New NumFOCUS grant for PyTables

[Francesc Alted]

Hi Developers,

We are happy to announce that we have got a grant for PyTables to leverage Blosc2 and the HDF5 enhanced direct chunk capabilities for *multidimensional* data.   For details on the grant, see: https://opencollective.com/pytables.

This is the logical extension of previous year's grant where we implemented optimized support for Blosc2 for Table objects (https://www.blosc.org/posts/blosc2-pytables-perf/).  With this new grant, we would extend that to CArray and EArray classes, bringing more performance in writing and reading hyperslices (either small or large) from datasets.

For this work, we are happy to announce that we will count with Ivan Vilata, one of the initial authors of PyTables (I'll add him to the list of current developers in the GitHub project soon).  Marta Iborra (from the Blosc team) and myself will also collaborate in the effort.

Regarding the implementation time, we are already starting the work, and hopefully will (mostly) complete it in a couple of months from now.  By then, we hope to have a PR ready for your inspection, and after reviewing and merging, we will do a new release (with binary wheels).  Tell us in case you have questions.

Cheers!

--

Francesc Alted

[Antonio Valentino]

Dear Francesc,

Il 22/06/23 13:13, Francesc Alted ha scritto:

Congratulations for the new grant.
Support for blosc2 in pytables is extremely important IMHO.

Welcome (back) to Ivan and Marta.

Look forward to play with the new features :).

cheers
--
Antonio Valentino

[Ivan Vilata i Balaguer]

Thanks, Antonio! Thanks Francesc for the introduction too.

It's great to be back in PyTables after so many years! I see that it got lots
of attention and care from you developers during this time. 🙂

See you around!

--
Ivan Vilata i Balaguer -- https://elvil.net/