Skip to content

[QEE] Remove SingleVHighPtElectron line

Luke Grazette requested to merge lugrazet-qee-remove-VHighElectron into qee_upgrade

Goal

Aiming for %Reoptimised Hlt2 selections for 2025: "lossless" first pass, the SingleVHighPTElectron line was identified as a large BW consumer, relatively high rate and a very high average event size. It also has seemingly no intended analysis and no interested parties. Check with line author if we can turn it off.

Based off of the milestone's starting results: Screenshot_2024-11-18_at_11.48.03

Changes

Removed Hlt2QEE_SingleVHighPTElectronFull and SpruceQEE_SingleVHighPTElectron lines.

Summary

The VHighPT electron line was added in feb-2022, Will:

I think I want to add one more line - without any custom PID type cuts, but with a higher pT threshold (to control the rate) - in case we find that custom PID is less reliable than we hope. Will try to add this today.

This line really only has two parameters, PT cut and prescale.
There's been a few concerns about the line that's been raised since then:

  • 35 GeV is very very high for a PT cut, and is potentially not even viable for electroweak physics usage : https://gitlab.cern.ch/LHCb-QEE/qee-hlt-run3/-/issues/37#note_8424856
  • It takes a large BW even with prescaling.
  • It takes a comparable amount of BW as SingleHighPtMuon line at HLT2 even when prescaled, but that line feeds a whole group of analyses while this line is a currently unused 'safety'.
  • Electron ID should be well understood at this point, so is this line necessary still?

For reference, the other existing lines single-electron lines:

  • Hlt2QEE_SingleHighPtElectronFull
  • Hlt2QEE_SingleHighPtElectronPrescaleFull: High PT single electron line with lower pT threshold, prescaled to reduce the rate.
  • Hlt2QEE_SingleHighPtElectronIsoFull: High PT single electron line with lower pT threshold, and isolation to keep the rate down.

All of the other electron lines were updated to include analysis-level cuts on HcalEoP/EcalEoP/PT from @/jiuzhao in !3363 (merged) which was deemed appropriate for efficiency (91% efficiency for Z->ee signal MC) and background rejection.

BW Changes

!4072 (comment 8752867)

  • A reduction of 170 MB/s (20%) of the qee full-stream BW
  • A reduction of 2 MB/s (1%) of the qee sprucing BW (makes sense as it was already being prescaled at sprucing).

ToDo:

Edited by Luke Grazette

Merge request reports

Loading