[Obscore Update] requirements from SIAv2 data access protocol

[Louys Mireille]

Hi all,

Thanks for your feedback.
It seems we agree on the content ( see below) and to expose them as a 
list of attributes that we can wrap them in one single class, to update 
the Obscore Data model.
Up to now the proposed name is : DimensionAndMappingSummary, long but 
descriptive , I admit.

Other suggestions could be :
- ArrayCoosDimensions
- ArrayCoosDimsNames
- ArrayDimAndCoos
- ArrayDimAndCalibLabels
-...

What would you think of those?
Any other suggestion?

In the meantime I will write an addendum for the ObsCore specification 
and submit it to the list by the end of this week.
Thanks for your comments ,
Mireille.

Le 31/07/2014 23:17, CresitelloDittmar, Mark a écrit :
> Mireille/Francois,
>
> Thanks for pinging this back to the top of my mail list :)
>
>
> On Thu, Jul 31, 2014 at 3:27 AM, François Bonnarel 
> <francois.bonnarel at astro.unistra.fr 
> <mailto:francois.bonnarel at astro.unistra.fr>> wrote:
>
>>     In order to keep the general approach applied for the Obscore
>>     data model , we suggest to define an additionnal class with the
>>     following data model fields :
>>
>>         1. Number of subarrays (usually but not always 1)
>>          "num_array"    integer
>>
>>         2. Number of array axes (dimensionality) of the main array
>>           "num_axes"   integer
>>
>>         3. Length of each axis
>>            "axis_length"  integer array, e.g., "512 512 3" is the
>>     simplest representation
>>
>>         4. Number of wcs axes  (may differ from num_axes)
>>             "num_wcs_axes"  integer
>>
>>         5. Coordinate type on each WCS axis
>>            "axis_type"    whitespace delimited list of axis types
>>             e.g RA-TAN DEC-TAN WAVELENGTH
>>
>
>     Sounds good. I think it is important that these additional new
>     class attributes could be referred somewhere by the time SIAV2
>     comes to recommandation.
>
>
> I agree, this content looks fine for ObsCore.
> If I recall, the goal is to have this ObsCore 1.1 update in place to 
> be referenced by the SIAV2 document.. correct?
>
>>     These form a coarse grain description of the Dataset dimensions
>>     and Mapping that could be called 'DimensionsAndMappingSummary'.
>>     This class can be referenced in Obscore, from the general
>>     Observation class, and in ImageDM from ObsDataset, or in the
>>     specialised Image classes : NDImage, SparseData if necessary.
>>
>>     Such a definition keeps the genericity of the Obscore DM. It
>>     should be reused for full interoperability in Image/Cube DM as a
>>     summary class.
>>     This is also the case for many classes in the CharacterisationDM .
>>     Here 'DimensionsAndMappingSummary' is a bridge between the two
>>     models at a coarse level of description of the 'Data' object.
>>
>     This is a very good and important point. Inclusion of this class
>     in ImageDM is important for future interoparability of several
>     protocols
>     So if adopted this class should be integrated in ImageDM diagrams
>
>
> I will be working heavily on the next draft over the next weeks...
> I agree that this summary class would be populated from the greater 
> ImageDM content.  I don't want to hi-jack this thread, so we
> should discuss how that connection is made separately.
>
>> I picked up 'DimensionsAndMappingSummary' as a class name and would 
>> be happy to get your feed back , whether it is understandable
>> with its attributes' list and if any shorter name can be proposed.
> I don't have any strong opinion about the lengthy, but very 
> descriptive  'DimensionsAndMappingSummary'.
>
> Mark
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/dm/attachments/20140819/763a9cce/attachment.html>