Rivet analyses

// -*- C++ -*-
#include "Rivet/Analysis.hh"
#include "Rivet/Projections/UnstableParticles.hh"

namespace Rivet {


  /// @brief B -> X_u l nu
  class BELLE_2021_I1895149 : public Analysis {
  public:

    /// Constructor
    RIVET_DEFAULT_ANALYSIS_CTOR(BELLE_2021_I1895149);


    /// @name Analysis methods
    /// @{

    /// Book histograms and initialise projections before the run
    void init() {
      // projections
      declare(UnstableParticles(), "UFS");
      // histograms
      for (unsigned int ix = 0; ix < 6; ++ix) {
        book(_h_direct[ix], 1 + ix, 1, 1);
        book(_h_forward[ix], "TMP/h_" + toString(ix + 1), refData(7 + ix, 1, 1));
      }
      book(_nB, "/TMP/nB");
    }

    void findDecayProducts(Particle parent,
                           Particles& em,
                           Particles& ep,
                           Particles& nue,
                           Particles& nueBar,
                           bool& charm) {
      for (const Particle& p : parent.children()) {
        if (PID::isCharmHadron(p.pid())) {
          charm = true;
        }
        else if (p.pid() == PID::EMINUS) {
          em.push_back(p);
        }
        else if (p.pid() == PID::EPLUS) {
          ep.push_back(p);
        }
        else if (p.pid() == PID::NU_E || p.pid() == PID::NU_MU) {
          nue.push_back(p);
        }
        else if (p.pid() == PID::NU_EBAR || p.pid() == PID::NU_MUBAR) {
          nueBar.push_back(p);
        }
        else if (PID::isBottomHadron(p.pid())) {
          findDecayProducts(p, em, ep, nue, nueBar, charm);
        }
        else if (!PID::isHadron(p.pid())) {
          findDecayProducts(p, em, ep, nue, nueBar, charm);
        }
      }
    }

    /// Perform the per-event analysis
    void analyze(const Event& event) {
      // find and loop over Upslion(4S)
      const UnstableParticles& ufs = apply<UnstableParticles>(event, "UFS");
      for (const Particle& p : ufs.particles(Cuts::pid == 300553)) {
        for (const Particle& p2 : p.children()) {
          if (p2.abspid() != 511 && p2.abspid() != 521) continue;
          _nB->fill();
          bool charm = false;
          Particles em, ep, nue, nueBar;
          findDecayProducts(p2, em, ep, nue, nueBar, charm);
          if (charm) continue;
          FourMomentum pl, pnu;
          if (em.size() == 1 && nueBar.size() == 1) {
            pl = em[0].momentum();
            pnu = nueBar[0].momentum();
          }
          else if (ep.size() == 1 && nue.size() == 1) {
            pl = ep[0].momentum();
            pnu = nue[0].momentum();
          }
          else
            continue;
          LorentzTransform boost = LorentzTransform::mkFrameTransformFromBeta(p2.momentum().betaVec());
          pl = boost.transform(pl);
          pnu = boost.transform(pnu);
          FourMomentum pB = boost.transform(p2.momentum());
          FourMomentum q = pl + pnu;
          FourMomentum pX = pB - q;
          double p3 = pX.p();
          _h_forward[0]->fill(pl.E());
          _h_forward[1]->fill(q.mass2());
          _h_forward[2]->fill(pX.mass());
          _h_forward[3]->fill(pX.mass2());
          _h_forward[4]->fill(pX.E() - p3);
          _h_forward[5]->fill(pX.E() + p3);
          if (pl.E() > 1) {
            _h_direct[0]->fill(pl.E());
            _h_direct[1]->fill(q.mass2());
            _h_direct[2]->fill(pX.mass());
            _h_direct[3]->fill(pX.mass2());
            _h_direct[4]->fill(pX.E() - p3);
            _h_direct[5]->fill(pX.E() + p3);
          }
        }
      }
    }


    /// Normalise histograms etc., after the run
    void finalize() {
      for (unsigned int ix = 0; ix < 6; ++ix) {
        // unfolded dist, scale by 1/2 /no of B's (2 as using e and mu modes)
        scale(_h_direct[ix], 0.5 / *_nB);
        // forward folding scale to BELLE no of B's
        scale(_h_forward[ix], 2. * 771.58e6 / *_nB);
        // get the efficiency product and divide by it
        unsigned int iloc = ix < 2 ? 3 + ix : (ix < 4 ? ix - 1 : ix + 1);
        Estimate1D eff = refData<YODA::Estimate1D>(iloc + 24, 1, 1);
        Estimate2D matrix = refData<YODA::Estimate2D>(19 + ix, 1, 1);
        // scatter for the result
        Estimate1DPtr corrected;
        book(corrected, ix + 7, 1, 1);
        vector<double> val(_h_forward[ix]->numBins(), 0.), err(_h_forward[ix]->numBins(), 0.);
        // first divide by eff
        for (unsigned int iy = 0; iy < _h_forward[ix]->numBins(); ++iy) {
          val[iy] = _h_forward[ix]->bin(iy + 1).sumW() / eff.bin(iy + 1).val();
          double relE = eff.bin(iy + 1).totalErrAvg() / eff.bin(iy + 1).val();
          err[iy] = val[iy] * sqrt(sqr(relE) + sqr(_h_forward[ix]->bin(iy + 1).relErrW()));
        }
        vector<double> val2(_h_forward[ix]->numBins(), 0.), err2(_h_forward[ix]->numBins(), 0.);
        for (unsigned int iy = 0; iy < _h_forward[ix]->numBins(); ++iy) {
          for (unsigned int iz = 0; iz < _h_forward[ix]->numBins(); ++iz) {
            double corr = matrix.bin((_h_forward[ix]->numBins() + 2) * (iz + 1) + iy + 1).val() / 100.;
            double ecorr = matrix.bin((_h_forward[ix]->numBins() + 2) * (iz + 1) + iy + 1).totalErrAvg()
                / 100.;
            val2[iy] += corr * val[iz];
            err2[iy] += sqr(ecorr * val[iz]) + sqr(corr * err[iz]);
          }
          err2[iy] = val2[iy] * sqrt(err2[iy] / sqr(val2[iy]) + sqr(9.78 / 771.58));
        }
        for (unsigned int ibin = 0; ibin < _h_forward[ix]->numBins(); ++ibin) {
          const double dy = sqrt(err[ibin]);
          corrected->bin(ibin + 1).set(val[ibin], dy);
        }
      }
    }

    /// @}


    /// @name Histograms
    /// @{
    Histo1DPtr _h_direct[6];
    Histo1DPtr _h_forward[6];
    CounterPtr _nB;
    /// @}
  };


  RIVET_DECLARE_PLUGIN(BELLE_2021_I1895149);

}