Z0->ee Analysis

1
Z0-gtee Analysis

• By Denis O. Damazio (BNL)
• (06/28/2004)
• Presented by Srini Rajagopalan

2
Introduction

• Z0-gtee analysis can be used for EM calibration
  studies.
• Z0-gtee may be used for simulation, reconstruction
  and physics performance evaluation.
• Z0-gtee is also suggested as a parton-parton
  luminosity monitor. (also W/-).
• Needs to understand trigger and Reconstruction
  efficiencies.

3
Framework

• The following analysis was performed using the
  Reconstruction in Offline software version 7.2.0.
• Data set from Z0-gtee run 002513 (dc1). DICE cut
  eta lt 2.7 and a e/- Pt cut of 17 GeV (300K
  events initially produced reduced to 149K).
• The 300 roottuple files produced by
  reconstruction are parsed and produce reduced
  data sets for different environments (truth,
  iPat, xKalman, clusters, Jets, L1 Trigger and
  general Rec).
• Such special roottuples are parsed to produce the
  pictures.

4
Truth Analysis (Z0)

• Truth is checked for the inclusivity of the data
  sample.
• Z0 Pt and Mass (fitted by Breit-Wigner curve.

5
Truth Analysis (e/-)

• The electron cuts (in Pt and Eta) can be easily
  spotted.

6
Reconstructed Clusters

• The Pt distribution of the clusters with
  eg_IsEM0 is similar to truth.
• IsEM clusters have smaller leakage.

7
Reconstructed Clusters

• E over p.
• EM clusters eta (showing the cracks).

8
Reconstructed e/-

• To evaluate the rec. performance, clusters away
  from truth electrons are discarded.
• For different distances Delta Pt is calculated.

9
Reconstructed e/-

• For different Pt Ranges, the Delta Pt between
  Truth and Rec is measured. Calorimeter resolution
  is worst for small Pt.

10
Reconstruction (Z0)

• Z0 features are reconstructed from the e/-
  analysis.
• Z0 Mass and Pt for truth and rec.

11
Reconstruction (Z0)

• Z0 P and Eta distribution for truth and Rec.
• P suffers from small reconstruction losses.

12
L1Trigger

• Trigger ROIs Pt.
• 2 2EM(10)
• 6 2EM(10)1EM(20)

13
Conclusions Future Steps

• Upgrade Analysis to Offline Software version
  8.2.0 (Python stuff). This work has already been
  started however, some reconstruction packages do
  not work (tracks and trigger).
• Look at DC2 preproduction files (used by
  Fabiola). Same problem as above.
• Start to use the AODs in order to add flexibility
  to the analysis (with Ketevi Assamagan's help).
• Start a deeper analysis of LVL1 and LVL2
  efficiencies (with Steve Armstrong's help).
• Expand to other data sets. Pile-up?! Background?!

14
Web Pages

• The latest results about this run and others
  being still analysed can be found at the
  following URL. http//www.usatlas.bnl.gov/damazi
  o/log/
• Comments and Suggestions are very welcome.
• The ROOT code can also be distributed.