Follow up for my question in Characterisation Data Model RFC

[François Bonnarel]

Hello fabien,
    I think the discussion is important , but only for the future 
because as Mireille
allready wrote the current implementation in xml is based on the axis first
approach only.
    For the Property first approach some people insisted that it can be the
best suited way to describe the coverage of a dataset with non 
independant axes
Of course this is really interesting at level 3 (support) and 4 
(sensitivity) of characterization where
data intervals and Response functions will really benefit from being 
defined in combined
spaces. A sensitivity on (x,y,lambda) can be such that it cannot be 
expressed as a
product of s(x,y)*s(lambda)... and then the bounds and central position
cannot be so easilly inferred from the separated supports ....
   X ray astronomers have some exmaples of that ,an we will assess that 
for the
next version ....

  Cheers
François

Fabien Chereau wrote:

> Hello,
>
> This was my question on the twiki 
> (http://www.ivoa.net/twiki/bin/view/IVOA/CharacterisationDataModelRFC):
>
>> "Hello, instead of having many collections of Properties for each
>> instance of Characterisation (like Resolution, SamplingPrecision
>> etc..) one for each axis as shown in Figure 4, why don't you have
>> only one collection of a CharacterisationAxis class which itself
>> would contains the properties for this axis? I think it would make
>> the design clearer. It is also convenient because to each
>> CharacterisationAxis instance can be associated a single coordinate
>> frame which also apply to each property."
>>
>> * Response (by IVOA.Mireille.Louys): The XML representation of the
>> model has this feature , that is each CharacterisationAxis is a node
>> underwhich we have the properties: coverage, resolution, Sampling for
>> this axis. 
>
>
>> For the model we want to keep the symetry , the
>> matrix-like design that is shown in the different examples. One could
>> search / group the metadata in a Property first order and have , for
>> example Resolution , for all axes, then Coverage, for all axes,
>> etc... 
>
>
> I don't understand why we would want to order the metadata in a 
> Property-first order since knowing what the value of the meta data is 
> without knowing for which CharacterisationAxis it is associated is not 
> very useful. For example in figure 4, if I go through the list of 
> Resolution and find that the first element has a resolution of 5.78, 
> it is meaningless if I don't know in which CharacterisationAxis (and 
> thus coordinate system and unit) it is. So at the end we always need 
> to link both, and it looks simpler to me if we decide that the 
> properties for a CharacterisationAxis are just its attributes.
>
>> When the axes are not independant, that is Space depends on
>> time, for instance, Resolution may be represented by a multi-variable
>> function of all axes, and will be hooked in the Resolution Class.
>
>
> In the general (and real) case where the axes are not independent, 
> having only one network of inter-connected generic 
> CharacterisationAxis classes (with methods or attribute such as 
> getCentralPosition, getBounds etc.) will be easier to manage than 
> having many Property classes connected to each other. Especially 
> because the properties of one axis are all derived from the same 
> quantity (for example the bounds are just the min/max of the support, 
> the Central Position is just the barycenter of the support etc..).
>
> Sorry if what I say is not very clear, I will try to explain what I 
> mean with more details in a future email.
>
> Fabien