Muon KE entering TMS is not right

[mdolce8]

There is something not quite right with the true muon energy as it enters TMS.

I am seeing from running
python draw_spill_3D_projections.py --outdir /exp/dune/data/users/mdolce/dune-tms/spills/truth -f /exp/dune/data/users/kleykamp/tmsreco_combined_files/2024-04-19_bfield_1p0T.tmsreco.root --only_true_tms_muons some unusual (and big) muon KEs from the truth as they enter TMS. I have pasted some of the output I was seeing below. I would say this happens around 10-20% of the time. So there is something not quite right with the true muon energy. It is calculated by using the MomentumTMSStart branch from the truth.

I ran this code from the branch: mdolce_drawspills_truehits. I added a few lines of code to Asa's script that loops through the Truth_info->PDG for each event and retrieves the muon true energy. And that was how I found this issue.

Here are a few examples of the issue:

Event: 0, Spill 6913, Muon KE at birth (LAr): 26.086267578125, Muon KE entering TMS: 173204.97505692, GeV.
Event: 0, Spill 8722, Muon KE at birth (LAr): 1.402595703125, Muon KE entering TMS: 173204.97505692, GeV.
Event: 0, Spill 8875, Muon KE at birth (LAr): 3.2294921875, Muon KE entering TMS: 173204.97505692, GeV.
Event: 0, Spill 9899, Muon KE at birth (LAr): 2.68086328125, Muon KE entering TMS: 173204.97505692, GeV.

[jdkio]

This may be fixed by this PR
#134 (comment)
We've fixed many issues since we made these files. We should remake the files with the most up to date code

[jdkio]

Momentum inside TMS or whatever it's called works better. It's not the momentum at the z position at the start of the TMS. Instead, it's the momentum when it crosses into the TMS fiducial, regardless of angle. For lar starting events, this is the same usually

[mdolce8]

Is it possible to remake the files with the up to date code? Maybe after the collaboration meeting?

Also to your second comment, I think that is what I used: truth.MomentumTMSStart to get the momentum, and calculate the KE from there. So I think this is right.

[jdkio]

Yeah, this is long overdue. I'm working on a validation suite and I'm still uncovering issues. But maybe we can't let perfect be the enemy of good