Rivet analyses

Measurements of ttbar charge asymmetry using dilepton final states in pp collisions at sqrt(s) = 8 TeV

Experiment: CMS (LHC)

Inspire ID: 1430892

Status: VALIDATED

Authors: - Jacob Linacre

References: - Phys. Lett. B 760 (2016) 365 - DOI:10.1016/j.physletb.2016.07.006 - arXiv: 1603.06221 - https://hepdata.net/record/ins1430892

Beams: p+ p+

Beam energies: (4000.0, 4000.0)GeV

Run details: - Dilepton ttbar events at $\sqrt{s}=8~\TeV$, where the leptons are prompt elecrons or muons (not from tau). No other cuts. All but one of the variables require top quarks in the event record.

Abstract: The charge asymmetry in tt̄ events is measured using dilepton final states produced in pp collisions at the LHC at $\sqrt{s}=8\:$TeV. The data sample, collected with the CMS detector, corresponds to an integrated luminosity of 19.5 fb−1. The measurements are performed using events with two oppositely charged leptons (electrons or muons) and two or more jets, where at least one of the jets is identified as originating from a bottom quark. The charge asymmetry is measured from differences in kinematic distributions, unfolded to the parton level, of positively and negatively charged top quarks and leptons. The tt̄ and leptonic charge asymmetries are found to be 0.011 +/- 0.011 (stat) +/- 0.007 (syst) and 0.003 +/- 0.006 (stat) +/- 0.003 (syst), respectively. These results, as well as charge asymmetry measurements made as a function of tt̄ system kinematic properties, are in agreement with predictions of the standard model. Particle-level addition to Rivet routine: While the analysis was performed at the parton-level only, Δ|η| is a purely leptonic variable and it has been checked that the results of the analysis would have been essentially unchanged had it been defined at particle-level using dressed leptons instead of using the parton-level top quark daughter leptons. We therefore include both particle- and parton-level versions of this distribution in the Rivet routine, with the former identified in the plot title. For same-flavour dilepton final states, the particle-level definition in the full phase space is problematic because the two leptons can come from fully-hadronic tt̄ plus a dilepton pair from radiation. Such pairs have invariant mass M ∼ 0 and produce a peak near zero in the Δ|η| distribution. We therefore select only the tt̄ → eμ final state, by requiring exactly one electron and exactly one muon. Note this means tt̄ → eμ events with additional dilepton pairs from radiation are vetoed. For PYTHIA8 this amounts to ~0.5% of tt̄ → eμ events - well below the level of sensitivity of the measured distribution. Histograms and covariance matrices: The error bars in the measured distributions should not be used for fitting because there are significant correlations between bins. The covariance matrices for the statistical and systematic uncertainties in each distribution can be found in hepdata. The single-differential cross sections in hepdata are normalised to unit area (i.e. the integral is equal to one), while the double-differential cross sections in hepdata are normalised to the sum of entries (such that the sum of all bin heights is equal to one). This should be taken into account when comparing the measured distributions to the Rivet results and when using the covariance matrices. Underflow and overflow bins: The lower and upper Δ|yt| and Δ|η| bins (starting and ending at -2 and +2, respectively) contain underflow and overflow events, i.e. the complete distribution from -infinity to +infinity is covered. Similarly, the upper Mtt̄, pTtt̄, and |ytt̄| bins contain overflow events up to +infinity.

Source code:CMS_2016_I1430892.cc

// -*- C++ -*-
#include "Rivet/Analysis.hh"
#include "Rivet/Projections/FinalState.hh"
#include "Rivet/Projections/IdentifiedFinalState.hh"
#include "Rivet/Projections/PromptFinalState.hh"
#include "Rivet/Projections/LeptonFinder.hh"
#include "Rivet/Projections/PartonicTops.hh"

namespace Rivet {


  /// CMS 8 TeV dilepton channel ttbar charge asymmetry analysis
  class CMS_2016_I1430892 : public Analysis {
  public:

    /// Constructor
    RIVET_DEFAULT_ANALYSIS_CTOR(CMS_2016_I1430892);


    /// Book histograms and initialise projections
    void init() {

      // Complete final state
      FinalState fs;

      // Projection for dressed electrons and muons
      IdentifiedFinalState photons(fs);
      photons.acceptIdPair(PID::PHOTON);

      IdentifiedFinalState el_id(fs);
      el_id.acceptIdPair(PID::ELECTRON);
      PromptFinalState electrons(el_id);
      declare(electrons, "Electrons");
      LeptonFinder dressed_electrons(electrons, photons, 0.1);
      declare(dressed_electrons, "DressedElectrons");

      IdentifiedFinalState mu_id(fs);
      mu_id.acceptIdPair(PID::MUON);
      PromptFinalState muons(mu_id);
      declare(muons, "Muons");
      LeptonFinder dressed_muons(muons, photons, 0.1);
      declare(dressed_muons, "DressedMuons");

      // Parton-level top quarks
      declare(PartonicTops(TopDecay::E_MU, PromptEMuFromTau::NO), "LeptonicPartonTops");


      // Booking of histograms

      // This histogram is independent of the parton-level information, and is an
      // addition to the original analysis. It is compared to the same data as
      // the parton-level delta_abseta histogram d05-x01-y01.
      book(_h_dabsetadressedleptons, "d00-x01-y01", _bins_dabseta);

      // The remaining histos use parton-level information
      book(_h_dabseta, "d05-x01-y01", _bins_dabseta);
      book(_h_dabsrapidity, "d02-x01-y01", _bins_dabsrapidity);

      // 2D histos
      book(_h_dabsrapidity_var[0], "d11-x01-y01", _bins_dabsrapidity, _bins_tt_mass);
      book(_h_dabseta_var[0], "d17-x01-y01", _bins_dabseta, _bins_tt_mass);

      book(_h_dabsrapidity_var[1], "d23-x01-y01", _bins_dabsrapidity, _bins_tt_pT);
      book(_h_dabseta_var[1], "d29-x01-y01", _bins_dabseta, _bins_tt_pT);

      book(_h_dabsrapidity_var[2], "d35-x01-y01", _bins_dabsrapidity, _bins_tt_absrapidity);
      book(_h_dabseta_var[2], "d41-x01-y01", _bins_dabseta, _bins_tt_absrapidity);

      // Profile histos for asymmetries
      book(_h_dabsrapidity_profile[0], "d08-x01-y01", _bins_tt_mass);
      book(_h_dabseta_profile[0], "d14-x01-y01", _bins_tt_mass);

      book(_h_dabsrapidity_profile[1], "d20-x01-y01", _bins_tt_pT);
      book(_h_dabseta_profile[1], "d26-x01-y01", _bins_tt_pT);

      book(_h_dabsrapidity_profile[2], "d32-x01-y01", _bins_tt_absrapidity);
      book(_h_dabseta_profile[2], "d38-x01-y01", _bins_tt_absrapidity);

    }


    /// Perform the per-event analysis
    void analyze(const Event& event) {


      // Use particle-level leptons for the first histogram
      const LeptonFinder& dressed_electrons = apply<LeptonFinder>(event, "DressedElectrons");
      const LeptonFinder& dressed_muons = apply<LeptonFinder>(event, "DressedMuons");

      const DressedLeptons dressedels = dressed_electrons.dressedLeptons();
      const DressedLeptons dressedmus = dressed_muons.dressedLeptons();

      const size_t ndressedel = dressedels.size();
      const size_t ndressedmu = dressedmus.size();

      // For the particle-level histogram, require exactly one electron and exactly one muon, to select the ttbar->emu channel.
      // Note this means ttbar->emu events with additional PromptFinalState dilepton pairs from the shower are vetoed - for PYTHIA8,
      // this affects ~0.5% of events, so the effect is well below the level of sensitivity of the measured distribution.
      if ( ndressedel == 1 && ndressedmu == 1 ) {

        const int electrontouse = 0, muontouse = 0;

        // Opposite-charge leptons only
        if ( sameSign(dressedels[electrontouse], dressedmus[muontouse]) ) {
          MSG_INFO("Error, e and mu have same charge, skipping event");
        } else {
          // Get the four-momenta of the positively- and negatively-charged leptons
          FourMomentum lepPlus = dressedels[electrontouse].charge() > 0 ? dressedels[electrontouse] : dressedmus[muontouse];
          FourMomentum lepMinus = dressedels[electrontouse].charge() > 0 ? dressedmus[muontouse] : dressedels[electrontouse];

          // Now calculate the variable
          double dabseta_temp = lepPlus.abseta() - lepMinus.abseta();

          fillWithUFOF( _h_dabsetadressedleptons, dabseta_temp);
        }

      }


      // The remaining variables use parton-level information.

      // Get the leptonically decaying tops
      const Particles leptonicpartontops = apply<ParticleFinder>(event, "LeptonicPartonTops").particlesByPt();
      Particles chargedleptons;

      unsigned int ntrueleptonictops = 0;
      bool oppositesign = false;

      if ( leptonicpartontops.size() == 2 ) {
        for (size_t k = 0; k < leptonicpartontops.size(); ++k) {

          // Get the lepton
          const Particle lepTop = leptonicpartontops[k];
          const auto isPromptChargedLepton = [](const Particle& p){return (isChargedLepton(p) && isPrompt(p, false, false));};
          Particles lepton_candidates = lepTop.allDescendants(firstParticleWith(isPromptChargedLepton), false);
          if ( lepton_candidates.size() < 1 ) MSG_WARNING("error, TopDecay::E_MU top quark had no daughter lepton candidate, skipping event.");

          // In some cases there is no lepton from the W decay but only leptons from the decay of a radiated gamma.
          // These hadronic PartonicTops are currently being mistakenly selected by TopDecay::E_MU (as of April 2017), and need to be rejected.
          // TopDecay::E_MU is being fixed in Rivet, and when it is the veto below should do nothing.
          /// @todo Should no longer be necessary -- remove
          bool istrueleptonictop = false;
          for (size_t i = 0; i < lepton_candidates.size(); ++i) {
            const Particle& lepton_candidate = lepton_candidates[i];
            if ( lepton_candidate.hasParentWith(Cuts::pid == PID::PHOTON) ) {
              MSG_DEBUG("Found gamma parent, top: " << k+1 << " of " << leptonicpartontops.size() << " , lepton: " << i+1 << " of " << lepton_candidates.size());
              continue;
            }
            if ( !istrueleptonictop && sameSign(lepTop,lepton_candidate) ) {
              chargedleptons.push_back(lepton_candidate);
              istrueleptonictop = true;
            }
            else MSG_WARNING("Error, found extra prompt charged lepton from top decay (and without gamma parent), ignoring it.");
          }
          if ( istrueleptonictop ) ++ntrueleptonictops;
        }
      }

      if ( ntrueleptonictops == 2 ) {
        oppositesign = !( sameSign(chargedleptons[0],chargedleptons[1]) );
        if ( !oppositesign ) MSG_WARNING("error, same charge tops, skipping event.");
      }

      if ( ntrueleptonictops == 2 && oppositesign ) {

        // Get the four-momenta of the positively- and negatively-charged leptons
        const FourMomentum lepPlus = chargedleptons[0].charge() > 0 ? chargedleptons[0] : chargedleptons[1];
        const FourMomentum lepMinus = chargedleptons[0].charge() > 0 ? chargedleptons[1] : chargedleptons[0];

        const double dabseta_temp = lepPlus.abseta() - lepMinus.abseta();

        // Get the four-momenta of the positively- and negatively-charged tops
        const FourMomentum topPlus_p4 = leptonicpartontops[0].pid() > 0 ? leptonicpartontops[0] : leptonicpartontops[1];
        const FourMomentum topMinus_p4 = leptonicpartontops[0].pid() > 0 ? leptonicpartontops[1] : leptonicpartontops[0];

        const FourMomentum ttbar_p4 = topPlus_p4 + topMinus_p4;

        const double tt_mass_temp = ttbar_p4.mass();
        const double tt_absrapidity_temp = ttbar_p4.absrapidity();
        const double tt_pT_temp = ttbar_p4.pT();
        const double dabsrapidity_temp = topPlus_p4.absrapidity() - topMinus_p4.absrapidity();

        // Fill parton-level histos
        fillWithUFOF( _h_dabseta, dabseta_temp);
        fillWithUFOF( _h_dabsrapidity, dabsrapidity_temp);

        // Now fill the same variables in the 2D and profile histos vs ttbar invariant mass, pT, and absolute rapidity
        for (int i_var = 0; i_var < 3; ++i_var) {
          double var;
          if ( i_var == 0 ) {
            var = tt_mass_temp;
          } else if ( i_var == 1 ) {
            var = tt_pT_temp;
          } else {
            var = tt_absrapidity_temp;
          }

          fillWithUFOF( _h_dabsrapidity_var[i_var], dabsrapidity_temp, var);
          fillWithUFOF( _h_dabseta_var[i_var], dabseta_temp, var);

          fillWithUFOF( _h_dabsrapidity_profile[i_var], dabsrapidity_temp, var, (_h_dabsrapidity->xMax() + _h_dabsrapidity->xMin())/2. );
          fillWithUFOF( _h_dabseta_profile[i_var], dabseta_temp, var, (_h_dabseta->xMax() + _h_dabseta->xMin())/2. );
        }

      }

    }


    /// Normalise histograms to unit area
    void finalize() {

      normalize(_h_dabsetadressedleptons);

      normalize(_h_dabseta);
      normalize(_h_dabsrapidity);

      for (int i_var = 0; i_var < 3; ++i_var) {
        normalize(_h_dabsrapidity_var[i_var]);
        normalize(_h_dabseta_var[i_var]);
      }

    }


  private:

    Histo1DPtr _h_dabsetadressedleptons, _h_dabseta, _h_dabsrapidity;
    Histo2DPtr _h_dabseta_var[3], _h_dabsrapidity_var[3];
    Profile1DPtr _h_dabseta_profile[3], _h_dabsrapidity_profile[3];

    const vector<double> _bins_tt_mass = {300., 430., 530., 1200.};
    const vector<double> _bins_tt_pT = {0., 41., 92., 300.};
    const vector<double> _bins_tt_absrapidity = {0., 0.34, 0.75, 1.5};
    const vector<double> _bins_dabseta = { -2., -68./60., -48./60., -32./60., -20./60., -8./60., 0., 8./60., 20./60., 32./60., 48./60., 68./60., 2.};
    const vector<double> _bins_dabsrapidity = {-2., -44./60., -20./60., 0., 20./60., 44./60., 2.};

    void fillWithUFOF(Histo1DPtr h, double x) {
      h->fill(std::max(std::min(x, h->xMax()-1e-9),h->xMin()+1e-9));
    }

    void fillWithUFOF(Histo2DPtr h, double x, double y) {
      h->fill(std::max(std::min(x, h->xMax()-1e-9),h->xMin()+1e-9), std::max(std::min(y, h->yMax()-1e-9),h->yMin()+1e-9));
    }

    void fillWithUFOF(Profile1DPtr h, double x, double y, double c) {
      h->fill(std::max(std::min(y, h->xMax()-1e-9),h->xMin()+1e-9), float(x > c) - float(x < c));
    }


  };


  RIVET_DECLARE_PLUGIN(CMS_2016_I1430892);


}