> Personally, I'd rather do it via a data model. Ie. the relevant
> columns have a utype (or whatever it ends up being called) which
> effectively says `I am the wavelength', `I am the intensity, or flux
> or whatever' and `I am the error on the intensity or flux' etc. etc.
So that's what a utype is. I've been asking for this in FITS for more
than a decade and it never raised a response apart from Don Wells
privately. We're obviously ahead of our time. People didn't have these
multi-component structured data unlike ourselves until comparatively
recently, hence didn't appreciate the problem.
Malcolm
|