Rivet development meeting

Thursday 10 Oct 2024, 16:00 → 18:00 Europe/Zurich

Zoom only

Monthly developer catch-up on Rivet and related tools.

 

 

16:00 → 16:20 Discussion

Rivet dev meeting 2024-10.pdf

Actions:

Immediate:

1. Aim for a 4.0.x patch release by end of Oct (get back into the pattern of frequent releases with minimal baggage)
2. Andy/Chris to validate the multi-arch Docker build and deployment CI for next release
3. Chris to add a Gitlab issue for chi2(Estimate, Estimate) at Python level, using covariance matrices constructed from the error breakdowns (in numpy)
4. Chris to trial multi-energy merging solution for highest-priority analysis (cf. Natalia), maybe adding a feature to allow analyses to list and access their booked histos by name, to use preloaded multi-energy data from the merging
5. All devs to think about whether some mechanism for disabling / targetting analysis builds would be useful (mainly for devs, maybe useful to users who e.g. only want the ATLAS analyses)
6. Chris to release new YODA with optional/disableable HDF5 feature
7. Chris to post plan for H5 Rivet mode dependence on H5 YODA, and detail of the bundled HighFive headers
8. Andy to contact ATLAS Phys Coord and spread the word of the Rivet tutorial more widely
9. All to discuss on MM and coordinate attending the Munich Aug 2025 MC workshop at the same time: application deadline is end Oct

Longer term:

1. Chris to migrate YODA coverage code to Rivet CI
2. Andy to discuss primary particle definition and LLPs more with Louie and Sihyun
3. Make public wrapper class for EDM4HEP writing from Rivet (AB/Sarah Williams)
4. Andy to add more view filtering on analysis coverage pages (and paginate further)
5. Rivet web page migration to new site... longstanding, needs action

 

16:20 → 16:40 Proposal for a YODA-H5 format

Speaker: Christian Gutschow (UCL (UK))

cg_yoda_hdf5.pdf