Delphes analysis of top quark polarization using Λb decays


Private page for internal exchange of information.

Put here: all practical information about MC samples; instructions to run Delphes and DelphesAnalysis; python code; small files; maybe plots.

MC samples

It is convenient to produce MC files in STDHEP format instead of LHE because STDHEP contains all hadrons and LHE doesn't.

Samples with 0,1,2 Λb decaying to muon and Λc in events of the kind

p p -> t tbar
       t    -> Wleptonic b
       tbar -> Whad b
are available here; username and password are in file webpage inside our svn repository.

The file needed to filter and to add the proper polarization (with EvtGen) is

Delphes set up

Use Delphes 3.1.2, which is the latest release. Instructions are here.

Set ROOT and python as the webpage instructs. In the specific case of this is the quickest way to get everything properly set up, exploiting the CMSSW settings:

cd [some CMSSW release]
setenv SCRAM_ARCH slc6_amd64_gcc481
cd [your Delphes-3.1.2 directory]
setenv PYTHONPATH `pwd`/python:$PYTHONPATH
In case you use bash instead of tc shell, just use export instead of setenv (with proper syntax).

Beware, at the moment you would have a compilation error on (fixed upon my request in the repository, will be therefore ok in future releases). To work around it: replace #include "Pythia.h" with #include "Pythia8/Pythia.h" in files modules/ and readers/DelphesPythia8.cpp.

Files I modified

To create a collection SoftMuons, took inspiration by this example:

To add impact parameter (and its uncertainty) to all the muons:

Configuration file:

DelphesAnalysis set up

Instructions for DelphesAnalysis at the bottom of this page are ok but:

  • don't execute . ./ at the beginning (you cannot anyway, as it was apparently removed from the release at some point; anyway, it was just setting PYTHONPATH and LD_LIBRARY_PATH as above)
  • chmod +x DelphesAnalysis/
  • as usual, the webpage assumes bash shell, so just translate export into setenv (e.g., setenv DelphesAnalysisCfg "") if you use tc shell
  • I didn't manage to run DelphesAnalysis/, but I am not sure that we need it anyway

Starting point could be the existing example ttbar analysis. To be modified such to consider:

  • a switch for events with and without Λb decaying into muon
  • event counting without b tagging
  • event reconstruction with two neutrinos

Note: maybe I will not use it in the end.

Root macro and scripts

This is what I am using right now (probably not the cleanest c++ code you have ever seen):

Pending issues

From Yevgeny:

  • an interesting thing I notice is that D0 have used my idea of the shortness of the Lambda_c lifetime for background reduction and they say that "it reduces the background significantly" (for the particular channel they study)
  • The thing is that for any channel the purity goes to zero in the limit of small fragmentation fraction times branching ratio. How do we know that we're not already in this regime for some (or all) of our exclusive channels?
  • Also, while on one hand the CDF and D0 measurements suffer from larger combinatoric background (as Andrea said), on the other hand our Lambda_b's have higher pT's so the resolution of the Lambda_c mass peak will be worse (is this correct?) if I ignore differences between the Tevatron and LHC detectors. Differences between detectors apparently don't help enough -- see my email from a few days ago about the Lambda_c mass window in Mario's simulation (which presumably approximates CMS) vs. CDF, unless Mario used a larger mass window than is actually needed.
  • I have another candidate for a very important but reducible background: Ds+ -> K+ K- pi+. Naively it's less important than the decay from the previous email because the Ds+ fragmentation fraction is smaller than that of D+ by a factor of 2.4, and furthermore the BR of this decay (5.4%) is smaller than that of the D+ decay by a factor of 1.7. But the reason it may still win is that the decay D+ -> pi+ K- pi+ will very often have the K- as the leading rather than the intermediate particle, so it won't pass the momentum-hierarchy cut, while Ds+ will frequently have the K- intermediate in momentum. This background can possibly be eliminated by requiring the invariant mass of three tracks to be inconsistent with the Ds+ mass for the K+ K- pi+ hypothesis. This would still leave us with contributions from K+ K- pi+ pi0, which has a comparably large BR (6.3%).

-- AndreaGiammanco - 07 Aug 2014

Topic attachments
I Attachment History Action Size Date Who Comment
Unknown file formatcc r1 manage 5.3 K 2014-08-18 - 11:52 AndreaGiammanco  
Header fileh AntiIsolation.h r1 manage 1.3 K 2014-08-18 - 11:53 AndreaGiammanco  
Unknown file formatcc r1 manage 3.4 K 2014-08-13 - 11:31 AndreaGiammanco  
Header fileh Delphes.h r1 manage 0.8 K 2014-08-13 - 11:31 AndreaGiammanco  
Header fileh DelphesClasses.h r1 manage 13.5 K 2014-08-18 - 17:13 AndreaGiammanco  
C source code filecpp DelphesSTDHEP.cpp r2 r1 manage 6.0 K 2014-08-13 - 11:29 AndreaGiammanco  
Unknown file formatcc r2 r1 manage 13.3 K 2014-08-13 - 11:30 AndreaGiammanco  
Header fileh DelphesSTDHEPReader.h r2 r1 manage 1.8 K 2014-08-13 - 11:30 AndreaGiammanco  
PNGpng DeltaEta_corrjet_lambdab.png r3 r2 r1 manage 10.4 K 2014-09-30 - 16:16 AndreaGiammanco  
PNGpng DeltaEta_jet_lambdab.png r3 r2 r1 manage 10.0 K 2014-09-30 - 16:17 AndreaGiammanco  
PNGpng DeltaPhi_corrjet_lambdab.png r3 r2 r1 manage 10.7 K 2014-09-30 - 16:17 AndreaGiammanco  
PNGpng DeltaPhi_jet_lambdab.png r3 r2 r1 manage 11.0 K 2014-09-30 - 16:17 AndreaGiammanco  
PNGpng E_ratio_corrjet_lambdab.png r3 r2 r1 manage 12.8 K 2014-09-30 - 16:18 AndreaGiammanco  
PNGpng E_ratio_jet_bquark.png r2 r1 manage 10.2 K 2014-09-30 - 16:18 AndreaGiammanco  
PNGpng E_ratio_jet_lambdab.png r3 r2 r1 manage 13.1 K 2014-09-30 - 16:18 AndreaGiammanco  
PNGpng E_ratio_lambdab_bquark.png r2 r1 manage 16.8 K 2014-08-28 - 19:32 AndreaGiammanco  
PNGpng E_ratio_lambdab_bquark_selected.png r1 manage 17.0 K 2014-08-28 - 19:32 AndreaGiammanco  
Header fileh ModulesLinkDef.h r1 manage 2.6 K 2014-08-18 - 11:53 AndreaGiammanco  
Unknown file formatcc r1 manage 18.9 K 2014-08-18 - 17:13 AndreaGiammanco  
C source code filec analyzeLambdab.C r7 r6 r5 r4 r3 manage 35.4 K 2014-09-03 - 22:17 AndreaGiammanco  
PNGpng cut_flow_0L.png r1 manage 11.0 K 2014-09-30 - 14:38 AndreaGiammanco  
PNGpng cut_flow_1L.png r2 r1 manage 10.9 K 2014-09-30 - 12:17 AndreaGiammanco  
PNGpng cut_flow_2L.png r2 r1 manage 10.6 K 2014-09-30 - 12:17 AndreaGiammanco  
Unknown file formattcl delphes_card_CMS.tcl r4 r3 r2 r1 manage 20.0 K 2014-08-22 - 07:22 AndreaGiammanco  
Unknown file formatcc r1 manage 8.8 K 2014-08-15 - 16:09 AndreaGiammanco  
Edit | Attach | Watch | Print version | History: r25 < r24 < r23 < r22 < r21 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r25 - 2015-04-24 - AndreaGiammanco
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback