UML diagrams for Observation data model

[Anita Richards]

> From: Mireille Louys <louys-at-newb6.u-strasbg.fr>
> Date: 2004-08-31Z14:58:20
>
> I've just installed new UML diagrams for the IVOA data model see the web
> pages at :  http://alinda.u-strasbg.fr/IDHA/dmobsIvoa/ Main changes from
> the previous version presented in Boston Interoperability meeting
> emphasize the Characterisation Class and the related classes: Location,
> Support , Bounds, etc...


Dear Mireille,

Well, almost everything I can think of is covered! I am amazed that is is
so concise, I was worried that (with all my nit-picking) we were building
a monstrosity but it is very concise and not intimidating. I cannot
find anything major missing; I think that some of the provisions e.g.
"DataReductionCuration" will turn out to be particularly useful.  I am
impressed that you ahve managed to serialise Characterisation and I will
try and impliment it on some data sets.

I looked at most of the classes and compared it with the Spectral model as
that was in my mind.... I have just a few questions:


------------------------------------------------------------


3rd Diag on the web page as linked above - bottom left corner

  StandardQuantity
      /      \
    Flux    Pos

these are just examples, not the only things allowed?

------------------------------------------------------------

Is there any provision for linking to information about position etc.
accuracy, e.g. photometric and astrometric calibration source catalogues?
"BibliographicReference" class might cover it but in some cases these will
not be published papers with a bibcode, or not in the useful form, but
electronic catalogues either in Vizier or at observatories.

------------------------------------------------------------

Back to my spectral hobby-horse:  How do I describe a data set which has
e.g. 1024 spectral channels identically spaced in frequency between e.g.
1709 - 1710 MHz?  Using "Filter" class minimalWavelength=1709 MHz etc. and
Resolution spectral domain = 1 kHz?

"Filter" class
Definition of minimalWavelength seems strange

"Grating" class
I am not sure what this does that isn't covered by filter, unless it is
just there as an example of domain-specific detail.

"Resolution" class
 in spectral domain : the linespread function
- what does this expect?  A single-value width and/or an expression e.g.
Gaussian parameters?

cheers

a


- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Dr. Anita M. S. Richards, AVO Astronomer
MERLIN/VLBI National Facility, University of Manchester,
Jodrell Bank Observatory, Macclesfield, Cheshire SK11 9DL, U.K.
tel +44 (0)1477 572683 (direct); 571321 (switchboard); 571618 (fax).