$t\bar{t}$ sample validation

We compare the events produced by different Monte-Carlo generators in order to identify and characterize the main systematic errors relevant for the measurement of $V_{tb}$. Besides our Madgraph All b samples we make also use of the following signal samples:

Sample Cross section (pb)
/tt0j_mT_70-alpgen/CMSSW_1_5_2-CSA07-2231/GEN-SIM-DIGI-RECO 334.51
/tt1j_mT_70-alpgen/CMSSW_1_5_2-CSA07-2211/GEN-SIM-DIGI-RECO 90.23
/tt2j_mT_70-alpgen/CMSSW_1_5_2-CSA07-2232/GEN-SIM-DIGI-RECO 18.75
/tt3j_mT_70-alpgen/CMSSW_1_5_2-CSA07-2212/GEN-SIM-DIGI-RECO 3.16

The comparison is done between inclusive distributions which means that a luminosity weighted sum of the $t\bar{t}+Njets$ distributions is made for each different conditions (Alpgen, Alpgen-fastsim, Madgraph). For each sample the luminosity is computed as $L_{sample}=\frac{N_{ev}(reco)}{\sigma_{process}}$. For each variable the distributions is weighted by $L^{-1}$ and acquires the dimensions of a cross section (pb).

Top production kinematics

Cross sections

Below we compare the cross section yields, at the generator level, for our private Madgraph+PYTHIA 6 samples with other $t\bar{t}$ samples. For completion we show the cross section obtained when the jet matching is done using $p_{T}^{2}$ ordered showers. However in the plots shown in the subsequent paragraphs we shall make use of the samples obtained using jet matching with virtuality ($Q^{2}$) ordered showers.

Table 1.1- Cross section yields for $t\bar{t}$ dilepton channel at generator level in pb
Event type 0 j 1 j 2 j 3 j 4 j Total K-factor
Madgraph + PYTHIA 6 $Q^{2}$ ordered showers w/ $Q_{cut}=30 GeV/c$ 11.00 7.97 3.53 2.10 n.a. 24.6 1.54 $\pm$ 0.03
Madgraph + PYTHIA 6 $p_{T}^{2}$ ordered showers w/ $Q_{cut}=20 GeV/c$ 4.43 3.97 2.54 2.92 n.a. 13.86 2.73 $\pm$ 0.05
ALPGEN 15.2 4.1 0.85 0.14 0.04 20.34 1.86 $\pm$ 0.03

Note that despite the large differences in the cross sections for Madgraph+PYTHIA6 obtained with different methods of jet matching the inclusive kinematics distributions for $t\bar{t}+Njets$ production is not affected, as it should. We also not that, depending on the parameters chosen for matching, the contribution of each jet multiplicity to the total cross section will be different. For more details see the sample validation report from MatchChecker utility. Questions:

  • why does the cross section decrease so much when $p_{T}^{2}$ ordered showers are used?
  • how does the jet matching procedure efficiency affect the cross section?


We proceed to compare the top production kinematics. You can find the basic plots in the following table. We have normalized each distribution by the luminosity of the sample used to build it and, as so, the counts are shown in units of cross section (pb).

Table 2.1 - Top kinematics at MC level (GEN)
$p_{T}$ spectrum $\eta$ distribution $\sum p_{T}$ for the $t\bar{t}$ pair $\Delta\phi$ between $t$ and $\bar{t}$ invariant mass of the $t\bar{t}$ pair



Table 3.1 - Jet kinematics at RECO level
Variable RECO RECO + matched to b-parton from top decay
Jet multiplicity n.a.
emf (electromagnetic fraction of the energy deposit of the jets)
$\Delta_{2}$ (b-tag discriminant for the high efficiency track counting algorithm

There is an overall good agreemenent between jet multiplicities generated by Alpgen and Madgraph. The main differences are seen in the electromagnetic fraction distribution and are most likely due to FastSim.


Table 4.1 - Lepton kinematics at RECO level for the tt~0j samples
Lepton Electron Muon
Variable RECO RECO + matched RECO RECO + matched
Lepton multiplicity n.a. n.a.
$\frac{E}{p}$ n.a. n.a.
tracker iso
calo iso

From this plots we conclude that there's a good agreement between the samples. Some comments:

  • there are some differences in lepton multiplicities
  • the low $\p_{T}$ electronic spectrum is understimated by FastSim
  • for the $\eta$ distribution of electrons we can clearly see similar drops in FullSim and FastSim in the ECAL barrel to endcap transition ($1.479\leq\eta\leq1.653$)
  • there's an underestimation of the the low E/p values for electrons in FastSim


Below we show the comparison of the MET. The main conclusion is that FastSim yields a softer spectrum than full simulation but is in good agreement in the region $30 < E_{T} <100 GeV$.

met.png |

-- PedroSilva - 04 Apr 2008
Latex rendering error!! dvi file was not created.

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng e_EoverP_comp.png r4 r3 r2 r1 manage 16.4 K 2008-04-28 - 11:52 PedroSilva  
PNGpng e_caloIso_comp.png r3 r2 r1 manage 16.3 K 2008-04-28 - 11:52 PedroSilva  
PNGpng e_eta_comp.png r4 r3 r2 r1 manage 14.0 K 2008-04-28 - 11:53 PedroSilva  
PNGpng e_matched_EoverP_comp.png r4 r3 r2 r1 manage 16.7 K 2008-04-28 - 11:53 PedroSilva  
PNGpng e_matched_caloIso_comp.png r3 r2 r1 manage 18.1 K 2008-04-28 - 11:53 PedroSilva  
PNGpng e_matched_eta_comp.png r4 r3 r2 r1 manage 14.5 K 2008-04-28 - 11:54 PedroSilva  
PNGpng e_matched_pT_comp.png r4 r3 r2 r1 manage 16.1 K 2008-04-28 - 11:55 PedroSilva  
PNGpng e_matched_trackerIso_comp.png r4 r3 r2 r1 manage 18.3 K 2008-04-28 - 11:54 PedroSilva  
PNGpng e_multiplicity_comp.png r4 r3 r2 r1 manage 12.1 K 2008-04-28 - 11:55 PedroSilva  
PNGpng e_pT_comp.png r4 r3 r2 r1 manage 16.7 K 2008-04-28 - 11:56 PedroSilva  
PNGpng e_trackerIso_comp.png r4 r3 r2 r1 manage 17.6 K 2008-04-28 - 11:56 PedroSilva  
PNGpng jets_emf_comp.png r5 r4 r3 r2 r1 manage 14.6 K 2008-04-28 - 11:24 PedroSilva  
PNGpng jets_eta_comp.png r6 r5 r4 r3 r2 manage 12.7 K 2008-04-28 - 11:24 PedroSilva  
PNGpng jets_matched_emf_comp.png r5 r4 r3 r2 r1 manage 15.3 K 2008-04-28 - 11:24 PedroSilva  
PNGpng jets_matched_eta_comp.png r5 r4 r3 r2 r1 manage 12.6 K 2008-04-28 - 11:25 PedroSilva  
PNGpng jets_matched_pT_comp.png r5 r4 r3 r2 r1 manage 16.8 K 2008-04-28 - 11:25 PedroSilva  
PNGpng jets_matched_trkCountHighEff_comp.png r5 r4 r3 r2 r1 manage 14.9 K 2008-04-28 - 11:26 PedroSilva  
PNGpng jets_multiplicity_comp.png r5 r4 r3 r2 r1 manage 14.8 K 2008-04-28 - 11:26 PedroSilva  
PNGpng jets_pT_comp.png r5 r4 r3 r2 r1 manage 16.9 K 2008-04-28 - 11:28 PedroSilva  
PNGpng jets_trkCountHighEff_comp.png r5 r4 r3 r2 r1 manage 14.8 K 2008-04-28 - 11:28 PedroSilva  
PNGpng met_comp.png r4 r3 r2 r1 manage 16.9 K 2008-04-28 - 11:42 PedroSilva  
PNGpng mu_EoverP_comp.png r2 r1 manage 13.6 K 2008-04-28 - 12:07 PedroSilva  
PNGpng mu_caloIso_comp.png r3 r2 r1 manage 15.6 K 2008-04-28 - 12:09 PedroSilva  
PNGpng mu_eta_comp.png r3 r2 r1 manage 14.3 K 2008-04-28 - 12:08 PedroSilva  
PNGpng mu_matched_EoverP_comp.png r2 r1 manage 13.4 K 2008-04-28 - 12:08 PedroSilva  
PNGpng mu_matched_caloIso_comp.png r3 r2 r1 manage 16.8 K 2008-04-28 - 12:09 PedroSilva  
PNGpng mu_matched_eta_comp.png r3 r2 r1 manage 14.2 K 2008-04-28 - 12:09 PedroSilva  
PNGpng mu_matched_pT_comp.png r3 r2 r1 manage 16.9 K 2008-04-28 - 12:10 PedroSilva  
PNGpng mu_matched_trackerIso_comp.png r3 r2 r1 manage 18.2 K 2008-04-28 - 12:11 PedroSilva  
PNGpng mu_multiplicity_comp.png r3 r2 r1 manage 11.8 K 2008-04-28 - 12:11 PedroSilva  
PNGpng mu_pT_comp.png r3 r2 r1 manage 16.9 K 2008-04-28 - 12:10 PedroSilva  
PNGpng mu_trackerIso_comp.png r3 r2 r1 manage 17.6 K 2008-04-28 - 12:10 PedroSilva  
PNGpng ttbar_dphi_comp.png r4 r3 r2 r1 manage 16.7 K 2008-04-28 - 11:37 PedroSilva  
PNGpng ttbar_eta_comp.png r4 r3 r2 r1 manage 15.8 K 2008-04-28 - 11:37 PedroSilva  
PNGpng ttbar_minv_comp.png r3 r2 r1 manage 17.7 K 2008-04-28 - 11:38 PedroSilva  
PNGpng ttbar_pT_comp.png r4 r3 r2 r1 manage 17.8 K 2008-04-28 - 11:38 PedroSilva  
PNGpng ttbar_sumpT_comp.png r4 r3 r2 r1 manage 19.8 K 2008-04-28 - 11:38 PedroSilva  
Edit | Attach | Watch | Print version | History: r13 < r12 < r11 < r10 < r9 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r13 - 2008-06-26 - PedroSilva
    • 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 & 2008-2023 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback