Add support for RapidSim names

[admorris]

I added support for the RapidSim particle-naming convention (oh how I wish they just used EvtGen names!)

It largely follows the same strategy as the LHCb names, but with the important difference of loading from $RAPIDSIM_ROOT/config/particles.dat if present, which is nice if you're using particle in e.g. a conda environment with rapidsim installed.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 64.58333% with 17 lines in your changes missing coverage. Please review.
✅ Project coverage is 97.30%. Comparing base (352f0c6) to head (5f019ef).

Files with missing lines	Patch %	Lines
src/particle/rapidsim/converters.py	48.27%	15 Missing ⚠️
src/particle/rapidsim/__init__.py	83.33%	1 Missing ⚠️
src/particle/rapidsim/data/__init__.py	83.33%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #756      +/-   ##
==========================================
- Coverage   97.89%   97.30%   -0.60%     
==========================================
  Files          31       35       +4     
  Lines        2619     2667      +48     
==========================================
+ Hits         2564     2595      +31     
- Misses         55       72      +17     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[admorris]

To improve the coverage, I should add a test that mocks $RAPIDSIM_ROOT being set

[eduardo-rodrigues]

Hello @admorris, I will look at this next week ...

One immediate reaction is indeed why RapidSim uses non PDG/EvtGen standard names in the 2020s!? It's more complication for absolute zero gain.

The obvious question is whether it would not be simpler in the end to sort the issue at source? In the end they must be using names as labels, right, so should be rather easy? Have you checked this with the authors?

I am not saying I am opposing to having another submodule as we already have for LHCb. Just asking for the record.

Thanks.

[admorris]

Hi Eduardo

That's a fair suggestion.

I just had a go at converting RapidSim's particles.dat to use EvtGen names, and I found 4 instances where the PDG ID does not match 🥲 particles.dat.txt

There are a lot of open PRs and issues in RapidSim which suggest it's not being actively maintained. But I'm sure Greig and Dan are still contactable.

[eduardo-rodrigues]

Hi.

I checked briefly your data file and all 4 misses are for wrongly assigned PDGIDs from RapidSim authors. I could find the correct replacements in https://github.com/scikit-hep/particle/blob/main/src/particle/data/conversions.csv.

I don't know how you and LHCb would like to proceed, but that seems like the main point here, else I am happy to merge this as-is if that helps with support for LHCb.

I see many LHCb analysts still using RapidSim, hence it is useful and should be maintained. If the authors are not active anymore, could the package be taken over by LHCb somehow? I see it is effectively dead, since the last release is from 2019 and the latest merge from 2 years ago - https://github.com/gcowan/RapidSim!

[eduardo-rodrigues]

There is now an extra line to skip given my added comment, right?

[eduardo-rodrigues]

This looks great and I only made a couple of trivial suggestions.

Another suggestion would be to add somewhere in the docs the link to RapidSim, probably in admin/dump_pdgid_to_rapidsim.py but also in the functions to and from?

I let you decide if you update the docs for this new functionality here or in a follow-up PR.

Last thing: did you then decide what to do overall, I mean given the exchange above? Even if some work is going to be done to revive and modernise RapidSim, it may still take a while before that's available. If helpful we can move on merging and releasing this asap. Just let me know.

Thank you again for the great and clear enhancement.

[eduardo-rodrigues]

Hi @admorris, I saw that RapidSim got moved to the LHCb organisation. Let me know your decision on this work.

[admorris]

I had a quick chat with Michal K a few weeks ago about it, and I think we agreed it's better not to break backwards compatibility with the names, but also that it should be easier for the user to provide their own particle table. At the moment you need to copy the whole config directory and override the parts you need.

The PDG ID bugs should be fixed though IMO.

I am undecided on how that impacts whether this PR is needed. Someone who wants to use Particle is probably proactive enough to provide their own particle table to RapidSim (of course we could ship one with it and make it a flag).

[eduardo-rodrigues]

OK, then we leave this open and see. Maybe you can chat to 1-2 usual suspects and see what their take is?

I agree that at least the PDG ID bugs ought to be fixed. Thanks.

[eduardo-rodrigues]

Hi. I know this is in standby. Just to say, FYI, there's been in the past days a great enhancement that simplifies conversions enormously, making them better designed, see #780.