Question on GSF track to track association for low pt electrons

[ArturAkh]

Dear Egamma experts,

During my studies to fix tau embedding
for central CMS production, I've had a closer look into low pt electron reconstruction, in particular into:

https://github.com/cms-sw/cmssw/blob/master/RecoEgamma/EgammaElectronProducers/plugins/LowPtGSFToTrackLinker.cc

This producer performs GSF track to track association by making use of a common track reference. To do that, it checks for the right electron seed index in preid vector indexed by electron seeds, and acesses the trackRef of the PreId. I wonder why this indirect access to track references is required, where a much more direct access is possible,

using

gref->seedRef().castTo<reco::ElectronSeedRef>()->ctfTrack()

instead of

preid->at(gref->seedRef().castTo<reco::ElectronSeedRef>().index()).trackRef()

Introducing the following line:

std::cout << "Considering GSF track no. " << igsf << " with seed reference CTF track (id, index): (" << gref->seedRef().castTo<reco::ElectronSeedRef>()->ctfTrack().id() << ", " << gref->seedRef().castTo<reco::ElectronSeedRef>()->ctfTrack().index() << "), determined track ref (id, index) from preid: (" << trk.id() << ", " << trk.index() << ")" << std::endl;

after LowPtGSFToTrackLinker.cc#L57

results in the following output, when running on an example input file:

Begin processing the 11th record. Run 320823, Event 169761289, LumiSection 113 on stream 0 at 27-Jun-2023 14:36:38.789 CEST
Considering GSF track no. 0 with seed reference CTF track (id, index): (9:1653, 0), determined track ref (id, index) from preid: (9:1653, 0)                                                                                                                                   
Considering GSF track no. 1 with seed reference CTF track (id, index): (9:1653, 1), determined track ref (id, index) from preid: (9:1653, 1)                                                                                                                                   
Begin processing the 12th record. Run 320823, Event 169393937, LumiSection 113 on stream 0 at 27-Jun-2023 14:36:39.414 CEST
Considering GSF track no. 0 with seed reference CTF track (id, index): (9:1653, 0), determined track ref (id, index) from preid: (9:1653, 0)                                                                                                                                   
Begin processing the 13th record. Run 320823, Event 170028693, LumiSection 113 on stream 0 at 27-Jun-2023 14:36:40.214 CEST
Considering GSF track no. 0 with seed reference CTF track (id, index): (9:1653, 1), determined track ref (id, index) from preid: (9:1653, 1) 

So all in all, both accesses to the track reference are equivalent in terms of id() and index().

Is there something I'm missing about preid, like backward-compatibility, or some kind of a corner case?

Cheers,

Artur

[cmsbuild]

A new Issue was created by @ArturAkh Artur Gottmann.

@Dr15Jones, @perrotta, @dpiparo, @rappoccio, @makortel, @smuzaffar can you please review it and eventually sign/assign? Thanks.

cms-bot commands are listed here

[makortel]

assign reconstruction

@cms-sw/egamma-pog-l2

[cmsbuild]

New categories assigned: reconstruction

@mandrenguyen,@clacaputo you have been requested to review this Pull request/Issue and eventually sign? Thanks

[swagata87]

tagging developer of low-pT electron reco @bainbrid

[ArturAkh]

Hi all,

More than two months have passed...

Is there any progress on this issue? I may create a pull request, since the change outlined above is valid by construction, if looking at the corresponding producers.

Cheers,

Artur

[bainbrid]

Hi @ArturAkh please go ahead if you feel the code is an improvement.

[ArturAkh]

Thanks @bainbrid,

Corresponding PR is created: #42690

[mandrenguyen]

+reconstruction
Issue seems resolved by #42690

[cmsbuild]

This issue is fully signed and ready to be closed.

[makortel]

@cmsbuild, please close