[coords/trans] - Relation of Frame to CoordSpace

CresitelloDittmar, Mark mdittmar at cfa.harvard.edu
Tue Oct 1 20:11:20 CEST 2019


All,

I'm reporting this on behalf of David Berry, who is working the Transform
model implementation thread (see twiki use case #2
<https://wiki.ivoa.net/twiki/bin/view/IVOA/STC2#Uses_cases> ).
The description there involves passing transform model Mapping instances
between gWCS and AST, and so didn't have any lien on the Coordinates model.

In the process, he has expanded the scope to a more useful thread for the
AST library, which is to:
  "import/export the full WCS associated with an image"
This includes descriptions of the Pixel axes, image axes, and the
associated transforms, which brings the Coords model into the scope.
The thread wants to specifically NOT include the coordinates themselves.
NOTE: this is similar to a usage discussed in the context of SIA2 for image
cutouts, so may be more broadly applicable.

In working this, he is having some trouble due to a change made some
iterations ago, back in 2017, which separated the Frame and CoordSpace
elements.  From that point, they have been linked only via the Coordinate
value itself, which references both a Frame and an axis of the CoordSpace.

To satisfy this extended usage case, would add a requirement to the Coords
model:
   "enable the description of a complete WCS ( axes, frames, mappings
between ) independent of the coordinates in that space"

This seems like a reasonable extension of the usage case to me.
I am looking into the options for satisfying that requirement and its
impact on the coords model currently in RFC.

Mark

PS: I'll also add this note to the coords model RFC page.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ivoa.net/pipermail/dm/attachments/20191001/f1fa2a5d/attachment.html>


More information about the dm mailing list