Online Monitoring of high pt electron trigger efficiencies

1
Online Monitoring of high pt electron trigger
efficiencies Geneva collaboration in the TB LAr
software
Manuel Diaz Gomez
Meeting ATLAS-GVA
8th December 2004
2
Talk Outline

• Introduction
• Trigger efficiency from data with Z?ee-
• Status Validation of DC2 Z?ee- events
• Geneva collaboration _at_ Test Beam LAr software
• Conclusions

3
Introduction

• High Pt electrons are involved in many important
  physic channels
• Offline validation of electron selection cuts
  using Monte Carlo (i.e TDR studies), gives a good
  hint about the expected selection efficiencies,
  but
• We dont know what the real background will be
  until ATLAS
• starts taking data, and therefore we cant
  just blindly trust Monte
• Carlo estimations
• Therefore we will need at some point to have a
  quick rough estimation of how the trigger
  system(e/gamma slice in particular) is behaving
  in an running environment with real data
• We define a method that will allow us to
  calculate single
• electron trigger efficiencies using real data

4
Trigger efficiency from data with Z?ee-
SINGLE TAG e25i
(Isolated electron with Ptgt25 GeV)
offline
Level2/EF
DOUBLE TAG 2e25i
offline
Level2/EF
Nz number of true Zs B- background entries
within mass window (fakes, etc..)
5
Zee DC2 production - Status

• Release 9.0.1 is out! (bug fix of 9.0.0)
• Zee DC2 production can still take a while (end of
  march)
• My thesis cant wait!!
• The output of DC2 wont be a ntuple(as in DC1),
  but an AOD (analysis object data)
• Need to define an event data model (tedious!)
• Stored in data base ? persistency classes
  (difficult in c!)
• The Trigger community wont make it for DC2
  (lack of manpower?)
• Production (good statistics) in a local cluster
  should not be a problem o)
• 2000 Zee events reconstructed so far (low
  luminosity, no pile-up)

6
DC2 Zee Pythia kinematics
7
DC2 Zee Level 2 Calorimeter Reconstruction
Distribution OK, butEnergy leaks outside the
clustering window is not recovered
25 of initial Zs reconstructed
8
DC2 Zee Level 2 Track Reconstruction
Pythia
IDScan
Pythia
SiTrack
IDScan
SiTrack
Only good matching tracks
All reconstructed tracks
Track reconstruction efficiency seems a bit low
82, Pt resolution OK , but need to investigate
further Bremmstralung effects
IDScan
SiTrack
9
Level 2 Trigger With Test Beam Data

• At combined test beam
• runs taken with ROD
• DSPs in physics mode
• ? energy and time
• Input to online Level 2
• trigger software
• Byte stream converter not
• ready
• ? only offline analysis
• Ongoing work to verify
• analysis chain and timing

10
LArDigits vs. LArRawChannels

• First reconstruction steps applied to 350 GeV
  proton data (run 2101614)
• Total energy and clusters reconstructed from raw
  data
• Total energy and clusters reconstructed directly
  from ROD/DSP digits
• Reasonable agreement ? different calibration
  constants used!

11
Conclusions

• A new method is being developed in order to
  provide online monitoring of high-Pt electron
  trigger efficiencies
• First results using DC1 fully simulated data
  proved feasibility
• Still more exhaustive study needed (i.e
  background)
• New Geant 4 simulation available (DC2)
• Despite many technical problems, reconstruction
  of interesting data should be rather quick (Uni,
  Trigger clusters)
• GVA participated actively in this year TB LAr
  software
• Physics analysis of TB data Starts now!
• TB reconstruction software released (9.1.1)
• GVA will be actively involved (Hayk, Imma,
  Arno,..)
• Next year will be very active! (DC2, thesis)