[ImageDM] Mapping

[Arnold Rots]

I want to come back to this statement (notwithstanding Doug's hope to avoid
this):




*In the case of STC, we did a comparison of STC to FITS WCS a while back.
STC does have some capability for this, but it is quite limited compared to
FITS WCS.*
It is just plain wrong. The transformations incorporated in STC are
perfectly
equivalent to the way FITS WCS handles them. It's all contained in the
PixelFrame: reference pixel, reference value, scale (optional angle) or CD
matrix,
and projection type.
The list of projection types is shorter, but it contains the most common
ones and
obviously can be extended. The PC matrix is missing and so are PV and PS,
but I believe that in practice the CD matrix or the CDELTs cover most, if
not all,
cases.
And the PixelCoordSystem allows multiple PixelFrames, facilitating linking
the
pixel array to alternate coordinate systems.
So, all the functionality is there and the mapping between STC and WCS is
unambiguous, in both directions, which tells me that there is no good reason
to embed WCS mappings in VO cube metadata, since there is a perfectly good
mechanism provided by IVOA standards.

Cheers,

  - Arnold


-------------------------------------------------------------------------------------------------------------
Arnold H. Rots                                          Chandra X-ray
Science Center
Smithsonian Astrophysical Observatory                   tel:  +1 617 496
7701
60 Garden Street, MS 67                                      fax:  +1 617
495 7356
Cambridge, MA 02138
arots at cfa.harvard.edu
USA
http://hea-www.harvard.edu/~arots/
--------------------------------------------------------------------------------------------------------------



On Mon, Nov 25, 2013 at 6:02 PM, Douglas Tody <dtody at nrao.edu> wrote:

> Hi Mark -
>
> I got the impression earlier that you were suggesting doing this by
> adding additional axes to Characterization; sure sounded like it.  In
> any case, for ImageDM/Spectral CoordSys is a place we can put transforms
> that don't fit into either Char or STC - FluxFrame/Photometry is an
> existing example.  In principle Mapping could be be moved there, however
> the complexity and size issue alone is sufficient to argue against this
> (also encapsulation etc. as I noted earlier).  You yourself argued a
> while back that Mapping was Data element specific and should be modeled
> as part of the Data element.
>
> In the case of STC, we did a comparison of STC to FITS WCS a while back.
> STC does have some capability for this, but it is quite limited compared
> to FITS WCS.  Hopefully we don't have to wade through all that again
> (hah!).  Compatibility with FITS WCS and easy transformation to and from
> our VO representations is mandatory if we wish to have any up-take by
> the non-VO community since the community is heavily invested in WCS,
> both in archive data and in client software.  STC could possibly be
> supported as an optional representation, e.g., to specify an output
> projection, and might work for simple projections so long as someone can
> work out the transformation to/from WCS.
>
>         - Doug
>
>
>
> On Mon, 25 Nov 2013, CresitelloDittmar, Mark wrote:
>
>  Doug,
>>
>> The Image and Spectral models have CoordSys outside of Characterisation,
>> and that object is described as basing from the STC astroCoordSys (?)
>> element.
>> So, I am not suggesting we replicate anything.. the STC recommendation
>> already supports the definition of relations (transforms) between
>> coordinate frames.
>> I am suggesting that we use that.
>>
>> I'll do my best to get a diagram of what I have in mind out to the group
>> today.. the Observation/Dataset separation took up my time this weekend.
>>
>> On Sun, Nov 24, 2013 at 6:07 PM, Douglas Tody <dtody at nrao.edu> wrote:
>>
>>  On Sun, 24 Nov 2013, CresitelloDittmar, Mark wrote:
>>>
>>>  ... I reiterate that the Mapping
>>>
>>>>
>>>> information, which defines coordinate systems, should be contained
>>>> within
>>>> the CoordSys umbrella, using existing VO standards as much as possible.
>>>>
>>>>
>>> The existing WCS formalism (as captured in Mapping) does a lot more than
>>> just define the coordinate systems used in the Mapping.  The current WCS
>>> model is comparable in size to Characterization.  Are we suggesting
>>> trying to replicate all of this within the Characterization model (e.g.,
>>> the CD matrix, tabular coordinate value / index arrays, etc.).  Note
>>> also, that in defining a WCS we (or a Photometric calibration and the
>>> like) are no longer merely defining the characteristics of the dataset.
>>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/dm/attachments/20131205/1c33f0b2/attachment-0001.html>