Runtime options for Runge-Kutta propagation. More...

#include <Acts/Propagator/GenericRungeKuttaStepper.hpp>

Inheritance diagram for Acts::GenericRungeKuttaStepper::Options:

Collaboration diagram for Acts::GenericRungeKuttaStepper::Options:

Public Member Functions
	Options (const GeometryContext &gctx, const MagneticFieldContext &mctx)
	Constructor.
void	setPlainOptions (const StepperPlainOptions &options)
	Set plain stepper options.
Public Member Functions inherited from Acts::StepperPlainOptions
	StepperPlainOptions (const GeometryContext &gctx, const MagneticFieldContext &mctx)
	StepperPlainOptions with context.

Public Attributes
ErrorEstimation	errorEstimation = ErrorEstimation::Embedded
	Error estimate for the adaptive step size.
double	fieldGradientEpsilon = 10 * UnitConstants::um
	Distance of the field lookups for the finite-difference gradient, if the field provider does not implement the gradient.
bool	includeFieldGradient = true
	Include the field gradient in the transport jacobian.
bool	jacobianInErrorEstimate = false
	Include the jacobian in the error estimate.
Public Attributes inherited from Acts::StepperPlainOptions
Dense	dense
	Options for dense material effects.
std::reference_wrapper< const GeometryContext >	geoContext
	Context object for the geometry.
double	initialStepSize = 10 * Acts::UnitConstants::m
	Initial step size.
std::reference_wrapper< const MagneticFieldContext >	magFieldContext
	Context object for the magnetic field.
unsigned int	maxRungeKuttaStepTrials = 10000
	Maximum number of Runge-Kutta steps for the stepper step call.
double	maxStepSize = std::numeric_limits<double>::max()
	Absolute maximum step size.
double	stepSizeCutOff = 0.
	Cut-off value for the step size.
double	stepTolerance = 1e-4
	Tolerance for the error of the integration.

Detailed Description

Runtime options for Runge-Kutta propagation.

Constructor & Destructor Documentation

◆ Options()

Acts::GenericRungeKuttaStepper::Options::Options	(	const GeometryContext &	gctx,
		const MagneticFieldContext &	mctx )

Constructor.

Parameters

gctx	Geometry context
mctx	Magnetic field context

Member Function Documentation

◆ setPlainOptions()

void Acts::GenericRungeKuttaStepper::Options::setPlainOptions ( const StepperPlainOptions & options )

Set plain stepper options.

Parameters

options Plain stepper options

Member Data Documentation

◆ fieldGradientEpsilon

double Acts::GenericRungeKuttaStepper::Options::fieldGradientEpsilon = 10 * UnitConstants::um

Distance of the field lookups for the finite-difference gradient, if the field provider does not implement the gradient.

The truncation error of the central difference is of order epsilon^2. The round-off error is of order eps_machine |B| / epsilon, which is about 1e-14 |B| per mm for 10 um. This is far below the gradient of a real field. 10 um is also small compared with a field-map cell, so the lookups usually stay in one cell.

◆ jacobianInErrorEstimate

bool Acts::GenericRungeKuttaStepper::Options::jacobianInErrorEstimate = false

Include the jacobian in the error estimate.

Its entries have mixed units, and the stepper uses the same tolerance for all of them.

Public Member Functions

Public Attributes