hf_egamma Namespace Reference

Functions
double	eSeLCorrected (double es, double el, double m, double b)
double	eSeLCorrected (double es, double el, double pc, double px, double py)
double	eSeLCorrected (double es, double el, int era)

Function Documentation

double hf_egamma::eSeLCorrected	(	double	es,
		double	el,
		double	m,
		double	b
	)

double hf_egamma::eSeLCorrected	(	double	es,
		double	el,
		double	pc,
		double	px,
		double	py
	)

Definition at line 7 of file HFEGammaSLCorrector.cc.

References log.

Referenced by eSeLCorrected(), and HFRecoEcalCandidateAlgo::produce().

                                                                              {
    double x = std::log(el / 100);
    double y = es / el;
    return pc + px * x + py * y;
  }

double hf_egamma::eSeLCorrected	(	double	es,
		double	el,
		int	era
	)

Definition at line 13 of file HFEGammaSLCorrector.cc.

References eSeLCorrected().

                                                      {
    double pc = 0.0, px = 0.0, py = 0.0;

    switch (era) {
      case (0):  //Data 41
        pc = -1.02388e-1;
        px = -1.51130e-1;
        py = 9.88514e-1;
        break;
      case (1):  //Fall 10 MC
        pc = -4.06012e-2;
        px = -1.34769e-1;
        py = 9.90877e-1;
        break;
      case (2):  //Spring 11 MC
        pc = 5.98732e-3;
        px = -1.74767e-1;
        py = 9.84610e-1;
        break;
      case (3):  //Summer 11 MC
        pc = -0.036416;
        px = -0.195854;
        py = 0.980633;
        break;
      case (4):  //July 5 Data
        pc = -0.008077;
        px = -0.216002;
        py = 0.976393;
        break;
    }

    //After fitting the 2D histogram, we find a y-intercept b, a slope m,
    //and a point x0 around which we choose to rotate the data points. We
    //will map (x,y) --> (x,y') where y' = pc + px*x + py*y, with
    //pc = sin(atan(m))*x0 - cos(atan(m))*(m*x0+b), px = -sin(atan(m)),
    //and py = cos(atan(m)). This transformation preserves the x-value of the
    //data point and takes y' to be the y-value of the original point after it
    //is rotated through angle atan(m) (to flatten the line of best fit) and
    //transposed vertically downward by b (to make the line of best fit coincide with the x-axis).

    return eSeLCorrected(es, el, pc, px, py);
  }