Last tracker hits not used in CKF

[ShujieL]

Environment: (where does this bug occur, have you tried other environments)

main branch. v24.09

Steps to reproduce: (give a step by step account of how to trigger the bug)

1. open a single particle reconstruction file, e.g. 5 GeV pion+ such that for every event, the particle should go through all tracking layers.
2. check the number of events with at least one valid tracker digi hit on the last tracker layer (outer MPGD barrel, or MPGD/AC-LGAD disk).
3. pull out the good measurements that are associated to each track, and check if there is at least one measurement on the last layer.

Expected Result: (what do you expect when you execute the steps above)

The number of events with digitized hits on the last layer should equal (up to the efficiency) to the number of tracks reconstructed with a measurement on the last layer

Actual Result: (what do you get when you execute the steps above)

Attached are the comparison of sim hits, digi hits, and measurement on each layer from
EPIC/RECO/24.09.0/epic_craterlake/SINGLE/pi+/10GeV/45to135deg/pi+_10GeV_45to135deg.0001.eicrecon.tree.edm4eic.root
While most of the MPGD hits are digitized after the threshold change at eic/epic#774, only half of those hits are used in track recon.

[ShujieL]

@simonge suggested me to check the phi distribution of TOF. While that looks alright, the outer MPGD missing hits definitely have some pattern.

@mposik1983 any thoughts?

[mposik1983]

Hi Shujie, thanks for looking at this. Do you happen to have the comparisons for theta, x,y, and z distributions?

…

________________________________ From: ShujieL ***@***.***> Sent: Thursday, October 31, 2024 1:52:07 AM To: eic/EICrecon ***@***.***> Cc: Matthew Posik ***@***.***>; Mention ***@***.***> Subject: [External] Re: [eic/EICrecon] Last tracker hits not used in CKF (Issue #1659) @simonge<https://github.com/simonge> suggested me to check the phi distribution of TOF. While that looks alright, the outer MPGD missing hits definitely have some pattern. @mposik1983<https://github.com/mposik1983> any thoughts? Screenshot.2024-10-30.at.22.48.51.png (view on web)<https://github.com/user-attachments/assets/aa22520e-7f75-44e8-b02a-2693a5b1367a> — Reply to this email directly, view it on GitHub<#1659 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAQ7622PTRSCWEOTPOFQKSDZ6HAQPAVCNFSM6AAAAABQ3H3OVGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINBZGA2TIMJXHE>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[ShujieL]

@mposik1983 Here's the z distribution

[mposik1983]

Hi Shujie, Thanks. So based on these distributions it looks like the hits are only being accepted around the detector perimeter, and hits are missing in the central area of the detector. I imagine if you look at theta vs. phi 2D distribution of the measurements you will see "holes" in the central of the detector planes. This to me smells like a bug in the ACTS detector implementation. I will relay these distributions at Mondays MPGD meeting. Matt Posik Research Associate Professor Temple University Department of Physics SERC 1925 N. 12th St. Philadelphia, PA 19122 USA TU Office: SERC Room 451 Physics Office: SERC Room 406/4th Floor

…

________________________________ From: ShujieL ***@***.***> Sent: Thursday, October 31, 2024 4:57 PM To: eic/EICrecon ***@***.***> Cc: Matthew Posik ***@***.***>; Mention ***@***.***> Subject: [External] Re: [eic/EICrecon] Last tracker hits not used in CKF (Issue #1659) @mposik1983<https://github.com/mposik1983> Here's the z distribution Screenshot.2024-10-31.at.13.55.43.png (view on web)<https://github.com/user-attachments/assets/e04ac4f4-3752-4984-8f0d-613a6695f942> — Reply to this email directly, view it on GitHub<#1659 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAQ762ZSYJVPKYILBH46LDLZ6KKUPAVCNFSM6AAAAABQ3H3OVGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINJQHAYDGMJRGU>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[mposik1983]

The issue with missing measurement hits has been fixed in PR#800 of the epic repository.