Retune inclusive HNL T track lines
This merge request aims to retune the T track HNL lines utilising some of the bandwidth cleared up by !5173 (merged) .
In particular, it aims to to:
- Weaken the displacement cut on the T track vertex to recover misreconstructed downstream tracks
- Loosen the mass cut on the N -> mu pi line to capture more N -> mu pi X decays (with X mainly a D meson)
The former change is particularly important as downstream tracking in Run 3 is known to have inefficiencies, especially for BSM particles. See, for instance this presentation here.
However, often misreconstructed downstream tracks have correctly reconstructed T track stubs. This allows the possibility of recovering these downstream events as T track events. In the attached plot, the tracking efficiency of HNL events, B_c -> mu (N -> mu pi) is shown where the mu mu pi final state is 'reconstructible' (i.e. has hits) as a long and two downstream tracks. The tracking efficiency of the state as the (correct) LDD final state is shown in blue. In orange is the tracking efficiency as a long and two T tracks; LTT. This shows something like ~60% of these downstream events can be recovered.
Given the short timeline for the review, these changes are deliberately kept minimal.
The per-line bandwidth impact is summarised below, but is minimal (+4 MB/s as opposed to the -30 MB/s from !5173 (merged)) [N.B. This now includes the loosened mass cut's effect as well. As expected, the change is minimal; +3 MB/s -> + 4 MB/s)
| Line | Reference Rate (kHz) | Reference Total Bandwidth (MB/s) | Comparison Rate (kHz) | Comparison Total Bandwidth (MB/s) | Difference in Rate (kHz) | Difference in Total Bandwidth (MB/s) |
|---|---|---|---|---|---|---|
| Hlt2QEE_N_PiMu_inclTTDecision | 0.0516 | 4.76968 | 0.0774 | 7.0915 | 0.0258 | 2.32182 |
| Hlt2QEE_N_EMuNu_inclTTDecision | 0.0774 | 5.36182 | 0.0903 | 6.93244 | 0.0129 | 1.57063 |