Data and services
Arnold Rots
arots at head-cfa.cfa.harvard.edu
Thu Sep 18 14:32:32 PDT 2003
Bob,
A while ago you asked about how to describe regions in a non-XLM
format. It occurred to me that Wil and I had a discussion that might
be helpful for your purposes. See the attached memo.
It was written for the query element, but I have attached a version
for the resource profile.
Another issue you raised was about how to handle limit transitions in
cyclical coordinates in the specification of polygons (like: which way
around when you go from RA=300 to RA=100).
I can see two solutions to this:
1. Allow values less than 0 as well as greater than 360:
- the example above would mean going "to the right"
- had you wanted to go the other way, you should have specified RA=300
to RA=460
2. Do not allow circle segments that are 180 degrees or more:
- the example would mean going "to the left"
- the other way would have required an intermediate point at, say, RA=200
I like (1) better and it is also easier to specify, but I suspect that
(2) is easier to parse.
- Arnold
Robert Hanisch wrote:
>
> ----- Original Message -----
> From: "Tony Linde" <ael at star.le.ac.uk>
> To: "'Ray Plante'" <rplante at poplar.ncsa.uiuc.edu>; "'Robert Hanisch'"
> <hanisch at stsci.edu>
> Cc: <registry at ivoa.net>
> Sent: Friday, September 12, 2003 3:18 PM
> Subject: RE: Data and services
>
>
> > I'd forgotten about Coverage. Maybe SkyService should stay so that when
> > solar and stp people create their schemas they can have a SolarService and
> > STPService.
> >
> > Maybe we need a DataService which includes a pointer to the root data
> > collection plus any other data-ish metadata and then SkyService,
> > SolarService and STPService can inherit from that?
> That might be ok. Though perhaps we don't have to build this structure just
> yet.
>
> > And should Coverage be renamed to SkyCoverage? Or is there enough
> > commonality between 'coverage' in the disciplines to deal with the
> > differences with one or two optional elements?
> Solar, Planetary, and STP will have different concepts for coverage.
>
> Bob
>
--------------------------------------------------------------------------
Arnold H. Rots Chandra X-ray Science Center
Smithsonian Astrophysical Observatory tel: +1 617 496 7701
60 Garden Street, MS 67 fax: +1 617 495 7356
Cambridge, MA 02138 arots at head-cfa.harvard.edu
USA http://hea-www.harvard.edu/~arots/
--------------------------------------------------------------------------
Incorporating STC Structures in VO Queries
==========================================
Define the following parameters (purely for convenience):
define CSys = (TimeFrame = (TimeScale='TT') and (TimeRefPosition='TELESCOPE'))
and (SpaceFrame = (CoordFrame='ICRS') and (CoordRefPosition='TELESCOPE'))
and (SpectralFrame = 'OBSERVATORY')
and (CoordFlavor = 'SPHERICAL')
define CArea = (TimeInterval = (StartTime=(ISOTime='2000-01-01'))
and (StopTime=(ISOTime='2002-12-31')))
and (Region = (Circle = (Center='60.0d,30.0d') and (Radius='10arcmin'))
or (Polygon = ((Vertex='23.0d,46.0d'),(Vertex='20.0d,46.0d')
(Vertex='20.0d,48.0d'),(Vertex='23.0d,48.0d'))))
and (SpectraLinterval = (LoLimit='1GHz') and (HiLimit='2GHz'))
define CSpec = (Time = (CoordResolution<='10s'))
and (Pos2Vector = (CoordResolution<='10arcsec,10arcsec'))
and (Spectrum = (CoordResolution<='50MHz'))
Specify the query:
select <SelectList> from <ResourceList> where <WhereClauses> and
SearchLocation = (CoordSystem = $CSys and CoordArea = $CArea
[ and CoordSpec = $CSpec ])
One could, of course, imagine pre-defining some defaults.
If one were to encapsulate the STC part in an XML document, the query
would reduce to:
select <SelectList> from <ResourceList> where <WhereClauses> and
SearchLocation = (http://MyPlace/MySearchLocation.xml)
Or one could make a hybrid:
select <SelectList> from <ResourceList> where <WhereClauses> and
SearchLocation = ((CoordSystem = http://nvo.org/standardCoordSys.xml)
and CoordArea = $CArea [ and CoordSpec = $CSpec ])
Resource Profile
================
Specify the coverage:
ResourceProfile = (CoordSystem = $CSys and CoordArea = $CArea and CoordSpec = $CSpec)
where:
define CSys = (TimeFrame = (TimeScale='TT') and (TimeRefPosition='TELESCOPE'))
and (SpaceFrame = (CoordFrame='ICRS') and (CoordRefPosition='TELESCOPE'))
and (SpectralFrame = 'OBSERVATORY')
and (CoordFlavor = 'SPHERICAL')
define CArea = (TimeInterval = (StartTime=(ISOTime='1999-07-23')))
and (Region = (Circle = (Center='60.0d,90.0d') and (Radius='180d')))
and (SpectraLinterval = (LoLimit='0.1keV') and (HiLimit='12keV'))
define CSpec = (Time = (CoordResolution>='0.000016s'))
and (Pos2Vector = (CoordResolution>='0.5arcsec,0.5arcsec'))
and (Spectrum = (CoordResolution>='2eV'))
Or do it in one expression::
ResourceProfile = (CoordSystem =
(TimeFrame = (TimeScale='TT') and (TimeRefPosition='TELESCOPE'))
and (SpaceFrame = (CoordFrame='ICRS') and (CoordRefPosition='TELESCOPE'))
and (SpectralFrame = 'OBSERVATORY')
and (CoordFlavor = 'SPHERICAL')
and CoordArea =
(TimeInterval = (StartTime=(ISOTime='1999-07-23')))
and (Region = (Circle = (Center='60.0d,90.0d') and (Radius='180d')))
and (SpectraLinterval = (LoLimit='0.1keV') and (HiLimit='12keV'))
and CoordSpec =
(Time = (CoordResolution>='0.000016s'))
and (Pos2Vector = (CoordResolution>='0.5arcsec,0.5arcsec'))
and (Spectrum = (CoordResolution>='2eV')))
It's not as bad as it looks at first, and it translates easily into
STC XML elements.
More information about the registry
mailing list