Trigger efficiency for high-pT electrons

06 December 2007

The plots below were mede from a few thousand single-electron AODs from 12.0.6 reconstruction. The data samples used were: 7006 (Et=120GeV), 7007 (Et=500GeV), and 7009 (Et=1TeV).  The efficiency for the e25i and e60 signatures was plotted versus the offline electron eta, assuming at most one reconstructed electron per event. In practice there were 26, 13 and 3 events in samples 7006, 7007 and 7009 where 2 electrons were reconstructed. This was considered negligible. The efficiency was obtained directly from the TriggerDecision object and normalised to the offline "loose electron" selection. I.e. for each signature, or combination of signatures, the efficiency was calculated as the number of events which passed the trigger signature and where a loose electron candidate had been found, divided by the number of events where a loose electron candidate had been found. The only selection applied to the events was the loose electron selection ( (isEM & 0x7) ==0 ). The numbers of events analysed for each sample can be found in the log files (see table). They were:  ET=120GeV:  6750 analysed (5972 had electron candidates); ET = 500 GeV: 19500 analysed (14824 had electron candidates); ET = 1 TeV: 4250 analysed (2983 had electron candidates).

A small Athena algorithm was used to produce a reduced TTree. A ROOT macro was used to run on this TTree and produce the plots. Both Athena code and ROOT macro are in the Atlas cvs repository at:
users/jgoncalo/TrigEgammaPerformanceR12/TrigEgammaPerformanceR12-00-00-04


The table below has links to the plots:

 7006 (ET = 120 GeV) : log
 7007 (ET = 500 GeV) : log 7009 (ET = 1 TeV) : log Combined
e25i
 eps, gif
 eps, gif
 eps, gif
 eps, gif (L1)
e60
 eps, gif
 eps, gif
 eps, gif
 eps, gif (L2)
e25i OR e60
 eps, gif
 eps, gif
 eps, gif
 eps, gif (EF)

This combined plot for the Event Filter shows the overall trend: loss of efficiency at high pT:
e255+e60 efficiency vs eta, for different electron ET

For reference, the CSC-EG-11 note documentation is the cvs repository at: /groups/egamma/CSC-EG-11.
RG - last modified 7/12/07