[DPG] Added unit test for validating reconstruction software

[alcaliva]

No description provided.

[github-actions]

O2 linter results: ❌ 2 errors, ⚠️ 0 warnings, 🔕 0 disabled

[alcaliva]

Hi @ktf,
the checks on build/O2Physics/o2/macOS and build/O2Physics/o2/macOS-arm keep failing since yesterday. Do you perhaps know what is the issue? Thanks a lot!
Alberto

[ktf]

@singiamtel is looking into it

[f3sch]

@alcaliva, there is also some very basic task we used in the past to validate the thinning procedure, which may could be used for this validation here, may need some adaptions for the MC tables though

[alcaliva]

Hi @f3sch,
thanks! We might also use that. I think that this code is sufficient to test randomness in the reconstruction software. If the pt distributions of the same events are exactly equal, then we can use this as a validation tool.

[f3sch]

indeed. Anyways, I think but I might be very well mistaken, even on cpu the tpc reco is not deterministic when using multiple threads, so a 1:1 mapping might not be possible.

[alcaliva]

Still, I think that quantifying the deviations due to this effect would be good.

[alibuild]

Error while checking build/O2Physics/o2 for bfe34e8 at 2025-09-11 23:55:

No log files found

Full log here.

[alcaliva]

Hi, can this PR be approved? the failure of o2/macOS is not caused by this

[ktf]

@singiamtel , I think you need to cleanup also this builder. It's probably the stale branch.