• 1) SUS-13-007: covers the single lepton + btag using the Lepton Spectrum (LS) and Delta Phi (DP) methods. Lepton studies are documented in AN-2013-035. Analysis notes: AN-12-390 and AN-12-416. Pre-approval date: Fri, Feb 1, 2013. Targeting Moriond.

  • 2) AOD rereco is used for periods A, B, and part C; prompt reco is used for part C and D.

  • 3) Efficiencies of leptonic and hadronic legs are shown to be not strongly correlated, thus they are measured separately. The lepton legs are mesured via Z T&P in events that are acquired by the analysis triggers (therefore with correctly biased kinematics).
    • The triggers used for electrons:
      • HLT CleanPFHT350 Ele5 CaloIdT CaloIsoVL TrkIdT TrkIsoVL PFMET[45,50] 
      • HLT CleanPFNoPUHT350 Ele5 CaloIdT CaloIsoVL TrkIdT TrkIsoVL PFMET[45,50] 
      • HLT CleanPFHT300 Ele15 CaloIdT CaloIsoVL TrkIdT TrkIsoVL PFMET[45,50] 
      • HLT CleanPFNoPUHT300 Ele15 CaloIdT CaloIsoVL TrkIdT TrkIsoVL PFMET[45,50] 
      • HLT CleanPFHT300 Ele[40,60] CaloIdVT CaloIsoVL TrkIdT 
        (control trigger)
      • HLT CleanPFNoPUHT300 Ele[40,60] CaloIdVT CaloIsoVL TrkIdT 
        (control trigger)
    • and for muons:
      • HLT PFHT350 Mu15 PFMET[45,50] 
      • HLT PFNoPUHT350 Mu15 PFMET[45,50] 
      • HLT PFHT400 Mu5 PFMET[45,50] 
      • HLT PFNoPUHT400 Mu5 PFMET[45,50] 
      • HLT Mu[40,60] PFHT350 
        (control trigger)
      • HLT Mu[40,60] PFNoPUHT350 
        (control trigger)

Summary tables:

Quantity PAT Object and Member Function barrel cut endcap cut comment*
type pat::Electron      
p_T ele->pt() ≥ 20 GeV    
abs(eta)
(ele->isEB() || ele->isEE()) && fabs(ele->superCluster()->eta()) <= 2.5 
     
Rel. Isolation charged + max (0., photon + neutral - (*eleRho)*Aeff) <0.15 same  
hadOverEM
ele->hadronicOverEm()
<0.12 <0.10  
abs(deltaPhi)
fabs(ele->deltaPhiSuperClusterTrackAtVtx())
<0.06 <0.03  
abs(deltaEta)
abs(ele->deltaEtaSuperClusterTrackAtVtx())
<0.004 <0.007  
sigmaIetaIeta
ele->scSigmaIEtaIEta()
<0.01 <0.03  
Conversion rejection
!ConversionTools::hasMatchedConversion(*gsfel,hConversions,beamSpotPosition)
True True Use the gsfel obtained with originalObjectRef; don't use the pat::Electron
missingHits
ele->gsfTrack()->trackerExpectedHitsInner().numberOfHits()
<= 1 <= 1  
dxy
fabs(ele->gsfTrack()->dxy(vertexPosition))
< 0.02 same vertexPosition is the position of the first good vertex
dz
fabs(ele->gsfTrack()->dz(vertexPosition))
< 0.1 same  
abs(reco - PF) PT
fabs(electron->pt() - (*ipfEl).pt())  
< 10 GeV same
(*ipfEl).particleId() == reco::PFCandidate::e && (*ipfEl) > 10.

veto electron selection

Quantity PAT Object and Member Function barrel cut endcap cut comment*
type pat::Electron      
p_T ele->pt() ≥ 15 GeV    
abs(eta)
ele->isEB() || ele->isEE()
     
Rel. Isolation charged + max (0., photon + neutral - (*eleRho)*Aeff) <0.15 same  
hadOverEM
ele->hadronicOverEm()
<0.15 no cut  
abs(deltaPhi)
fabs(ele->deltaPhiSuperClusterTrackAtVtx())
<0.8 <0.7  
abs(deltaEta)
abs(ele->deltaEtaSuperClusterTrackAtVtx())
<0.007 <0.01  
sigmaIetaIeta
ele->scSigmaIEtaIEta()
<0.01 <0.03  
dxy
fabs(ele->gsfTrack()->dxy(vertexPosition))
< 0.04 same vertexPosition is the position of the first good vertex
dz
fabs(ele->gsfTrack()->dz(vertexPosition))
< 0.2 same  

Muons are selected using the POG tight muon recommended selection at https://twiki.cern.ch/twiki/bin/view/CMSPublic/SWGuideMuonId

Quantity PAT Object and Member Function Cut Comment
muon type
pat::Muon
   
ID
muon -> isGlobalMuon() 
True  
ID
muon -> isPFMuon()
True  
pT
muon -> pt()
≥ 20 GeV  
abs(eta)
muon-> eta() 
≤ 2.4  
pf Rel. Isolation
pfRelIso =  (muon->pfIsolationR03().sumChargedHadronPt 
+ max(0., muon->pfIsolationR03().sumNeutralHadronEt 
+ muon->pfIsolationR03().sumPhotonEt - 0.5*muon->pfIsolationR03().sumPUPt ) ) 
/ muon->pt()
< 0.12 Don't use pat::Muon->isoDeposit()->depositWithin(0.3)
normChi2
muon->globalTrack()->chi2() / muon->globalTrack()->ndof()
≤ 10  
nValMuonHits
muon->globalTrack()->hitPattern().numberOfValidMuonHits()
> 0  
numMatchedStations
muon->numberOfMatchedStations()
> 1  
pixelHits
muon->innerTrack()->hitPattern().numberOfValidPixelHits() 
> 0  
numTrackerLayersWithMeasurement
muon->track()->hitPattern().trackerLayersWithMeasurement()
> 5  
dxy
fabs(muon->innerTrack()->dxy(vertexPosition))
< 0.02 vertexPosition is the position of the first good vertex
dz
fabs(muon->innerTrack()->dz(vertexPosition))
< 0.5  
abs(reco - PF)
fabs(muon->pt() - (*ipfMu).pt()) 
< 5 GeV
(*ipfMu).particleId() == reco::PFCandidate::mu && (*ipfMu) > 10.

Muon Veto Selection

Quantity PAT Object and Member Function Cut Comment
Mu type
pat::Muon
   
ID
muon->isGlobalMuon() || muon->isTrackerMuon() 
True  
ID
muon -> isPFMuon() 
True  
pT
muon -> pt() 
≥ 15 GeV  
abs(eta)
fabs(muon -> eta())
≤ 2.5  
pf Rel. Isolation
pfRelIso =  (muon->pfIsolationR03().sumChargedHadronPt 
+ max(0., muon->pfIsolationR03().sumNeutralHadronEt 
+ muon->pfIsolationR03().sumPhotonEt - 0.5*muon->pfIsolationR03().sumPUPt ) ) 
/ muon->pt()
< 0.2  
dxy
fabs(muon->innerTrack()->dxy(vertexPosition)) 
< 0.2 vertexPosition is the position of the first good vertex
dz
fabs(muon->innerTrack()->dz(vertexPosition)) 
< 0.5 vertexPosition is the position of the first good vertex

  • 5) Lepton reconstruction and ID efficiency scale factors between data and MC were determined with the T&P method. Backrgound subtraction (and simulation) was applied in the fit. The scale factors are computed as the ratio of constant fits on the efficiencies over the threshold value.

Selection data/MC (full simulation) data/MC (fast simulation)
Electron ID + Isolation (average) 0.9987 ± 0.0015(stat.)±0.02(sys.) 0.9613 ± 0.0014(stat.)±0.08(sys.)
Barrel   0.9900 ± 0.0007(stat.)±0.03(sys.)
Endcap   0.8725 ± 0.0014(stat.)±0.07(sys.)
Electron GSF tracking 0.9944 ± 0.0003(stat.)±0.02(sys.) 0.9902 ± 0.0004(stat.)±0.01(sys.)
Muon ID1 0.9981 ± 0.0007(stat.)±0.02(sys.) 0.9531 ± 0.0007(stat.)±0.03(sys.)
Muon ID2 1.0015 ± 0.0006(stat.)±0.01(sys.) 0.9902 ± 0.0005(stat.)±0.01(sys.)
Muon track 0.9953 ± 0.0001(stat.)±0.005(sys.) 0.9951 ± 0.0001(stat.)±0.005(sys.)

Trigger efficiencies: scale factor is to be applied on Monte Carlo. Individual triggers are listed with scale factors and systematic errors. "Overall" is a value that is compatible with the result of every trigger in the table. In case of the muons, additionally, a correction is to be applied on the scale factor different for barrel and endcap to account for eta dependence.

Electron trigger path Scale factor
CleanPFHT350_Ele5_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET45     
0.96 ± 0.04
CleanPFHT350_Ele5_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET50     
0.96 ± 0.04
CleanPFNoPUHT350_Ele5_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET45 
0.94 ± 0.07
CleanPFNoPUHT350_Ele5_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET50 
0.95 ± 0.08
CleanPFHT300_Ele15_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET45    
0.97 ± 0.07
CleanPFHT300_Ele15_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET50    
0.97 ± 0.07
CleanPFNoPUHT300_Ele15_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET45
0.98 ± 0.03
CleanPFNoPUHT300_Ele15_CaloIdT_CaloIsoVL_TrkIdT_TrkIsoVL_PFMET50
0.98 ± 0.04
CleanPFHT300_Ele40_CaloIdVT_TrkIdT                              
0.96 ± 0.02
CleanPFNoPUHT300_Ele40_CaloIdVT_TrkIdT                          
0.96 ± 0.02
CleanPFHT300_Ele60_CaloIdVT_TrkIdT                              
0.96 ± 0.02
CleanPFNoPUHT300_Ele60_CaloIdVT_TrkIdT                          
0.96 ± 0.02
Ele30_CaloIdVT_TrkIdT                                           
0.96 ± 0.02
Ele25_CaloIdVT_CaloIsoVL_TrkIdVL_TrkIsoT_TriCentralPFNoPUJet30  
0.97 ± 0.02
Overall 0.96 ± 0.05

Muon trigger path Scale factor
IsoMu24                  
0.88 ± 0.02
IsoMu24_eta2p1           
0.90 ± 0.01
Mu40_PFHT350             
0.91 ± 0.02
Mu40_PFNoPUHT350         
0.91 ± 0.02
Mu60_PFHT350             
0.91 ± 0.02
Mu60_PFNoPUHT350         
0.92 ± 0.01
PFHT400_Mu5_PFMET45      
0.92 ± 0.05
PFHT400_Mu5_PFMET50      
0.92 ± 0.05
PFNoPUHT400_Mu5_PFMET45  
0.93 ± 0.04
PFNoPUHT400_Mu5_PFMET50  
0.93 ± 0.04
PFHT350_Mu15_PFMET45     
0.93 ± 0.05
PFHT350_Mu15_PFMET50     
0.93 ± 0.05
PFNoPUHT350_Mu15_PFMET45 
0.92 ± 0.04
PFNoPUHT350_Mu15_PFMET50 
0.92 ± 0.05
Overall 0.91 ± 0.04
Scale factor for abs(eta) < 0.9 1.076
Scale factor for abs(eta) &gt 0.9 0.928

  • 6) Fakes are attributed to QCD and are derived from data. (eg Section 4.2 in AN-12-390)
  • 7) Both answers are no.
  • 8) Systematic errors on trigger and selection efficiencies are given in 5), more details are in AN-13-035.

-- ViktorVeszpremi - 2013-02-15

Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r3 - 2013-02-15 - ViktorVeszpremi
 
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback