towards 3D data model

[Reagan Moore]

Doug:
The Data Format Description Language group of the 
Global Grid Forum is making progress on building 
XML-based representations of the structure of 
arbitrary binary data sets. Within the year it 
should be possible to read a description of the 
structure of a data set, and then automate the 
extraction of data.

Reagan Moore



>Francois -
>
>The characterization data model (also other general dataset data models
>such as identification etc.) apply to all datasets hence would apply to
>3D data as well.
>
>For modeling 3D data in DAL, more is required as we need to model the
>actual dataset.  Currently we have 2D sky images (SIA) and various kinds of
>tabular spectrophotometric data (SSA).  My feeling is that what is required
>for spectral datacubes is a generalization of the 2D image model to 3D.
>This would build upon the spectrophotometric metadata developed for SSA to
>describe the observable, plus FITS WCS to describe the coordinate system.
>
>This is probably what is needed for regularly gridded data such as a
>spectral data cube.  It is also possible that for some data of this sort,
>e.g., MOS or IFU data where the spectra have already been extracted, that
>we should use the SSA data model instead.  It would be easy to extend
>this to collections of spectra, which is what one has in this case.
>
>It is possible that to support spectral data cubes we should just
>fold the support into a future version of SIA - whether or not this is
>advisable depends upon how much it complicates things.  It may be that
>the complications would justify a new data access interface for 3D data.
>In particular, we don't know at this point what is required to cutout
>regions of a large spectral data cube.  We need to understand the science
>drivers better to analyze this, i.e., what sorts of analysis do we want
>to target with this capability.
>
>	- Doug
>
>
>On Mon, 14 Feb 2005, Francois Bonnarel wrote:
>
>>  dear Giovanni, Phillipe and Igor,
>>     I am happy that Euro 3D people will join the IVOA data model effort
>>  This connection was missing since a long time. I am not personnaly sure
>>  that creating a specific 3D data model is the best we can do to achieve
>>  that. There is an ongoing "Characterization" data model effort supposed
>>  to deal with any kind of observations and data sets.  (material can be found
>>  on the list about that, and Giovanni, you can 
>>remember we discussed this last
>>  October). A subgroup is supposed to produce a draft long before Kyoto for
>>  discussion.
>>  My immediate proposal is that somebody among Euro3D people can join the
>>  subgroup  and judge the pertinance of the ongoing work from the 3D point
>>  of view.
>>    Your opinion about that ?
>>  Cheers
>>  François
>>
>>  =====================================================================
>>  Francois   Bonnarel               Observatoire Astronomique de Strasbourg
>>  CDS (Centre de donnees          11, rue de l'Universite
>>  astronomiques de Strasbourg)    F--67000 Strasbourg (France)
>>
>>  Tel: +33-(0)3 90 24 24 11       WWW: 
>>http://cdsweb.u-strasbg.fr/people/fb.html
>>  Fax: +33-(0)3 90 24 24 25       E-mail: bonnarel at astro.u-strasbg.fr
>>  ---------------------------------------------------------------------