eta.h

Go to the documentation of this file.

// @(#)root/mathcore:$Id$
// Authors: W. Brown, M. Fischler, L. Moneta    2005

 /**********************************************************************
  *                                                                    *
  * Copyright (c) 2005 , FNAL MathLib Team                             *
  *                                                                    *
  *                                                                    *
  **********************************************************************/


// Header source file for function calculating eta
//
// Created by: Lorenzo Moneta  at 14 Jun 2007


#ifndef ROOT_Math_GenVector_eta
#define ROOT_Math_GenVector_eta  1

#include "Math/GenVector/etaMax.h"


#include <limits>
#include <cmath>


namespace ROOT {

  namespace Math {

     namespace Impl {

    /**
        Calculate eta given rho and zeta.
        This formula is faster than the standard calculation (below) from log(tan(theta/2)
        but one has to be careful when rho is much smaller than z (large eta values)
        Formula is  eta = log( zs + sqrt(zs^2 + 1) )  where zs = z/rho

        For large value of z_scaled (tan(theta) ) one can appoximate the sqrt via a Taylor expansion
        We do the approximation of the sqrt if the numerical error is of the same order of second term of
        the sqrt.expansion:
        eps > 1/zs^4   =>   zs > 1/(eps^0.25)

        When rho == 0 we use etaMax (see definition in etaMax.h)

     */
        template<typename Scalar>
        inline Scalar Eta_FromRhoZ(Scalar rho, Scalar z) {
           if (rho > 0) {

              // value to control Taylor expansion of sqrt
              static const Scalar big_z_scaled = pow(std::numeric_limits<Scalar>::epsilon(), static_cast<Scalar>(-.25));

              Scalar z_scaled = z/rho;
              if (std::fabs(z_scaled) < big_z_scaled) {
                 return log(z_scaled + sqrt(z_scaled * z_scaled + 1.0));
              } else {
                 // apply correction using first order Taylor expansion of sqrt
                 return z > 0 ? log(2.0 * z_scaled + 0.5 / z_scaled) : -log(-2.0 * z_scaled);
              }
           }
           // case vector has rho = 0
           else if (z==0) {
              return 0;
           }
           else if (z>0) {
              return z + etaMax<Scalar>();
           }
           else {
              return z - etaMax<Scalar>();
           }

        }


        /**
           Implementation of eta from -log(tan(theta/2)).
           This is convenient when theta is already known (for example in a polar coorindate system)
        */
        template<typename Scalar>
        inline Scalar Eta_FromTheta(Scalar theta, Scalar r) {
           Scalar tanThetaOver2 = tan(theta / 2.);
           if (tanThetaOver2 == 0) {
              return r + etaMax<Scalar>();
           }
           else if (tanThetaOver2 > std::numeric_limits<Scalar>::max()) {
              return -r - etaMax<Scalar>();
           }
           else {
              return -log(tanThetaOver2);
           }

        }

     } // end namespace Impl

  } // namespace Math

} // namespace ROOT


#endif /* ROOT_Math_GenVector_etaMax  */