RooCMSShape.cc

Go to the documentation of this file.

/*****************************************************************************
 * Project: CMS detector at the CERN
 *
 * Package: PhysicsTools/TagAndProbe/RooCMSShape
 *
 *
 * Authors:
 *   Nadia Adam, Princeton - neadam@princeton.edu
 *   Adam Hunt, Princeton  - ahunt@princeton.edu
 *   Kalanand Mishra, Fermilab - kalanand@fnal.gov
 *
 * Description:
 *   Defines a probability density function which has exponential decay 
 *   distribution at high mass beyond the pole position (say, Z peak)  
 *   but turns over (i.e., error function) at low mass due to threshold 
 *   effect. We use this to model the background shape in Z->ll invariant 
 *   mass.
 * History:
 *   
 *
 *****************************************************************************/
 
#include "PhysicsTools/TagAndProbe/interface/RooCMSShape.h"
 
ClassImp(RooCMSShape);
 
RooCMSShape::RooCMSShape(const char* name,
                         const char* title,
                         RooAbsReal& _x,
                         RooAbsReal& _alpha,
                         RooAbsReal& _beta,
                         RooAbsReal& _gamma,
                         RooAbsReal& _peak)
    : RooAbsPdf(name, title),
      x("x", "x", this, _x),
      alpha("alpha", "alpha", this, _alpha),
      beta("beta", "beta", this, _beta),
      gamma("gamma", "gamma", this, _gamma),
      peak("peak", "peak", this, _peak) {}
 
RooCMSShape::RooCMSShape(const RooCMSShape& other, const char* name)
    : RooAbsPdf(other, name),
      x("x", this, other.x),
      alpha("alpha", this, other.alpha),
      beta("beta", this, other.beta),
      gamma("gamma", this, other.gamma),
      peak("peak", this, other.peak) {}
 
Double_t RooCMSShape::evaluate() const {
  // ENTER EXPRESSION IN TERMS OF VARIABLE ARGUMENTS HERE
 
  //Double_t erf = TMath::Erfc((alpha - x) * beta);
  Double_t erf = RooMath::erfc((alpha - x) * beta);
  Double_t u = (x - peak) * gamma;
 
  if (u < -70)
    u = 1e20;
  else if (u > 70)
    u = 0;
  else
    u = exp(-u);  //exponential decay
  return erf * u;
}