Monte Carlo Validation

Goal

To write a package to do streamline Monte Carlo Validation. Action items:

• Things to check for top:
  • W(qq) inv mass
  • W(lnu) inv mass distinguishing the three possible flavors
  • top (qqq) inv mass
  • top (qlnu) inv mass
  • dN/dn_genjet
  • 1st and 2nd associated jet pT and eta
  • branching ratio of ttbar decays (had vs semilep) vs ttbar

• Things to check for Z and W
  • All similar, except the top mass of course.

Users

• ThiagoTomei
• FlaviaDias
• MarcoDias

Procedures

Test files

• /hdacs/shared/tt1j.root # /hdacs/shared/tt2j.root

CVS

• Set up CVS:

• Check out the code - make sure to be inside a directory CMSSW_X_Y_Z/src/SomeAnalyzer beforehand!

• Check in the code after making changes.

Links

• http://lcgapp.cern.ch/project/simu/HepMC/HepMC20110/html/classHepMC_1_1GenEvent.html
• http://lcgapp.cern.ch/project/simu/HepMC/HepMC20110/html/classHepMC_1_1GenParticle.html
• http://lcgapp.cern.ch/project/simu/HepMC/HepMC20110/html/classHepMC_1_1GenVertex.html

Flavia's Log

• To reconstruct t and tbar, I located the vertex that has a t (tbar) ingoing, and took the outgoing particles of the vertices. However, some of these top vertices also got particles other than the W and b: uubar, gluon. I've (under Thiago's advice) just ignored it and tried to reconstruct the top mass with just W's and b's. Solved: status 2 particles.

• Status stuff - status 2 particles for showering.

• To compile with TFileService, take care to put the lines in the BuildFile in the right place (lines above, not below).

• Successfully reconstructed invariant mass of the W's from hadronic and leptonic decays.

• Successfully reconstructed invariant mass of t and tbar from hadronic and leptonic decays.

• For the jet reconstruction, what must be done is the following:
  • In the _cfg.py file: use a module that converts HepMC to GenParticles
  • In the _cfg.py file: use a module that makes GenJets.
  • In the ValidationAnalyzer file: get the GenJets with something like: edm::Handle<GenJetCollection>. Match those jets with the EXTRA partons - that means, if we are dealing with ttbar+2partons, you want to match those jets to the 2 partons, NOT to the ttbar or the b.
  • By "match" I really mean "the deltaR in between the parton and the jet is smaller than some number" - at least the size of the jet (usually 0.5), but preferably smaller.

• The steps taken:
  • In the _cfg.py file: Added some processes: (process.load)
    • RecoJets/Configuration/RecoGenJetsAll_cff
    • RecoJets/Configuration/GenJetParticles_cff
    • RecoJets/JetProducers/sisCone5GenJets_cff
  • In the process . p = cms . Path:
    • process.genParticles * process.genParticlesForJets * process.sisCone5GenJets * process.demo
  • In the BuildFile:
    • use name=DataFormats/JetReco
  • Includes in the .cc code:
    • DataFormats/JetReco/interface/GenJet.h
    • DataFormats/JetReco/interface/GenJetCollection.h

-- ThiagoTomei - 28 Jan 2009 -- FlaviaDias - 17 Feb 2009