Presentation on theme: "DM WORKING GROUP MADRID OCT 2005. WG STATUS Spectral DM: document updated and reformatted Spectral DM Java library prototype nearly ready Characterization:"— Presentation transcript:
DM WORKING GROUP MADRID OCT 2005
WG STATUS Spectral DM: document updated and reformatted Spectral DM Java library prototype nearly ready Characterization: CDS group held wide discussions. B. Thomas working on Quantity/Char/STC approach. A. Richards presented radio example. Goal for this meeting: agree on new version draft as basis for going to PR next interop. 3D data: I. Chilingarian presenting approach based on SIA and Characterization. Service implementation later in 05.
WG STATUS Jonathan – summary and SED Pedro – catalogs Mireille – characterization Igor – 3D data Matteo – lines Arnold – STC Jonathan - summary
WG STATUS Space-Time Coordinates: Rots proposed STC-Lite for simple use. RFC drew significant comments, rev in work Goal for this meeting: decision to proceed on STC-Lite and agreement on approach. Result: WG does not like the idea of two schemas. Atomic Line Data: Osuna, Dubernet, etc. - first internal draft in work, discussed in Paris last week. Goal for this meeting: gather input and discuss draft. Source catalogs: early draft in preparation (Osuna) Goal for this meeting: reconcile different approaches, gather requirements from other WGs
Summary of current activities Characterization: GOOD – PR by May? Spectrum/SED GOOD – PR by May? Spectral Line GOOD – PR by May if feedback ok Catalog container MEDIUM/GOOD PR by mid 06? Source MEDIUM - research STC DIFFICULTIES Quantity MEDIUM – reuse issue in work
Characterization Describe the context of a dataset in the VO – Revised and improved model seems well received – No remaining controversies in model? – Serialization issue: link with STC, progress made on doing this – New document by Dec 2005 – On track for PR May 2006; Rec soon after?
Describes 1D spectra, time series, SEDs – Proposal to simplify by separating SED aspect from simple 1D spectrum aspect; minor change to document – Need to upgrade to keep in synch with STC and Char. – Missing keywords in FITS serialization need to be added – Still some hesitation about FITS serialization approach – No showstoppers. Expect new document Dec 2005, hope to iterate minor changes to a 1.0 document Jan 2006 to PR – Aim for Rec by May 2006 Spectral Model
Describes theoretical and observational archives of atomic and molecular transition data – DM meeting in Paris last week – First public presentation here – Model is simple and in good shape – v0.2 document public imminently – Need feedback from community but hopeful we can push quickly forward to a v1.0 and a PR by May 2006 Spectral Line Model
Split into Catalog container and Source efforts – Catalog container model has several different versions but they should be easily reconciled. Should be possible to get a v0.5 or so by early 2006 and a PR in mid – Source model is an attempt to provide structured astronomical knowledge e.g. for advanced catalog queries. This work is still at an R&D stage. – Spirited discussion in WG and in pre-Interop meeting. But I don't think there are serious disagreements in the catalog container work. Catalog and Source Models
STC Space-Time Coordinates: – RFC drew significant comments – need to revise – New revision by end of month? – Serialization link with Quantity being studied – F. Ochsenbein: Use of STC in VOTable – unsolved – How to do simple cases? DM group did not like the STCLite idea of having 2 rival schemas; VOEvent group reportedly does not like the Xinclude approach – STC: TECHNICALLY MATURE BUT STILL CONTROVERSIAL – Will propose ways to address these problems by Dec 05
Hot issues Catalogs: – Implicit VOTable model – series of simple tables – Simple cells, tree hierarchy, plus controlled vocab (Williams, McDowell, in general discussion) – Specialized source model with small number of predefined fields (Osuna) – Abstract, general containers to hold any kind of data (Thomas) Need more detailed requirements from other working groups (e.g. registry)
Hot issues STC – Full STC versus role of simple cases and defaults – Requiring all of an STC instance vs reuse of elements – role of Quantity in XML serialization of STC Use cases and plan: – (1) User inspection of metadata: what is this data anyway? requires software to create STC instance and to display – (2) Do my object's coords lie in this dataset? requires stc:inside() function, possibly with coord convs – (3) Convert dataset to new coord system AST (ex Starlink) implementation; CDS, ESO using Regions part
DM strategies We need to deploy VO services useful for astronomers - so we must identify simple but useful things to provide, and model the interfaces DM model document -> class library and DAL access protocol -> implies transmission format Basic reusable DM components - Characterization, STC, Curation, SEDCoord etc. Use these and UCD tree in other specialized models - AtomicLine, Planets etc.
VOs talking to each other IVOA, EGSO, IHY.... It's hard enough for the IVOA folks to talk to each other - we must be realistic about the amount of coordination possible across VOs Ideally, common DM components would help a lot. At a minimum, expressing the mapping between the DMs offers a route to interoperability. Maybe now is the time to identify the relevant people and have an exploratory meeting