Agenda Changes: ImageDM discussion during Education I

[Laurino, Omar]

I agree with Tim. And I believe that ModelingByUtypes is a bad idea. A data
model is not a list of strings.

However if one wants, for some reason, to fall back on the 1-to-1 mapping,
I would rather use the same strings for FITS Keywords and UTYPEs, which at
least is more portable.

Omar.


On Tue, Jun 4, 2013 at 10:21 AM, Tim Jenness <tjenness at cornell.edu> wrote:

>
> On Jun 4, 2013, at 05:15 , François Bonnarel <
> francois.bonnarel at astro.unistra.fr> wrote:
>
> >
> > If we think to provide WCS (or Mapping) for an image along the lines of
> the example in your slides what would be the utypes ?
> > I found those with their equivalent WCS FITS Keywords
> > PixelCoordSystem.PixelCoordFrame.Cart2DrefFrame.projection ---> second
> part of CTYPE
> > PixelCoordSystem.PixelCoordFrame.Cart2DrefFrame.Transform2 ----> "CDELT"
> keywords
> > PixelCoordSystem.PixelCoordFrame.CoordRefPos.Vector2DCoordinate ---->
> "CRVAL" Keywords
> > PixelCoordSystem.PixelCoordFrame.CARTESIAN.naxes ----> NAXES
> > PixelCoordSystem.PixelCoordFrame.ReferencePixel.Pixel2D.Name1 ---->
> First part of CTYPE1
> > PixelCoordSystem.PixelCoordFrame.ReferencePixel.Pixel2D.Name2 ----->
> First part of CTYPE2
> > PixelCoordSystem.PixelCoordFrame.ReferencePixel.Pixel2D.Value2  ------>
> CRPIX Keywords
> >
>
> I've always found the concept of UTYPES with WCS a bit restrictive. Ok, so
> they can handle simple FITS WCS but more complex WCS mappings quickly get
> out of control. The moves towards stackable mappings inside WCS (from STScI
> and Starlink) don't fit at all into the old simple FITS keyword WCS. Don't
> you really want the WCS UTYPE to refer to an STC-S blob?
>
> --
> Tim Jenness
> CCAT
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/dm/attachments/20130604/f67cdc79/attachment-0001.html>