Update: bjet cross sections w Andreas

1
Update b-jet cross section(s)(w/ Andreas Greg)
Rob Snihur McGill University CDF Canada mtg.,
Sept. 04, 2007

• Matching Tests
• Jet pT corrections

2
Matching Tests

• Use matching for
• Tagging efficiency (requires a pure sample of MC
  b-jets)
• B-fraction (tagged jets matched to MC b-jets)
• Absolute pt-jet corrections (cal jets to hadron
  jets)
• Matching details matter!
• Metric use delta-r
• Pt ordering
• All jets or selected jets
• Kinematic cuts
• Tag -1?
• Parton ID uds, c, b, g
• Direction of matching

3
Cartoons

• h

Cal Had 1 1 2 2 3
Cal Had 1 1 2 2 3
x
y
x
y
1
3
2
1
2
z
4
Delta R Plots cal vs had

• Top me (old)
• Bottom Greg (new)

5
Deriving jet pT corrections

• On a jet-by-jet basis, correct the measured
  pT(cal) ? pT(hadron)
• Absolute
• Relative
• Offset (due to UE, MI, etc.)
• Must also be consistent with unfolding
  corrections
• Corrects the shape of pT spectrum due to
  resolution effects
• Follow kT, demonstrate 2 methods
• Bin in pT of calorimeter jets
• Bin in pT(mean) (pT(cal) pT(had)) / 2

6
Absolute jet pT correction

• Match cal jets to hadron jets
• Plot
• pT(had) - pT(cal)
• pT(had) / pT(cal)
• Versus pT(cal) pT(mean)
• Extract mean y-values and x-values
• Plot fit it

7
pT(had) in bins of pT(cal)

• h

8
pT(had)-pT(cal) in bins of pT(cal)

• h

9
pT(had)/pT(cal) in bins of pT(cal)
10
pT(had) in bins of pT(mean)
11
pT(cal) in bins of pT(mean)
12
pT(had)-pT(cal) in bins of pT(mean)
13
pT(had)/pT(cal) in bins of pT(mean)
14
Immediate next steps

• From the previous plots, extact mean values. For
  the following plots
• pT(had)-pT(cal) vs pt(cal) in bins of pT(cal)
• pT(had)/pT(cal) vs pt(mean) in bins of pT(mean)
• Fit these with a polynomial

15
Plans and next steps

• Analyze new MC
• Unfolding correction
• Reweighting
• To match the data pT spectrum
• Theory

16
Taggable Efficiency?

• Only taggable jets get a tag value
• Taggable defn Not only pt and eta/y cuts
• See y vs. pt plots of JetClu and SecVtx
• Number of tracks?