November 2007

Fully hadronic channel

• Some features :
  • Large QCD backgrounds
  • Need not to take into account the missing energy (neutrinos)
  • Largiest branching ratio ~ 46%
  • Two leading jets are considered as b-jets
    • b-tagging are not performed in this analysis but it will be taken into account later

Framework and samples

• Framework
  • CMSSW_1_5_2
  • SusyAnalyzer V01-01-00
  • ROOT
• Samples
  • tt0j_mT_70_alpgen ~ tt4j_mT_70_alpgen
    • DBS data discovery page
    • QCD backgrounds are contained at the generation stage.

Preliminary

Jet efficiency

• Tried to make efficiency plots by varying the number of jets in events
• Tried to see the effect to the efficiency by adjusting the parameters corresponding to jest in analysis.cfg
  • Event rejection decision : rej_BadHardJet = TRUE ro FALSE
  • Primary and final accpetance pt cut, 20 GeV and 30 GeV, respectively
  • Also eta cuts for primary and final acceptance : Max = 5
  • Extrapolation error adjustment(in cm) :
    • reco_jetD0ErrorThresh = 0.05
    • reco_jetDzErrorThresh = 0.10
    • reco_bjetD0ErrorThresh = 0.5
    • reco_bjetDzErrorThresh = 0.20
• 4jets event for example
  • Rejection Bad Hard jet = True
    • jetD0 and DzErrorThresh : jetD0ErrorThresh = 0 & jetDzErrorThresh = 0 ( bjet cases are non-zero )
    • 
    • bjetD0 and DzErrorThresh : bjetD0ErrorThresh = 0 & bjetDzErrorThresh = 0
    • 
  • Rejection Bad Hard jet = false
    • Extrapolation error adjustment are all zero
    • 
    • Extrapolation error adjustment are non-zero
    • 
• Simple conclusion :
  • Efficiency increased when the extrapolation error adjustment implemented, especially non-bjet cuts
  • Rejection bad hard jet cuts can decrease the efficiency slightly

Event selection

Sum of non-leading jests

• Non-leading jet total transverse energy obtained removing the two most energetic jets
• Sum_3 >= 148 GeV
• 

Verifying jet-parton-matching

• The six partons in fully hadronic channel are matched to six reconstruced jets by picking the matching which minimises the sum of the angular seperation between reconstructed jet and matched parton. CMS Notes
  • good jet-parton-matching : Each of the six partons and jets differ only by 15(deg) and the jet-reconstructed tops also differ only by 15(deg) from their corresponding parton-level direction.
  • half-good jet-parton-matching : Three of the partons and jets forming one top differ only by 15(deg) and this jet-reconstructed top also differ only by 15(deg) from his corresponding parton-level direction.
  • bad jet-parton-matching : Everthing else.
• SusyAnalyzer has own MatchObject class, so I tried to verify how it is efficient by using this method.
• As results, the value of 15(deg) is somewhat arbitrary but all matching objects are good.
• i.e. total six jets and partons, so 6 * 15 = 90(deg). the distribution for good jet-parton-mathing shows well below 90(deg)

The other event selection cuts

Centrality

• Centrality C >= 0.68
  • the fraction of the hard scatter energy going in the transverse plane.

Aplanarity

• Aplanarity A >= 0.024
  • A = 3/2 * Q : Q being the smallest of the three normalized eigenvalues of the sphericity tensor
  • Choose 3rd eigenvalues in TMatrixDEigen Class
• Sphericity tensor
  • http://cepa.fnal.gov/psm/simulation/mcgen/lund/pythia_manual/pythia6.3/pythia6301/node213.html
• Calculating Eigen values of Sphericity tensor
  • How to use the linear algebra packages(LAPACK) in ROOT
    • Matrix references : http://root.cern.ch/root/html/MATRIX_Index.html
  • TMatrixT, TMatrixDEigen