LinearFit Class Reference

#include <LinearFit.h>

Public Member Functions
void	fit (const std::vector< float > &x, const std::vector< float > &y, int ndat, const std::vector< float > &sigy, float &slope, float &intercept, float &covss, float &covii, float &covsi) const

Detailed Description

Straight line fit for data with errors on one coordinate

Definition at line 9 of file LinearFit.h.

Member Function Documentation

void LinearFit::fit	(	const std::vector< float > &	x,
		const std::vector< float > &	y,
		int	ndat,
		const std::vector< float > &	sigy,
		float &	slope,
		float &	intercept,
		float &	covss,
		float &	covii,
		float &	covsi
	)		const

x_i, y_i: measurements of y(x_i), sigy_i: error (sigma) on y_i slope, intercept: fitted parameters covss, covii, covsi: covariance matrix of fitted parameters, s denoting slope, i denoting intercept

Definition at line 3 of file LinearFit.cc.

References diffTwoXMLs::g1, diffTwoXMLs::g2, and i.

Referenced by LinearFitErrorsIn2Coord::slope().

{

  float g1 = 0, g2 = 0;
  float s11 = 0, s12 = 0, s22 = 0;
  for (int i = 0; i != ndat; i++) {
    float sy2 = sigy[i] * sigy[i];
    g1 += y[i] / sy2;
    g2 += x[i]*y[i] / sy2;
    s11 += 1. / sy2;
    s12 += x[i] / sy2;
    s22 += x[i]*x[i] / sy2;
  }

  float d = s11*s22 - s12*s12;
  intercept = (g1*s22 - g2*s12) / d;
  slope = (g2*s11 - g1*s12) / d;

  covii =  s22 / d;
  covss =  s11 / d;
  covsi = -s12 / d;
  
}