class HKalRungeKutta: public TObject

 Constructor that sets variables. It is still neccessary to set the field
 map pointer with
 setFieldMap(HMdcTrackGField *fpField, Double_t fpol)
 before Runge-Kutta propation can be used.

 Calculates energy loss due to bremsstrahlung and ionization.
 Radiation loss is done for electrons/positrons only.
 Only yields proper results if particle charge is +/- 1.
 Return value is energy loss in MeV/particle charge.

 Input and output:
 fProc:      Process noise matrix. Contribution of energy loss will be added.
 stateVec:   track state vector before energy loss. q/p of state will be modified.

 Input:
 length:     track length in mm
 qp:         charge/momentum before energy loss in MeV/c
 Z:          atomic number.
 ZoverA:     atomic number / atomic mass of passed material
 density:    density of passed material in g/mm^3
 radLength:  radiation length of passed material in mm
 exciteEner: mean excitation energy of passed material in MeV
 pid:        Geant particle ID.

 Returns the ionization energy loss for thin materials with Bethe-Bloch formula.
 - dE/dx = K/A * Z * z^2 / beta^2 * (0.5 * ln(2*me*beta^2*gamma^2*T_max/I^2) - beta^2)
 T_max = (2*me*beta^2*gamma^2) / (1 + 2*gamma*me/mass + me^2/mass^2)

 beta:       v/c of particle
 mass:       mass of the particle in MeV/c^2 traversing through the material
 ZoverA:     atomic number / atomic mass of passed material
 density:    density of passed material in g/mm^3
 exciteEner: mean excitation energy of passed material in MeV
 z:          atomic number of particle

 Calculates energy loss dE/dx in MeV/mm due to ionization for relativistic electrons/positrons.

 For formula used, see:
 Track fitting with energy loss
 Stampfer, Regler and Fruehwirth
 Computer Physics Communication 79 (1994), 157-164

 qp:         particle charge divided by momentum
 ZoverA:     atomic number / atomic mass of passed material
 density:    density of material in g/mm^3
 exciteEner: mean excitation energy in MeV
 pid:        Geant particle id (2 = positron, 3 = electron)

 Calculates energy loss due to bremsstrahlung for electrons with Bethe-Heitler formula.
 Sign will be negative if propagating in forward direction, i.e. particle looses energy.

 Input and Output:
 fProc:     process noise matrix (will be modified)

 Input:
 length:    track length in mm
 mass:      particle mass in MeV/c^2
 qp:        charge/momentum before energy loss in MeV/c
 radLength: radiation length of material in mm.

 Calculates the field integral (Tm) by numerical integration
 along the stepping points :  | sum over i  x(i+1)-x(i) x B(i) |
 where x and B are position and field vectors.
 setFillPointArrays(kTRUE) has to set before call propagate
 to fill the point and field arrays which are needed for
 the calculation.

 Calculate B-field (T) in x,y,z (mm).

 Output:
 btos: field vector at position xv in Tesla.

 Input:
 xv:   return field at this position vector. Coordinates must be in mm.
 fpol: field scaling factor * polarity.

 Calculate B-field (T) in x,y,z (mm).

 Output:
 btos: field vector at position xv in Tesla.

 Input:
 xv:   return field at this position vector. Coordinates must be in mm.
 fpol: field scaling factor * polarity.

 Calculates material budget for a Runge-Kutta step between two points.

 Output:
 A:          mass number of mixture of passed material.
 Z:          atomic number of mixture of passed material.
 density:    density of mixture of passed material in g/mm^3.
 radLength:  radiation length of mixture of passed material in mm.
 exciteEner: mean excitation energy of mixture of passed material in MeV.

 Input:
 posFrom:    position of track before the Runge-Kutta step.
 posTo:      position of track after the Runge-Kutta step.
 layerFrom:  measurement layer at posFrom.
 layerTo:    measurement layer at posTo.

 Add multiple scattering to the process noise covariance.

 Input and output:
 fProc:     Process noise covariance. Contribution of multiple scattering is added to this matrix.

 Input:
 stateVec:  Track state vector at start of an RK step.
 length:    Track length in mm.
 radLength: Radiation length of passed material in mm.
 beta:      v/c of particle.
 pid:       Geant particle ID.

 Resets error flags and clears arrays that store points/field along the trajectory.
 Call this every time when fitting a new track.

 Check propagator matrix for INFs/NaNs or too large entries.

 Reduces values in the state vector that have become too large to avoid
 overflows or underflows and print a warning.
 Returns whether track parameters have been reduced or not as a boolean.

 sv: 5-dimensional track state vector. May be modified in case of errors.

 Propagate a particle trajectory to a plane using a 4th order Runge-Kutta.
 Returns true if no errors were encountered during track propagation.

 Output:
 fProp:         Propagator matrix.
 fProc:         Process noise matrix with contributions of multiple scattering and energy loss.
 stateVecTo:    Track state vector (x,y,tx,ty,q/p) at target plane.

 Input:
 stateVecFrom:  Track state vector at (x,y,tx,ty,q/p) at source plane.
 planeFrom:     Plane to propagate from.
 planeTo:       Plane to propagate to.
 measLayerFrom: Detector element in which the track is currently in. Used to calculate passed materials.
 measLayerTo:   Detector element the track will move to. Used to calculate passed materials.
 pid:           Geant particle ID. Needed for process noise calculations.
 propDir:       propagation direction kIterForward or kIterBackward.
 bCalcJac:      Calculate the propagator matrix.

 Propagate the track state along a straight line in its current direction.
 (x',y',z') = (x,y,z) + dz * (tx,ty,1)

 Output:
 fPropChange: Change in propagator matrix.

 Input & output:
 stateVec:    Current track state vector (x,y,tx,ty,qp).

 Input:
 zPos:        Current z position of track.
 dz:          Step length in z coordinate.

 From the position and direction stored in the track state vector, propagate the track
 to a target plane using a straight line. The track state and reference layer are updated
 and the propagator matrix is calculated. The function returns the length of the straight line.
 The class variable trackPosAtZ must contain the current z-value of the track.

 Output:
 fPropChange: Change in propagator matrix

 Input and output:
 stateVec:    Current Track state vector.
 zPos:        Current z-position of track.

 Input:
 planeTo:     Plane to propagate to.
 propDir:     Propagation direction.

 One step of track tracing from track state.

 Input and output:
 stateVec:  Starting track parameters.
 fPropStep: Change in propagator matrix for this step.

 Input:
 totStep:   Step size.
 bCalcJac:  Update the Jacobian (propagator matrix).