Wrong muon direction

[tcuisset]

The muon nb 5 of event 1458825976 (first one in file : /eos/user/t/tcuisset/fireworks/H4muMartina2024_multiEvts.root) seems to be out of place. It has positive eta but is drawn in the negative endcap. Also the muon chambers look wrong.

See here for the original report.
The event is from run 380074, lumi 746, event 1458825976 of /Muon0/Run2024C-MINIv6NANOv15-v1/MINIAOD

[alja]

@tcuisset

Thank you for the report!

Muon track propagation has a couple of different cases:
https://github.com/alja/FireworksWeb/blob/main/Muons/src/FWMuonBuilder.cc

I will debug which one has the error.

[alja]

@tcuisset

The highlighted muon track has five track fitting points defined in reco::Muon::matches(). The source is here proxy builder :

Trajectory point (101.078545, -119.547157, 587.650024)
Trajectory point (101.078644, -119.547279, 587.650024)
Trajectory point (145.313721, -162.522797, 816.161255)
Trajectory point (169.901337, -188.055145, 947.838745)
Trajectory point (186.004639, -206.174515, 1037.338745)
Trajectory point (214.713745, -436.675323, -975.200012)
Trajectory point (278.810394, -411.896088, -1065.694946)
Trajectory point (97.047211, -115.134903, 565.710022)
Trajectory point (-147.216248, 80.412315, -571.469971)
Trajectory point (97.765839, -115.838524, 569.549988)

The 3 matching muon chambers are located in the negative Z direction.
The track propagation tries to fit all 10 of them.

Here is a screenshot of the same muon, where you can see the 3 muon chambers in the negative Z direction:

I don't have the permissions to access the original report in https://cms-pub-talk.web.cern.ch/t/physics-briefing-for-moriond-2026/47670/5. It is hard to tell where the bug is.
Perhaps reco::Muon data is not consistent. The possibility is that the geometry is incorrect.