Motivation

The main motivation is to understand the trigger correction for electrons in 2013 pPb data.

Cluster Ratio

The cluster ratio (clusters per event in triggered data/ clusters per event in MB data) has been used for the trigger correction, as it takes into account the trigger efficiency in a straightforward. Another solution would be to split the trigger correction into 2 parts: Trigger rejection factor and trigger efficiency for a specific particle -- which includes the methods using in off line analysis.

Trigger and Data:

• EG1 (11 GeV threshold), EG2 (7 GeV threshold)
• The triggers can overlap, EG2 has a downscale (~ 16.5% according to Log Book).
• Triggered sample --> LHC13d,e,f pass2 (Total ~1M EG1 (11GeV), ~ 0.5 M (7GeV))
• MB sample ---> LHC13c pass2 (75 M evts), LHCd,e,f pass2(around 2M evts)
• There is also a L0 trigger (at ~ 3 GeV). LHC13d,e,f (~ 0.4 Mevts)

First Estimation

• To use LHC13def/LHC13c MB, LHC13d,e,f have low MB stats
• A better solution seems to be EGE(1,2)/L0 x L0/MB in LHC13d,e,f. Since L0 threshold is around 3 GeV, there is plenty of statistics for the ratio L0/MB and L1 triggers/L0. As a consequence, this type o estimation reduces the bias that there might be using different periods for triggered and MB data.

Results for the ratios (Examples made by Cristiane).

LHC13e/LHC13c (Threshold ~ 7 GeV)

All the Clusters	Cluster+T.M. requirement

LHC13e/LHC13c (Threshold ~ 11 GeV)

All the Clusters	Cluster+T.M. requirement

LHCdef/LHC13c Both thresholds

EG2 (7 GeV threshold)	EG1(11GeV threholds)

Conclusions

The results are quite different when there is the T.M. requirement.

Remarks/Questions

• Is T.M. efficiency different in the triggered sample?
• Is the electron trigger efficiency really higher than pi0, photons?
  • How can one disentangle these efficiencies in a straightforward way?