Applications Working Group : June 2025

VOParquet Note and implementations

The VOParquet convention that describes associating rich VOTable metadata with bulk data in Parquet format was published as an IVOA Note in January 2025. I will briefly recap the content of this Note, and report current implementations of it, as well as some utility tools available in STILTS for validating, debugging and tuning files written using the VOParquet convention.

We present the Hierarchical Adaptive Tiling Scheme (HATS), a spatial indexing framework developed to enable efficient querying and cross-matching across massive astronomical datasets. HATS adapts to non-uniform sky coverage and data density, providing a scalable and flexible structure for modern survey data. In this talk, we will outline the structure of the HATS IVOA Note, discuss current implementations, and invite feedback on community adoption.

VOTable has had basic unicode support from day one; there is the unicodeChar datatype – which would work just fine if there was only TABLEDATA. Alas, in BINARY serialisation, unicodeChar is supposed to do UCS-2, which has been obsolete since at least 2011 and is hardly implemented anywhere anymore (if you think you have UCS-2, you probably have UTF-16). On the other hand, we now live in a world of UTF-8, and 8-bit encodings are a thing of the past. Perhaps it is time to allow UTF-8 in char and deprecate unicodeChar?

PyVO Update

Current 1.6.2 release and future plans.

Multi-Order HEALPix Map implementation in CDS HEALPix Rust

Multi-Order HEALPix Maps (MOMs) are kind of MOCs in which a value is associated to each cell. It's also an extension of BMOCs to non-boolean values. An example of application is (chi2) compressed density maps. After an introduction and examples, we are going to present the choices made in the Rust implementation, and how MOC, BMOC and MOM serializations could be unified (also for Time and Frequency), and operations performed from streamed inputs, thanks to the ZUNIQ numbering.

Data format needs for LISA

LISA will detect a range for gravitational wave sources, both transient and persistent. While the first prototypes for the data processing are being developed, we are starting to imagine the format for this new type of data so that it can be used by a wide range of astronomers.