Dijet Analyses for ICD Resolution and Response

1
Dijet Analyses for ICD Resolution and Response

• Barry Spurlock
• The University of Texas at Arlington
• April 8, 2004

2
OUTLINE

• Motivation
• DPCC and How We Used It
• Data
• Analyses Results
• Conclusions

3
Motivation

• Dijet study of QCD background (P13)
• ICD-MG Task Force (ICD weighting factor)
• Initial analyses performed
• 150k JT95TT events
• More data
• fixed corrected
• Needed more processing power
• DPCC

4
DPCCDistributed Parallel Computing Cluster

• Joint effort including UTA CSE dept., UTA HEP
  dept., and UT Southwest Medical Center (Dallas)
• 97 machines/193 processors
• Pentium 4 Xeons at 2.4-2.6 GHz
• Over 72 terabytes disk space
• The DPCC is supported, in part, by the NSF grant
  EIA-0216500

5
General Proceedure

• Submit_to_PBS
• Assign job names
• Submits run_in_batch
• run_in_batch
• Copy data files, root macro, tarfile
• File list root macro are chosen using the job
  name
• Tarfile includes C program, header, MakeClasses
• Runs the root macro as a ROOT batch job
• Root macro makes classes, chains the data files,
  assigns values to global variables, runs code
• Compilation problems

6
Data Data Processing

• Started with thumbnails from QCD group skim of
  P13 jet data
• 200GB, 1281 files, 9.7M events
• tmbfixer d0correct
• Used d0tools to run
• Took time to find right version of tmbfixer
• d0correct compiled locally
• Skimmed out JT45TT, JT65TT JT95TT
• Skimmed out events with exactly two good JCCB
  jets, back-to-back (10)
• 0.5ltemflt.95, chflt0.4, hotlt10.0, n90gt1.0
• Events 170k JT45TT, 160k JT65TT, 37k JT95TT

7
Analyses

• ICD resolution versus ICD weighting factor
• Dijet asymmetry and response by eta for various
  ICD weighting factors
• Least squares best fit

8
ICD Resolution

• CC-ICD and CC-MG events used
• Cut on narrow ET band of CC jet
• Sigma/E½ used as the resolution of probe
• Resolution plotted against ICD factor for 6 ET
  bins (60, 75, 85, 100, 120, 150)

9
Dijet Asymmetry Response

• CC control jet in narrow ET range
• Other jet used as probe
• Plotted asymmetry and response by eta bin
• Plots made for 6 ET bands (60, 75, 85, 100, 120,
  150) and various values of ICD weighting factor
• 1.0, 1.5, 2.0, 2.4, 2.8, 3.0, 3.2, 3.4, 3.6, 3.8,
  4.0, 4.2, 4.4, 5.0

10
JT65TT/100 GeV/Factor 1.0
11
JT65TT/100 GeV/Factor 1.5
12
JT65TT/100 GeV/Factor 2.0
13
JT65TT/100 GeV/Factor 2.4
14
JT65TT/100 GeV/Factor 2.8
15
JT65TT/100 GeV/Factor 3.0
16
JT65TT/100 GeV/Factor 3.2
17
JT65TT/100 GeV/Factor 3.4
18
JT65TT/100 GeV/Factor 3.6
19
JT65TT/100 GeV/Factor 3.8
20
JT65TT/100 GeV/Factor 4.0
21
JT65TT/100 GeV/Factor 4.2
22
JT65TT/100 GeV/Factor 4.4
23
JT65TT/100 GeV/Factor 5.0
24
Best Fit (Least Squares)

• ET from probe jet is broken into parts
• ETicd icdf ET ETccmg ccmgET
• ETch chfET ETecmg ecmgET
• ETrem ET ETicd - ETccmg - ETecmg Etch
• A linear combination of these is set equal to the
  ET of the control jet.
• A least squares fit is used to calculate the
  values of the factors for each part
• This was performed over a variety of eta ranges
  and ET bands.
• This was also performed where some of the
  coefficients were fixed to 1

25
Example of Least Squares Fit
26
Results/Conclusions

• Submitting 133 jobs at 3 hours each and having it
  done the next day is very nice!
• But getting all the results in such a short time
  is a bit overwhelming.
• The results will be discussed at the ICD-MG task
  force meeting tomorrow
• At a quick glance, 3.8 seems about right.