Time Series Cube DM - IVOA Note

CresitelloDittmar, Mark mdittmar at cfa.harvard.edu
Fri Mar 31 14:49:29 CEST 2017


DMers,

I was reviewing some of the 'modified' model content yesterday and realized
something may have been lost.
I'd like to make a clarification on the requirements with respect to the
(value + error) object which I have in the coords model and here will call
Measurement.

The modeling in the last drop to volute, prior to this discussion was:
  + define pattern for this Object (value == coorvalue + frameaxis
reference ) + errors
  + implement this pattern in the domains needed for the cube work.

This discussion adds 2 points/requirement to fold into this model:
  1) a statement that this Measurement need only be modeled once, a
measurement is a measurement and need not be redefined in each domain.
  2) a generic user (plotter) should be able to find/use the content of
this Object with as little 'specialized knowledge' as possible.

So, from this... I simulated a modeling where the Measurement was concrete
and NOT domain specialized.  Moved all the domain specialization to the
(value + axisref) level.

I think this works for the above new requirements, but I think it breaks
another.

  3) ability to specify a domain-specific Measurement in other objects.
       eg: A Source has a Position.. presumably this is a ((value+axis) +
errors) Measurement restricted to the Spatial domain...similarly Target.pos

I'd like to confirm #3, but I'm assuming this is true, and will be
re-working to see about satisfying all three without getting too messy.

Mark



​
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ivoa.net/pipermail/dm/attachments/20170331/d83ddbaf/attachment.html>


More information about the dm mailing list