Data Model Working Group Sessions: May 2024, Sydney, Australia

Back to main program

DM Session 1: High Energy focus

DM 1 - Tuesday, May 21, 14:00 -- 15:30, C122
Speaker Time Title Materials
Mathieu Servillat 10' + 2' Overview based on the Note prepared by the HE group

Following earlier discussions over severals years, a High Energy group formed at IVOA [1] in 2023, and started to write an IVOA Note [2] to explain the specificities of High Energy Data and potential objectives for a High Energy Interest Group within IVOA.


Ian Evans 15' + 3' Why High Energy Data Are Different

We review high energy astrophysics data, using Chandra X-ray Observatory data as an example. Some key differences in calibration and data analysis approaches that are used when handling energetic photons are discussed, together with the implications for data models. We also consider more broad data analysis techniques that are used routinely by Chandra observers and in the Chandra Source Catalog (and that are being used ever more widely throughout the astronomical community), and why measurement models must be expanded to handle them.

Bruno Khelifi and Mathieu Servillat 10' + 2' High Energy Data in Gamma-ray Astronomy

Extending on the previous presentation on X-ray data, we stress additional specificities of Gamma-ray astronomy data, particularly for Cherenkov astronomy. The Gammapy project and VODF (VHE Open Data Format) initiative will be presented to expose those specificities.

Mathieu Servillat and all 15' + 3' Experience building ObsCore service for HE data

Photon counting in High Energy Astronomy leads to the generation of lists of events used as a starting point for further data analysis. Though ObsCore can expose data product of type "event", such services are virtually non-existent in the VO. We initiate the discussion on how ObsCore can be used for High Energy event-lists based on the current service hosted at Paris Astronomical Data Centre exposing the H.E.S.S. public data release.

Mark Cresitello-Dittmar 15' + 3' Data models: compatibility with High Energy data

A brief review of the data models supporting high energy data (eg: event lists), and possible paths for enabling use cases with more complex data products and workflows
    Open Discussion  

DM Session 2: General Session

DM 2 - Wednesday, May 22, 11:00--12:30, C122
Speaker Time Title Materials
Paul Harrison 15' + 3' !VODML enhancement requests and Data Model Development

A series of extensions to VO-DML will be presented, with a roadmap to VO-DML 1.2 - all the proposed new features of VO-DML 1.1 (and some of 1.2) have been tested withing the VODML tooling for some time now. Some of the new features of the VO-DML tooling such as JSON schema will also be presented.

In addition some ideas for a different Data Model roadmap will be presented.

Laurent Michel 15' + 3' MANGO, a Model for Annotating Generic Objects

Mango is a DataModel proposal designed to represent open-ended collections of complex properties such as we find in data tables.
We consider a property to be complex if it is built with multiple parameters, possibly associated with complex errors, flags or even other properties.
The purpose of Mango is to facilitate the processing of data tables property by property rather than column by column, which requires clients to infer the property components.
The properties supported by Mango can be modeled by classes imported from other data models (Meas/Coords, photDM …) or by built-in classes.
François Bonnarel 15' + 3' !FoV data model: integrating a rendering class

Mathieu Servillat 15' + 3' One-step Provenance updates

One-step Provenance proposes a simplified structure to describe the provenance of an entity as a succession of steps, based on the IVOA Provenance Data Model. We will present the last updates on the propose data model, as well as improvement of the voprov Python package to handle this concept.

Joint Session:

DM/DAL/APPS - Thursday, May 23, 16:00 -- 17:30, C122
Speaker Time Title Materials
Mark Cresitello-Dittmar 15' Introduction

This will primarily be a discussion session to consider options for how Data Models are to be delivered and consumed by clients (DAL, APPS) with vastly different requirements.
The IVOA Data Models to date have been focused primarily on Data Discovery and Access via interactions with database tables. The results of these are served by simple VOTables and/or DataLink to native format files. More recent Data Model work ( Provenance, Cube family, Mango ) have been directed to more complex use cases and are consequently, more detailed than is desired to serve simple use cases directly. In addition, we have the prospect of integrating the Common Archive Observation Model (CAOM) into our Data Model suite, which has significant concept overlap with existing models, but serves its target clients very well.

How do we develop/present models rich enough to serve the most complex cases AND serve clients who don't require that level of flexibility WITHOUT creating a maze of data models with overlapping content? Bring your ideas!
Laurent Michel 10' Epic Propagation project: case in point

    Open Discussion  
Edit | Attach | Watch | Print version | History: r9 < r8 < r7 < r6 < r5 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r9 - 2024-05-15 - MathieuServillat
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright © 2008-2024 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback