b-tag subgroup 

ATLAS Physics Analysis Retreat (Mar/07 SLAC)

Basic Startup Projects:

• Is a b-jet a b-jet ?

  Examine tag and fake rates given truth as a function of pT, eta, and any other variable which tag performance may potentially depend on.  How are the charm jet doing ?  

• What are vertex reconstruction resolutions like for the various taggers ?

  In particular, 2D vs 3D. How well is the primary vertex reconstructed ?

• Track impact parameter resolution 
  
  At an even more fundamental level, the track impact parameter resolution for various momenta and eta can provide a very basic feel for what can be expected for b-tagging performance.

• Jet flavor truth labelling

What flavor one labels a jet is tricky in MC (is it a b, c, light, or tau?). ATLAS has an algorithm with a very small delta-R cut between the jet axis and the meson. Is this appropriate? This involves looking at MC truth, tag rates, as a function of delta-R, etc.

Documentations:

Flavor Tagging Working Group Wiki
Official b-tagging (R12) Tutorial
    the SLAC copy of samples for the release 12 tutorial are located at $ATLROOT/work/data1/btag/tutorial/ 
b-tagging Session at Oct/06 Trigger&Physics Week

Data Samples

• Standard analysis using ATHENA

  For the current release 12 tagging code, the only samples (ESD and AOD) available are those for the b-tag tutorial at $ATLROOT/work/data1/btag/tutorial/. 

  The CSC11 AOD samples can be found from the SLAC dataset page:

  • b signal samples from TTbar, WH(H->bb) and ttH AOD samples.

  • c sample in WH(H->cc) AOD sample.

  • light flavors in QCD dijets J2,J3,J4 samples and WH(H->uu) sample.

    Note that these CSC11 samples have incompatible EDM  with the latest release 12 code. 
 

• Using the SAN to do tagging (as opposed to the AOD)