[PWGLF] Add customV0 pairing for Sigma+ PCM photon

[hfribert]

@fmazzasc

This PR adds the option to build the photon candidates itself instead of using V0Datas:

• This is needed since the SVertexer is built for V0's from the primary vertex, which is not the case for the Sigma+'s photon (weak decay: Sigma+ --> p + pi0 --> p + 2 gamma (only 1 gamma reconstructed)). In MC local tests, this yields ~3x as many true Sigma+ as the V0Datas.
• Pairing similar to the SVertexer: precuts on the daughters' polar angle difference and on the circles touching, then a vertex fit in every compatible collision (TPC-only daughters moved to the collision time, as in the strangenessbuilder)
• For TPC-only daughters, the compatible collisions are taken from their full TPC time range
• The proton daughter determines the collision (has ITS). For the photon deduplication (same photon in several candidates), only the Sigma+ candidate with the smallest DCA of the reconstructed mother to the PV is written
• Also included additional photon & candidate variables, as well as histograms, for the direct comparison between the two different methods and for additional cuts in post-analysis

[github-actions]

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

[romainschotter]

Hi @hfribert ! It looks good. However, it is not clear to me how this is different than what is currently being done in the propagationService. Can you please elaborate a bit more?
Maybe we could discuss it tomorrow at the PAG-Strangeness meeting?
Thank you very much!

[hfribert]

Hi @romainschotter, thanks for having a look! The propagationService, as far as I saw, takes the SVertexer V0s as input. The SVertexer however applies cuts towards the primary vertex (e.g. mTPCTrackMaxDCAXY2ToMeanVertex), which seem to be too tight for many of the Sigma+ decays (after a couple of cm's). Many of these photons therefore wouldn't be available via the propagationService V0s.

[romainschotter]

In propagationService, deduplication is performed at the photon-building stage, whereas here it is done on the photon mother (the Sigma). Is there a specific reason or benefit to handling it this way?
Thank you very much!

[hfribert]

The reason to deduplicate using the mother is that the photon does not point to the PV. For photon-based deduplication, I would also need to use the proton daughter, e.g., the CPA of the photon to the decay vertex (p + 2gamma). I also tried deduplicating using the proton-photon dca, which seems to yield fewer true candidates than taking the reconstructed mother dca-to-PV as a criterion.

[romainschotter]

(sorry misclick)

[alibuild]

Error while checking build/O2Physics/code-check for 3502361 at 2026-10-05 21:32:

## sw/BUILD/O2Physics-code-check-latest/log
--
========== List of issues found ==========
++ echo 'Found 8 errors and 7 warnings.'
Found 8 errors and 7 warnings.
++ [[ 8 -gt 0 ]]
++ cat /sw/BUILD/26fac670f82450b48678cf3157501ed10d4b91bf/O2Physics-code-check/errors.txt
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:303:28: error: no header providing "TAxis" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:304:12: error: no header providing "size_t" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:310:29: error: no header providing "TH1" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:333:29: error: no header providing "TH2" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:642:33: error: no header providing "o2::aod::track::extensions::TPCTimeErrEncoding" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:707:21: error: no header providing "o2::math_utils::CircleXYf_t" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:730:43: error: no header providing "o2::aod::track::TrackTimeAsym" is directly included [misc-include-cleaner,-warnings-as-errors]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:784:77: error: no header providing "o2::constants::lhc::LHCBunchSpacingNS" is directly included [misc-include-cleaner,-warnings-as-errors]
++ [[ 7 -gt 0 ]]
++ cat /sw/BUILD/26fac670f82450b48678cf3157501ed10d4b91bf/O2Physics-code-check/warnings.txt
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:416:26: warning: use 'contains' to check for membership [readability-container-contains]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:642:5: warning: uninitialized record type: 'timeEncoding' [cppcoreguidelines-pro-type-member-init]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:643:27: warning: do not access members of unions; consider using (boost::)variant instead [cppcoreguidelines-pro-type-union-access]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:656:21: warning: use designated initializer list to initialize 'TrackCand' [modernize-use-designated-initializers]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:690:33: warning: use designated initializer list to initialize 'SVCand' [modernize-use-designated-initializers]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:1348:37: warning: casting (double + 0.5) to integer leads to incorrect rounding; consider using lround (#include <cmath>) instead [bugprone-incorrect-roundings]
PWGLF/TableProducer/Strangeness/sigmaplusbuilder.cxx:1707:5: warning: use auto when initializing with a template cast to avoid duplicating the type name [modernize-use-auto]
++ [[ 8 -gt 0 ]]
++ exit 1
--

Full log here.