view the properties
of the NDF including its dimensionality and shape.
You don't have to know the name of the AXIS component if you run
KAPPA:SETAXIS.
% setaxis myndf mode=delete
> The only thing I can think of is that I had loaded up figaro and kappa in
> the opposite order. i.e. figaro first.
The order shouldn't matter as we tried to avoid name clashes, and
consequently Keith Shortridge and I renamed a few Figaro and KAPPA
commands respectively. Unfortunately, the Starlink Panel later
requested a merger of some packages into Figaro, and the namespace
aspect was overlooked. These few clashes (CADD, CSUB spring to mind)
aren't related to Sarah's problem.
Malcolm
|