#include <TimeStepper.hpp>

Collaboration diagram for TimeStepper:

Public Member Functions
	TimeStepper (double startTime, double endTime, double dt, bool enforceConstantTimeStep=false, std::vector< double > additionalTimes=std::vector< double >())

void	AdvanceOneTimeStep ()

double	GetTime () const

double	GetNextTime () const

double	GetNextTimeStep ()

double	GetIdealTimeStep ()

bool	IsTimeAtEnd () const

unsigned	EstimateTimeSteps () const

unsigned	GetTotalTimeStepsTaken () const

void	ResetTimeStep (double dt)

Private Member Functions
double	CalculateNextTime ()

template<class Archive >
void	serialize (Archive &archive, const unsigned int version)

Private Attributes
double	mStart

double	mEnd

double	mDt

unsigned	mTotalTimeStepsTaken

unsigned	mAdditionalTimesReachedDeprecated

double	mTime

double	mNextTime

double	mEpsilon

std::vector< double >	mAdditionalTimesDeprecated

Friends
class	TestTimeStepper

class	boost::serialization::access

Detailed Description

A helper class that provides a robust way to deal with time loops.

An incremented integer counter is used to calculate the current time and ensure that the correct end time.

Definition at line 52 of file TimeStepper.hpp.

Constructor & Destructor Documentation

TimeStepper::TimeStepper	(	double	startTime,
		double	endTime,
		double	dt,
		bool	enforceConstantTimeStep = `false`,
		std::vector< double >	additionalTimes = `std::vector<double> ()`
	)

Create a new time stepper over some simulation interval. Time units are not specified, but all parameters should have consistent units.

Parameters

startTime	the start of the interval
endTime	the end of the interval
dt	the time step to use.
enforceConstantTimeStep	If this is true the stepper estimates whether non-constant timesteps will be used and quits if so.
additionalTimes	If the timestepper needs to stop at certain additional times, they can be passed in in this std::vector. Defaults to empty. These times must be in ascending order. DEPRECATION: Note that these additional times are checked in order to ensure that the stepper will stop at these times anyway. (For example we want to check that new events such as electrodes switching on will be detected in a regular time loop.) New additional times are not added but instead throw an exception.

Definition at line 43 of file TimeStepper.cpp.

References CalculateNextTime(), Divides(), EstimateTimeSteps(), EXCEPTION, mDt, mEnd, mEpsilon, mNextTime, and mStart.

Member Function Documentation

void TimeStepper::AdvanceOneTimeStep ( )

Step forward one step in time and update member variables.

Note that this should be used in conjuction with (not) IsTimeAtEnd because increment past the end of a simulation will throw an exception

Definition at line 123 of file TimeStepper.cpp.

References CalculateNextTime(), EXCEPTION, mNextTime, mTime, and mTotalTimeStepsTaken.

Referenced by AbstractGeneralizedRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractOneStepIvpOdeSolver::InternalSolve(), NhsModelWithBackwardSolver::RunDoNotUpdate(), AbstractOneStepIvpOdeSolver::Solve(), AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), CvodeAdaptor::Solve(), AbstractCvodeSystem::Solve(), AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), AbstractGeneralizedRushLarsenCardiacCell::SolveAndUpdateState(), AbstractRushLarsenCardiacCell::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< SIZE >::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< 0u >::SolveAndUpdateState(), and LinearParabolicPdeSystemWithCoupledOdeSystemSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::SolveAndWriteResultsToFile().

double TimeStepper::CalculateNextTime ( )

private

Returns: Compute what the time will be at the next time step.

Definition at line 106 of file TimeStepper.cpp.

References mAdditionalTimesDeprecated, mDt, mEnd, mEpsilon, mStart, and mTotalTimeStepsTaken.

Referenced by AdvanceOneTimeStep(), ResetTimeStep(), and TimeStepper().

unsigned TimeStepper::EstimateTimeSteps ( ) const

Returns: Estimate number of time steps remaining, which may be an overestimate. Used to reserve memory for writing intermediate solutions.

Definition at line 170 of file TimeStepper.cpp.

References mDt, mEnd, and mStart.

Referenced by ExtendedBidomainProblem< DIM >::DefineWriterColumns(), AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::DefineWriterColumns(), AbstractOneStepIvpOdeSolver::Solve(), CvodeAdaptor::Solve(), AbstractCvodeSystem::Solve(), and TimeStepper().

double TimeStepper::GetIdealTimeStep ( )

Returns: Get the size of the ideal time step (as set in the constructor. Note when a fixed time step is used GetNextTimeStep() == GetIdealTimeStep() until the end of the simulation when GetNextTimeStep() == 0.0

Definition at line 160 of file TimeStepper.cpp.

References mDt.

double TimeStepper::GetNextTime ( ) const

Returns: Get the value time will take at the next step.

Definition at line 144 of file TimeStepper.cpp.

References mNextTime.

Referenced by AbstractOneStepIvpOdeSolver::Solve(), AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), CvodeAdaptor::Solve(), AbstractCvodeSystem::Solve(), AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), and LinearParabolicPdeSystemWithCoupledOdeSystemSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::SolveAndWriteResultsToFile().

double TimeStepper::GetNextTimeStep ( )

Returns: Get the size of the next time step which will be taken.

Note that this is often reported as the ideal timestep.

The actual time step is mNextTime - mTime which could be evaluated as GetNextTime() - GetTime() = mNextTime - mTime ~= mStart + (mTotalTimeStepsTaken + 1)*mDt - mTime ~= mStart + (mTotalTimeStepsTaken + 1)*mDt - (mStart + mTotalTimeStepsTaken*mDt) ~= mDt This wouldn't evaluate to mDt in general but to within 2*mEpsilon <= 2*DBL_EPSILON*mEnd*mDt When mEnd and/or mStart are a long way from zero then the value reported mNextTime - mTime will differ from mDt, even when the actual step is mDt

Definition at line 149 of file TimeStepper.cpp.

References mAdditionalTimesDeprecated, mDt, mEnd, mNextTime, and mTime.

Referenced by AbstractOneStepIvpOdeSolver::InternalSolve(), and AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve().

double TimeStepper::GetTime ( ) const

Returns: Get the time.

Definition at line 139 of file TimeStepper.cpp.

References mTime.

Referenced by AbstractGeneralizedRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractOneStepIvpOdeSolver::InternalSolve(), AbstractOneStepIvpOdeSolver::Solve(), AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), AbstractCvodeSystem::Solve(), AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), AbstractGeneralizedRushLarsenCardiacCell::SolveAndUpdateState(), AbstractRushLarsenCardiacCell::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< SIZE >::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< 0u >::SolveAndUpdateState(), and LinearParabolicPdeSystemWithCoupledOdeSystemSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::SolveAndWriteResultsToFile().

unsigned TimeStepper::GetTotalTimeStepsTaken ( ) const

Returns: The number of times AdvanceOneTimeStep() has been called SINCE the last time ResetTimeStep() was called.

Definition at line 175 of file TimeStepper.cpp.

References mTotalTimeStepsTaken.

Referenced by AbstractOneStepIvpOdeSolver::Solve(), AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), CvodeAdaptor::Solve(), AbstractCvodeSystem::Solve(), and LinearParabolicPdeSystemWithCoupledOdeSystemSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::SolveAndWriteResultsToFile().

bool TimeStepper::IsTimeAtEnd ( ) const

Returns: True when GetTime == endTime.

Definition at line 165 of file TimeStepper.cpp.

References mEnd, and mTime.

Referenced by AbstractGeneralizedRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractRushLarsenCardiacCell::ComputeExceptVoltage(), AbstractOneStepIvpOdeSolver::InternalSolve(), NhsModelWithBackwardSolver::RunDoNotUpdate(), AbstractOneStepIvpOdeSolver::Solve(), AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), CvodeAdaptor::Solve(), AbstractCvodeSystem::Solve(), AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve(), AbstractGeneralizedRushLarsenCardiacCell::SolveAndUpdateState(), AbstractRushLarsenCardiacCell::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< SIZE >::SolveAndUpdateState(), AbstractBackwardEulerCardiacCell< 0u >::SolveAndUpdateState(), and LinearParabolicPdeSystemWithCoupledOdeSystemSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::SolveAndWriteResultsToFile().

void TimeStepper::ResetTimeStep ( double dt )

Set the time step to use for adaptive time integration. Note that this also resets mStart to be the current time and zeroes mTotalTimeStepsTaken.

Parameters

dt	is the new time-step

Definition at line 180 of file TimeStepper.cpp.

References CalculateNextTime(), mDt, mNextTime, mStart, mTime, and mTotalTimeStepsTaken.

Referenced by AbstractDynamicLinearPdeSolver< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::Solve().

template<class Archive >

void TimeStepper::serialize	(	Archive &	archive,
		const unsigned int	version
	)

inlineprivate

Parameters

archive	the archive
version	the current version of this class

Definition at line 193 of file TimeStepper.hpp.

References CHASTE_CLASS_EXPORT, mAdditionalTimesDeprecated, mAdditionalTimesReachedDeprecated, mDt, mEnd, mEpsilon, mNextTime, mStart, mTime, and mTotalTimeStepsTaken.

Friends And Related Function Documentation

friend class boost::serialization::access

friend

Needed for serialization.

Definition at line 185 of file TimeStepper.hpp.

Member Data Documentation

std::vector<double> TimeStepper::mAdditionalTimesDeprecated

private

DEPRECATED Vector to store the additional times the stepper must stop at.

Definition at line 182 of file TimeStepper.hpp.

Referenced by CalculateNextTime(), GetNextTimeStep(), and serialize().

unsigned TimeStepper::mAdditionalTimesReachedDeprecated

private

DEPRECATED: The number of times one of the 'additional times' has been passed.

Definition at line 163 of file TimeStepper.hpp.

Referenced by serialize().

double TimeStepper::mDt

private

The size of time step.

Definition at line 157 of file TimeStepper.hpp.

Referenced by CalculateNextTime(), EstimateTimeSteps(), GetIdealTimeStep(), GetNextTimeStep(), ResetTimeStep(), serialize(), and TimeStepper().

double TimeStepper::mEnd

private

The end time.

Definition at line 154 of file TimeStepper.hpp.

Referenced by CalculateNextTime(), EstimateTimeSteps(), GetNextTimeStep(), IsTimeAtEnd(), serialize(), and TimeStepper().

double TimeStepper::mEpsilon

private

An architecture-dependent scaling factor. This is so that we can compare relative to the end time when mEnd is large. mEpsilon = DBL_EPSILON when mEnd is small = mEnd*DBL_EPSILON when mEnd is large

Definition at line 176 of file TimeStepper.hpp.

Referenced by CalculateNextTime(), serialize(), and TimeStepper().

double TimeStepper::mNextTime

private

What the time will be after the next time step.

Definition at line 169 of file TimeStepper.hpp.

Referenced by AdvanceOneTimeStep(), GetNextTime(), GetNextTimeStep(), ResetTimeStep(), serialize(), and TimeStepper().