Skip to content

Thresholds for calo seeded ROI in the InDet software

Christos Anastopoulos requested to merge ATLAS-EGamma/athena:CaloSeeded_ROI_InDet_thresholds into master

I am trying to revisit some of the discussion from ATLASRECTS-4597

We know that moving to topoclusters from SW , effectively decreased the threshold where:

  • A. We do TRT alone
  • B. We brem recover Silicon

We also have confirmed that increasing the thresholds to the same level as SW will take us back cpu wise.

The SW had a turn on curve that can be seen here https://atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/PAPERS/PERF-2017-01/fig_02a.png yellow points. So ~2.75 GeV seems good compromise for TRT-alone (since topo cluster start even below 1 GeV).

For the rest, we still start now at 1.5 GeV I made 2 bins 1.5-2.5, 2.5-3.5 and tighten ever so slightly the lower part selection.

Mentioning @amete , @rnaranjo and @amorley @nstyles

The ideal situation is to get a slight decrease in spot timing with ~0 effetc in e/gamma ART. as was discussed in ATLASRECTS-4597

For completeness this is for the InDet tracking. Where it allows for a retry with brem hypo or seeding TRT. For selecting tracks, from the totality of InDet tracks, to do GSF obviously different criteria are employed.

Edited by Christos Anastopoulos

Merge request reports