#include <elxPreconditionedGradientDescent.h>

Detailed Description

template<class TElastix>
class elastix::PreconditionedGradientDescent< TElastix >

A gradient descent optimizer with a decaying gain.

This class is a wrap around the AdaptiveStochasticPreconditionedGradientOptimizer class. It takes care of setting parameters and printing progress information. For more information about the optimization method, please read the documentation of the AdaptiveStochasticPreconditionedGradientOptimizer class.

The parameters used in this class are:

Parameters:

Optimizer: Select this optimizer as follows:
(Optimizer "PreconditionedGradientDescent")

MaximumNumberOfIterations: The maximum number of iterations in each resolution.
example: (MaximumNumberOfIterations 100 100 50)
Default/recommended value: 500.

MaximumNumberOfSamplingAttempts: The maximum number of sampling attempts. Sometimes not enough corresponding samples can be drawn, upon which an exception is thrown. With this parameter it is possible to try to draw another set of samples.
example: (MaximumNumberOfSamplingAttempts 10 15 10)
Default value: 0, i.e. just fail immediately, for backward compatibility.

SP_a: The gain at each iteration is defined by
$a(k) = SP\_a / (SP\_A + k + 1)^{SP\_alpha}$ .
SP_a can be defined for each resolution.
example: (SP_a 3200.0 3200.0 1600.0)
The default value is 400.0. Tuning this variable for you specific problem is recommended.

SP_A: The gain at each iteration is defined by
$a(k) = SP\_a / (SP\_A + k + 1)^{SP\_alpha}$ .
SP_A can be defined for each resolution.
example: (SP_A 50.0 50.0 100.0)
The default/recommended value is 50.0.

SP_alpha: The gain at each iteration is defined by
$a(k) = SP\_a / (SP\_A + k + 1)^{SP\_alpha}$ .
SP_alpha can be defined for each resolution.
example: (SP_alpha 0.602 0.602 0.602)
The default/recommended value is 0.602.

See also: StochasticPreconditionedGradientOptimizer

Definition at line 72 of file elxPreconditionedGradientDescent.h.

Inheritance diagram for elastix::PreconditionedGradientDescent< TElastix >:

[legend]

Data Structures
struct	SettingsType

Public Types
using	ConstPointer = itk::SmartPointer< const Self >

using	ITKBaseType = typename Superclass2::ITKBaseType

using	Pointer = itk::SmartPointer< Self >

using	Self = PreconditionedGradientDescent

using	Superclass1 = AdaptiveStochasticPreconditionedGradientDescentOptimizer

using	Superclass2 = OptimizerBase< TElastix >

Public Types inherited from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer
using	ConstPointer = SmartPointer< const Self >

using	Pointer = SmartPointer< Self >

using	Self = AdaptiveStochasticPreconditionedGradientDescentOptimizer

using	Superclass = StochasticPreconditionedGradientDescentOptimizer

Public Types inherited from itk::StochasticPreconditionedGradientDescentOptimizer
using	ConstPointer = SmartPointer< const Self >

using	Pointer = SmartPointer< Self >

using	PreconditionType = vnl_sparse_matrix< PreconditionValueType >

using	PreconditionValueType = DerivativeType::ValueType

using	ScaledCostFunctionPointer = ScaledCostFunctionType::Pointer

using	ScaledCostFunctionType = ScaledSingleValuedCostFunction

using	ScalesType = NonLinearOptimizer::ScalesType

using	Self = StochasticPreconditionedGradientDescentOptimizer

enum	StopConditionType

using	Superclass = PreconditionedGradientDescentOptimizer

Public Types inherited from itk::PreconditionedGradientDescentOptimizer
using	ConstPointer = SmartPointer< const Self >

using	Pointer = SmartPointer< Self >

using	PreconditionType = vnl_sparse_matrix< PreconditionValueType >

using	PreconditionValueType = DerivativeType::ValueType

using	ScaledCostFunctionPointer = ScaledCostFunctionType::Pointer

using	ScaledCostFunctionType = ScaledSingleValuedCostFunction

using	ScalesType = NonLinearOptimizer::ScalesType

using	Self = PreconditionedGradientDescentOptimizer

enum	StopConditionType { MaximumNumberOfIterations , MetricError , MinimumStepSize }

using	Superclass = ScaledSingleValuedNonLinearOptimizer

Public Types inherited from itk::ScaledSingleValuedNonLinearOptimizer
using	ConstPointer = SmartPointer< const Self >

using	Pointer = SmartPointer< Self >

using	ScaledCostFunctionPointer = ScaledCostFunctionType::Pointer

using	ScaledCostFunctionType = ScaledSingleValuedCostFunction

using	ScalesType = NonLinearOptimizer::ScalesType

using	Self = ScaledSingleValuedNonLinearOptimizer

using	Superclass = SingleValuedNonLinearOptimizer

Public Types inherited from elastix::OptimizerBase< TElastix >
using	ElastixType = TElastix

using	ITKBaseType = itk::Optimizer

using	ParametersType = typename ITKBaseType::ParametersType

using	RegistrationType = typename ElastixType::RegistrationBaseType

using	Self = OptimizerBase

using	Superclass = BaseComponentSE< TElastix >

Public Types inherited from elastix::BaseComponentSE< TElastix >
using	ConfigurationPointer = Configuration::Pointer

using	ElastixType = TElastix

using	RegistrationType = typename ElastixType::RegistrationBaseType

using	Self = BaseComponentSE

using	Superclass = BaseComponent

Public Member Functions
virtual void	AfterEachIteration ()

virtual void	AfterEachResolution ()

virtual void	AfterRegistration ()

virtual void	BeforeEachResolution ()

virtual void	BeforeRegistration ()

	elxClassNameMacro ("PreconditionedGradientDescent")

virtual bool	GetAutomaticParameterEstimation () const

virtual const char *	GetClassName () const

virtual const unsigned long &	GetMaximumNumberOfSamplingAttempts ()

	ITK_DISALLOW_COPY_AND_MOVE (PreconditionedGradientDescent)

virtual void	MetricErrorResponse (itk::ExceptionObject &err)

virtual void	ResumeOptimization ()

virtual void	SetAutomaticParameterEstimation (bool _arg)

virtual void	SetCurrentPositionPublic (const ParametersType &param)

virtual void	SetMaximumNumberOfSamplingAttempts (unsigned long _arg)

virtual void	StartOptimization ()

Public Member Functions inherited from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer
virtual const char *	GetClassName () const

virtual double	GetSigmoidMax () const

virtual double	GetSigmoidMin () const

virtual double	GetSigmoidScale () const

virtual bool	GetUseAdaptiveStepSizes () const

	ITK_DISALLOW_COPY_AND_MOVE (AdaptiveStochasticPreconditionedGradientDescentOptimizer)

virtual void	SetSigmoidMax (double _arg)

virtual void	SetSigmoidMin (double _arg)

virtual void	SetSigmoidScale (double _arg)

virtual void	SetUseAdaptiveStepSizes (bool _arg)

Public Member Functions inherited from itk::StochasticPreconditionedGradientDescentOptimizer
virtual void	AdvanceOneStep ()

virtual const char *	GetClassName () const

virtual double	GetCurrentTime () const

virtual double	GetInitialTime () const

virtual double	GetParam_a () const

virtual double	GetParam_A () const

virtual double	GetParam_alpha () const

	ITK_DISALLOW_COPY_AND_MOVE (StochasticPreconditionedGradientDescentOptimizer)

virtual void	SetInitialTime (double _arg)

virtual void	SetParam_a (double _arg)

virtual void	SetParam_A (double _arg)

virtual void	SetParam_alpha (double _arg)

virtual void	StartOptimization ()

Public Member Functions inherited from itk::PreconditionedGradientDescentOptimizer
virtual void	AdvanceOneStep ()

const cholmod_common *	GetCholmodCommon () const

const cholmod_factor *	GetCholmodFactor () const

virtual const char *	GetClassName () const

virtual double	GetConditionNumber () const

virtual unsigned int	GetCurrentIteration () const

virtual double	GetDiagonalWeight () const

virtual const DerivativeType &	GetGradient ()

virtual double	GetLargestEigenValue () const

virtual const double &	GetLearningRate ()

virtual double	GetMinimumGradientElementMagnitude () const

virtual const unsigned long &	GetNumberOfIterations ()

virtual const DerivativeType &	GetSearchDirection ()

virtual double	GetSparsity () const

virtual const StopConditionType &	GetStopCondition ()

virtual const double &	GetValue ()

	ITK_DISALLOW_COPY_AND_MOVE (PreconditionedGradientDescentOptimizer)

virtual void	MetricErrorResponse (ExceptionObject &err)

virtual void	ResumeOptimization ()

virtual void	SetDiagonalWeight (double _arg)

virtual void	SetLearningRate (double _arg)

virtual void	SetMinimumGradientElementMagnitude (double _arg)

virtual void	SetNumberOfIterations (unsigned long _arg)

virtual void	SetPreconditionMatrix (PreconditionType &precondition)

virtual void	StartOptimization ()

virtual void	StopOptimization ()

Public Member Functions inherited from itk::ScaledSingleValuedNonLinearOptimizer
virtual const char *	GetClassName () const

const ParametersType &	GetCurrentPosition () const override

virtual bool	GetMaximize () const

virtual const ScaledCostFunctionType *	GetScaledCostFunction ()

virtual const ParametersType &	GetScaledCurrentPosition ()

bool	GetUseScales () const

virtual void	InitializeScales ()

	ITK_DISALLOW_COPY_AND_MOVE (ScaledSingleValuedNonLinearOptimizer)

virtual void	MaximizeOff ()

virtual void	MaximizeOn ()

void	SetCostFunction (CostFunctionType *costFunction) override

virtual void	SetMaximize (bool _arg)

virtual void	SetUseScales (bool arg)

Public Member Functions inherited from elastix::OptimizerBase< TElastix >
void	AfterRegistrationBase () override

void	BeforeEachResolutionBase () override

ITKBaseType *	GetAsITKBaseType ()

const ITKBaseType *	GetAsITKBaseType () const

virtual const char *	GetClassName () const

	ITK_DISALLOW_COPY_AND_MOVE (OptimizerBase)

virtual void	SetCurrentPositionPublic (const ParametersType &param)

virtual void	SetSinusScales (double amplitude, double frequency, unsigned long numberOfParameters)

Public Member Functions inherited from elastix::BaseComponentSE< TElastix >
void	AddTargetCellToIterationInfo (const char *const name)

Configuration *	GetConfiguration () const

ElastixType *	GetElastix () const

xl::xoutbase &	GetIterationInfoAt (const char *const name)

RegistrationType *	GetRegistration () const

	ITK_DISALLOW_COPY_AND_MOVE (BaseComponentSE)

int	RemoveTargetCellFromIterationInfo (const char *const name)

void	SetConfiguration (Configuration *_arg)

void	SetElastix (ElastixType *_arg)

Public Member Functions inherited from elastix::BaseComponent
virtual void	AfterEachIteration ()

virtual void	AfterEachIterationBase ()

virtual void	AfterEachResolution ()

virtual void	AfterEachResolutionBase ()

virtual void	AfterRegistration ()

virtual void	AfterRegistrationBase ()

virtual int	BeforeAll ()

virtual int	BeforeAllBase ()

virtual void	BeforeEachResolution ()

virtual void	BeforeEachResolutionBase ()

virtual void	BeforeRegistration ()

virtual void	BeforeRegistrationBase ()

virtual const char *	elxGetClassName () const

const char *	GetComponentLabel () const

	ITK_DISALLOW_COPY_AND_MOVE (BaseComponent)

	itkTypeMacroNoParent (BaseComponent)

void	SetComponentLabel (const char *label, unsigned int idx)

Static Public Member Functions
static Pointer	New ()

Static Public Member Functions inherited from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer
static Pointer	New ()

Static Public Member Functions inherited from itk::StochasticPreconditionedGradientDescentOptimizer
static Pointer	New ()

Static Public Member Functions inherited from itk::PreconditionedGradientDescentOptimizer
static Pointer	New ()

Static Public Member Functions inherited from itk::ScaledSingleValuedNonLinearOptimizer
static Pointer	New ()

Static Public Member Functions inherited from elastix::BaseComponent
template<typename TBaseComponent >
static auto	AsITKBaseType (TBaseComponent *const baseComponent) -> decltype(baseComponent->GetAsITKBaseType())

static void	InitializeElastixExecutable ()

static bool	IsElastixLibrary ()

Protected Types
using	FixedImageType = typename RegistrationType::FixedImageType

using	ImageGridSamplerPointer = typename ImageGridSamplerType::Pointer

using	ImageGridSamplerType = itk::ImageGridSampler< FixedImageType >

using	ImageRandomCoordinateSamplerPointer = typename ImageRandomCoordinateSamplerType::Pointer

using	ImageRandomCoordinateSamplerType = itk::ImageRandomCoordinateSampler< FixedImageType >

using	ImageRandomSamplerBasePointer = typename ImageRandomSamplerBaseType::Pointer

using	ImageRandomSamplerBaseType = itk::ImageRandomSamplerBase< FixedImageType >

using	ImageSamplerBasePointer = typename ImageSamplerBaseType::Pointer

using	ImageSamplerBaseType = itk::ImageSamplerBase< FixedImageType >

using	MovingImageType = typename RegistrationType::MovingImageType

using	ProgressCommandPointer = typename ProgressCommand::Pointer

using	ProgressCommandType = ProgressCommand

using	RandomGeneratorPointer = typename RandomGeneratorType::Pointer

using	RandomGeneratorType = itk::Statistics::MersenneTwisterRandomVariateGenerator

using	SettingsVectorType = typename std::vector< SettingsType >

Protected Types inherited from itk::PreconditionedGradientDescentOptimizer
using	cholmod_l = int CInt

Protected Member Functions
virtual void	AddRandomPerturbation (const ParametersType &initialParameters, ParametersType &perturbedParameters, double sigma)

virtual void	AutomaticParameterEstimation ()

virtual void	GetScaledDerivativeWithExceptionHandling (const ParametersType &parameters, DerivativeType &derivative)

	PreconditionedGradientDescent ()

virtual void	PrintSettingsVector (const SettingsVectorType &settings) const

virtual void	SampleGradients (const ParametersType &mu0, double &sigma1, double &sigma2)

virtual void	SetSelfHessian ()

virtual	~PreconditionedGradientDescent ()

Protected Member Functions inherited from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer
	AdaptiveStochasticPreconditionedGradientDescentOptimizer ()

virtual void	UpdateCurrentTime ()

virtual	~AdaptiveStochasticPreconditionedGradientDescentOptimizer ()

Protected Member Functions inherited from itk::StochasticPreconditionedGradientDescentOptimizer
virtual double	Compute_a (double k) const

	StochasticPreconditionedGradientDescentOptimizer ()

virtual void	UpdateCurrentTime ()

virtual	~StochasticPreconditionedGradientDescentOptimizer ()

Protected Member Functions inherited from itk::PreconditionedGradientDescentOptimizer
virtual void	CholmodSolve (const DerivativeType &gradient, DerivativeType &searchDirection, int solveType=CHOLMOD_A)

	PreconditionedGradientDescentOptimizer ()

void	PrintSelf (std::ostream &os, Indent indent) const

virtual	~PreconditionedGradientDescentOptimizer ()

Protected Member Functions inherited from itk::ScaledSingleValuedNonLinearOptimizer
virtual void	GetScaledDerivative (const ParametersType &parameters, DerivativeType &derivative) const

virtual MeasureType	GetScaledValue (const ParametersType &parameters) const

virtual void	GetScaledValueAndDerivative (const ParametersType &parameters, MeasureType &value, DerivativeType &derivative) const

void	PrintSelf (std::ostream &os, Indent indent) const override

	ScaledSingleValuedNonLinearOptimizer ()

void	SetCurrentPosition (const ParametersType &param) override

virtual void	SetScaledCurrentPosition (const ParametersType &parameters)

	~ScaledSingleValuedNonLinearOptimizer () override=default

Protected Member Functions inherited from elastix::OptimizerBase< TElastix >
virtual bool	GetNewSamplesEveryIteration () const

	OptimizerBase ()=default

virtual void	SelectNewSamples ()

	~OptimizerBase () override=default

Protected Member Functions inherited from elastix::BaseComponentSE< TElastix >
	BaseComponentSE ()=default

	~BaseComponentSE () override=default

Protected Member Functions inherited from elastix::BaseComponent
	BaseComponent ()=default

virtual	~BaseComponent ()=default

Protected Attributes
unsigned int	m_NumberOfGradientMeasurements

unsigned int	m_NumberOfSamplesForExactGradient

RandomGeneratorPointer	m_RandomGenerator

SettingsVectorType	m_SettingsVector

double	m_SigmoidScaleFactor

Protected Attributes inherited from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer
DerivativeType	m_PreviousSearchDirection

Protected Attributes inherited from itk::StochasticPreconditionedGradientDescentOptimizer
double	m_CurrentTime { 0.0 }

Protected Attributes inherited from itk::PreconditionedGradientDescentOptimizer
cholmod_common *	m_CholmodCommon

cholmod_factor *	m_CholmodFactor { nullptr }

cholmod_sparse *	m_CholmodGradient { nullptr }

double	m_ConditionNumber { 1.0 }

DerivativeType	m_Gradient

double	m_LargestEigenValue { 1.0 }

double	m_LearningRate { 1.0 }

DerivativeType	m_SearchDirection

double	m_Sparsity { 1.0 }

StopConditionType	m_StopCondition { MaximumNumberOfIterations }

Protected Attributes inherited from itk::ScaledSingleValuedNonLinearOptimizer
ScaledCostFunctionPointer	m_ScaledCostFunction

ParametersType	m_ScaledCurrentPosition

Protected Attributes inherited from elastix::BaseComponentSE< TElastix >
ConfigurationPointer	m_Configuration {}

itk::WeakPointer< TElastix >	m_Elastix {}

RegistrationType *	m_Registration {}

Private Attributes
	elxOverrideGetSelfMacro

bool	m_AutomaticParameterEstimation

bool	m_AutomaticParameterEstimationDone

unsigned long	m_CurrentNumberOfSamplingAttempts

unsigned long	m_MaximumNumberOfSamplingAttempts

bool	m_PreconditionMatrixSet

unsigned long	m_PreviousErrorAtIteration

Member Typedef Documentation

◆ ConstPointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ConstPointer = itk::SmartPointer<const Self>

Definition at line 84 of file elxPreconditionedGradientDescent.h.

◆ FixedImageType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::FixedImageType = typename RegistrationType::FixedImageType

protected

Samplers:

Definition at line 190 of file elxPreconditionedGradientDescent.h.

◆ ImageGridSamplerPointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageGridSamplerPointer = typename ImageGridSamplerType::Pointer

protected

Definition at line 199 of file elxPreconditionedGradientDescent.h.

◆ ImageGridSamplerType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageGridSamplerType = itk::ImageGridSampler<FixedImageType>

protected

Definition at line 198 of file elxPreconditionedGradientDescent.h.

◆ ImageRandomCoordinateSamplerPointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageRandomCoordinateSamplerPointer = typename ImageRandomCoordinateSamplerType::Pointer

protected

Definition at line 197 of file elxPreconditionedGradientDescent.h.

◆ ImageRandomCoordinateSamplerType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageRandomCoordinateSamplerType = itk::ImageRandomCoordinateSampler<FixedImageType>

protected

Definition at line 196 of file elxPreconditionedGradientDescent.h.

◆ ImageRandomSamplerBasePointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageRandomSamplerBasePointer = typename ImageRandomSamplerBaseType::Pointer

protected

Definition at line 195 of file elxPreconditionedGradientDescent.h.

◆ ImageRandomSamplerBaseType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageRandomSamplerBaseType = itk::ImageRandomSamplerBase<FixedImageType>

protected

Definition at line 194 of file elxPreconditionedGradientDescent.h.

◆ ImageSamplerBasePointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageSamplerBasePointer = typename ImageSamplerBaseType::Pointer

protected

Definition at line 193 of file elxPreconditionedGradientDescent.h.

◆ ImageSamplerBaseType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ImageSamplerBaseType = itk::ImageSamplerBase<FixedImageType>

protected

Definition at line 192 of file elxPreconditionedGradientDescent.h.

◆ ITKBaseType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ITKBaseType = typename Superclass2::ITKBaseType

Definition at line 106 of file elxPreconditionedGradientDescent.h.

◆ MovingImageType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::MovingImageType = typename RegistrationType::MovingImageType

protected

Definition at line 191 of file elxPreconditionedGradientDescent.h.

◆ Pointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::Pointer = itk::SmartPointer<Self>

Definition at line 83 of file elxPreconditionedGradientDescent.h.

◆ ProgressCommandPointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ProgressCommandPointer = typename ProgressCommand::Pointer

protected

Definition at line 187 of file elxPreconditionedGradientDescent.h.

◆ ProgressCommandType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::ProgressCommandType = ProgressCommand

protected

Definition at line 186 of file elxPreconditionedGradientDescent.h.

◆ RandomGeneratorPointer

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::RandomGeneratorPointer = typename RandomGeneratorType::Pointer

protected

Definition at line 185 of file elxPreconditionedGradientDescent.h.

◆ RandomGeneratorType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::RandomGeneratorType = itk::Statistics::MersenneTwisterRandomVariateGenerator

protected

Other protected typedefs

Definition at line 184 of file elxPreconditionedGradientDescent.h.

◆ Self

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::Self = PreconditionedGradientDescent

Standard ITK.

Definition at line 80 of file elxPreconditionedGradientDescent.h.

◆ SettingsVectorType

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::SettingsVectorType = typename std::vector<SettingsType>

protected

Definition at line 181 of file elxPreconditionedGradientDescent.h.

◆ Superclass1

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::Superclass1 = AdaptiveStochasticPreconditionedGradientDescentOptimizer

Definition at line 81 of file elxPreconditionedGradientDescent.h.

◆ Superclass2

template<class TElastix >

using elastix::PreconditionedGradientDescent< TElastix >::Superclass2 = OptimizerBase<TElastix>

Definition at line 82 of file elxPreconditionedGradientDescent.h.

Constructor & Destructor Documentation

◆ PreconditionedGradientDescent()

template<class TElastix >

elastix::PreconditionedGradientDescent< TElastix >::PreconditionedGradientDescent ( )

protected

◆ ~PreconditionedGradientDescent()

template<class TElastix >

virtual elastix::PreconditionedGradientDescent< TElastix >::~PreconditionedGradientDescent ( )

inlineprotectedvirtual

Definition at line 202 of file elxPreconditionedGradientDescent.h.

Member Function Documentation

◆ AddRandomPerturbation()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::AddRandomPerturbation	(	const ParametersType &	initialParameters,
		ParametersType &	perturbedParameters,
		double	sigma
	)

protectedvirtual

Helper function that adds a random perturbation delta to the input parameters, with delta ~ sigma * N(0,I). Used by SampleGradients.

◆ AfterEachIteration()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::AfterEachIteration ( )

virtual

Reimplemented from elastix::BaseComponent.

◆ AfterEachResolution()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::AfterEachResolution ( )

virtual

Reimplemented from elastix::BaseComponent.

◆ AfterRegistration()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::AfterRegistration ( )

virtual

Reimplemented from elastix::BaseComponent.

◆ AutomaticParameterEstimation()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::AutomaticParameterEstimation ( )

protectedvirtual

Estimates some reasonable values for the parameters SP_a, SP_alpha (=1), SigmoidMin, SigmoidMax (=1), and SigmoidScale.

◆ BeforeEachResolution()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::BeforeEachResolution ( )

virtual

Reimplemented from elastix::BaseComponent.

◆ BeforeRegistration()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::BeforeRegistration ( )

virtual

Methods invoked by elastix, in which parameters can be set and progress information can be printed.

Reimplemented from elastix::BaseComponent.

◆ elxClassNameMacro()

template<class TElastix >

elastix::PreconditionedGradientDescent< TElastix >::elxClassNameMacro ( "PreconditionedGradientDescent< TElastix >" )

Name of this class. Use this name in the parameter file to select this specific optimizer. example: (Optimizer "PreconditionedGradientDescent")

◆ GetAutomaticParameterEstimation()

template<class TElastix >

virtual bool elastix::PreconditionedGradientDescent< TElastix >::GetAutomaticParameterEstimation ( ) const

virtual

◆ GetClassName()

template<class TElastix >

virtual const char * elastix::PreconditionedGradientDescent< TElastix >::GetClassName ( ) const

virtual

Run-time type information (and related methods).

Reimplemented from itk::AdaptiveStochasticPreconditionedGradientDescentOptimizer.

◆ GetMaximumNumberOfSamplingAttempts()

template<class TElastix >

virtual const unsigned long & elastix::PreconditionedGradientDescent< TElastix >::GetMaximumNumberOfSamplingAttempts ( )

virtual

Get the MaximumNumberOfSamplingAttempts.

◆ GetScaledDerivativeWithExceptionHandling()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::GetScaledDerivativeWithExceptionHandling	(	const ParametersType &	parameters,
		DerivativeType &	derivative
	)

protectedvirtual

Helper function, which calls GetScaledValueAndDerivative and does some exception handling. Used by SampleGradients.

◆ ITK_DISALLOW_COPY_AND_MOVE()

template<class TElastix >

elastix::PreconditionedGradientDescent< TElastix >::ITK_DISALLOW_COPY_AND_MOVE ( PreconditionedGradientDescent< TElastix > )

◆ MetricErrorResponse()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::MetricErrorResponse ( itk::ExceptionObject & err )

virtual

Stop optimization and pass on exception.

◆ New()

template<class TElastix >

static Pointer elastix::PreconditionedGradientDescent< TElastix >::New ( )

static

Method for creation through the object factory.

◆ PrintSettingsVector()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::PrintSettingsVector ( const SettingsVectorType & settings ) const

protectedvirtual

Print the contents of the settings vector to elxout.

◆ ResumeOptimization()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::ResumeOptimization ( )

virtual

Set the SelfHessian as a preconditioning matrix and call Superclass' implementation. Only done when m_PreconditionMatrixSet == false; If automatic gain estimation is desired, then estimate SP_a, SP_alpha SigmoidScale, SigmoidMax, SigmoidMin. After that call Superclass' implementation.

Reimplemented from itk::PreconditionedGradientDescentOptimizer.

◆ SampleGradients()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::SampleGradients	(	const ParametersType &	mu0,
		double &	sigma1,
		double &	sigma2
	)

protectedvirtual

Measure some derivatives, exact and approximated. Returns the sigma1 and sigma2. Needed for the automatic parameter estimation. Gradients are measured at position mu_n, which are generated according to: mu_n - mu_0 ~ N(0, perturbationSigma P ); perturbationSigma = sigma1 = 1/N g_0' P g_0, where g_0 is the exact gradient at mu_0, and N the number of parameters. sigma2 = 1/N sum_n g_n' P g_n / (sum_n n), where g_n is the approximated gradient at mu_0. NB: sigma1 = sqr(sigma_1), sigma2 = sqr(sigma_2), compared to my notes.

◆ SetAutomaticParameterEstimation()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::SetAutomaticParameterEstimation ( bool _arg )

virtual

Set/Get whether automatic parameter estimation is desired. If true, make sure to set the maximum step length.

The following parameters are automatically determined: SP_a = 2*SP_A, SP_alpha (=1), SigmoidMin, SigmoidMax (=1), SigmoidScale. A usually suitable value for SP_A is 20, which is the default setting, if not specified by the user.

◆ SetCurrentPositionPublic()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::SetCurrentPositionPublic ( const ParametersType & param )

inlinevirtual

Add SetCurrentPositionPublic, which calls the protected SetCurrentPosition of the itkAdaptiveStochasticPreconditionedGradientDescentOptimizer class.

Reimplemented from elastix::OptimizerBase< TElastix >.

Definition at line 144 of file elxPreconditionedGradientDescent.h.

◆ SetMaximumNumberOfSamplingAttempts()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::SetMaximumNumberOfSamplingAttempts ( unsigned long _arg )

virtual

Set the MaximumNumberOfSamplingAttempts.

◆ SetSelfHessian()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::SetSelfHessian ( )

protectedvirtual

Get the SelfHessian from the metric and submit as Precondition matrix

◆ StartOptimization()

template<class TElastix >

virtual void elastix::PreconditionedGradientDescent< TElastix >::StartOptimization ( )

virtual

Check if any scales are set, and set the UseScales flag on or off; after that call the superclass' implementation.

Reimplemented from itk::StochasticPreconditionedGradientDescentOptimizer.

Field Documentation

◆ elxOverrideGetSelfMacro

template<class TElastix >

elastix::PreconditionedGradientDescent< TElastix >::elxOverrideGetSelfMacro

private

Definition at line 255 of file elxPreconditionedGradientDescent.h.

◆ m_AutomaticParameterEstimation

template<class TElastix >

bool elastix::PreconditionedGradientDescent< TElastix >::m_AutomaticParameterEstimation

private

Definition at line 265 of file elxPreconditionedGradientDescent.h.

◆ m_AutomaticParameterEstimationDone

template<class TElastix >

bool elastix::PreconditionedGradientDescent< TElastix >::m_AutomaticParameterEstimationDone

private

Definition at line 266 of file elxPreconditionedGradientDescent.h.

◆ m_CurrentNumberOfSamplingAttempts

template<class TElastix >

unsigned long elastix::PreconditionedGradientDescent< TElastix >::m_CurrentNumberOfSamplingAttempts

private

Definition at line 259 of file elxPreconditionedGradientDescent.h.

◆ m_MaximumNumberOfSamplingAttempts

template<class TElastix >

unsigned long elastix::PreconditionedGradientDescent< TElastix >::m_MaximumNumberOfSamplingAttempts

private

Private variables for the sampling attempts.

Definition at line 258 of file elxPreconditionedGradientDescent.h.

◆ m_NumberOfGradientMeasurements

template<class TElastix >

unsigned int elastix::PreconditionedGradientDescent< TElastix >::m_NumberOfGradientMeasurements

protected

Definition at line 207 of file elxPreconditionedGradientDescent.h.

◆ m_NumberOfSamplesForExactGradient

template<class TElastix >

unsigned int elastix::PreconditionedGradientDescent< TElastix >::m_NumberOfSamplesForExactGradient

protected

Definition at line 208 of file elxPreconditionedGradientDescent.h.

◆ m_PreconditionMatrixSet

template<class TElastix >

bool elastix::PreconditionedGradientDescent< TElastix >::m_PreconditionMatrixSet

private

Private variables for self Hessian support.

Definition at line 263 of file elxPreconditionedGradientDescent.h.

◆ m_PreviousErrorAtIteration

template<class TElastix >

unsigned long elastix::PreconditionedGradientDescent< TElastix >::m_PreviousErrorAtIteration

private

Definition at line 260 of file elxPreconditionedGradientDescent.h.

◆ m_RandomGenerator

template<class TElastix >

RandomGeneratorPointer elastix::PreconditionedGradientDescent< TElastix >::m_RandomGenerator

protected

RandomGenerator for AddRandomPerturbation.

Definition at line 212 of file elxPreconditionedGradientDescent.h.

◆ m_SettingsVector

template<class TElastix >

SettingsVectorType elastix::PreconditionedGradientDescent< TElastix >::m_SettingsVector

protected

Variable to store the automatically determined settings for each resolution.

Definition at line 205 of file elxPreconditionedGradientDescent.h.

◆ m_SigmoidScaleFactor

template<class TElastix >

double elastix::PreconditionedGradientDescent< TElastix >::m_SigmoidScaleFactor

protected

Definition at line 209 of file elxPreconditionedGradientDescent.h.

Detailed Description

Data Structures

Public Types

Public Member Functions

Static Public Member Functions

Protected Types

Protected Member Functions

Protected Attributes

Private Attributes

Member Typedef Documentation

◆ ConstPointer

◆ FixedImageType

◆ ImageGridSamplerPointer

◆ ImageGridSamplerType

◆ ImageRandomCoordinateSamplerPointer

◆ ImageRandomCoordinateSamplerType

◆ ImageRandomSamplerBasePointer

◆ ImageRandomSamplerBaseType

◆ ImageSamplerBasePointer

◆ ImageSamplerBaseType

◆ ITKBaseType

◆ MovingImageType

◆ Pointer

◆ ProgressCommandPointer

◆ ProgressCommandType

◆ RandomGeneratorPointer

◆ RandomGeneratorType

◆ Self

◆ SettingsVectorType

◆ Superclass1

◆ Superclass2

Constructor & Destructor Documentation

◆ PreconditionedGradientDescent()

◆ ~PreconditionedGradientDescent()

Member Function Documentation

◆ AddRandomPerturbation()

◆ AfterEachIteration()

◆ AfterEachResolution()

◆ AfterRegistration()

◆ AutomaticParameterEstimation()

◆ BeforeEachResolution()

◆ BeforeRegistration()

◆ elxClassNameMacro()

◆ GetAutomaticParameterEstimation()

◆ GetClassName()

◆ GetMaximumNumberOfSamplingAttempts()

◆ GetScaledDerivativeWithExceptionHandling()

◆ ITK_DISALLOW_COPY_AND_MOVE()

◆ MetricErrorResponse()

◆ New()

◆ PrintSettingsVector()

◆ ResumeOptimization()

◆ SampleGradients()

◆ SetAutomaticParameterEstimation()

◆ SetCurrentPositionPublic()

◆ SetMaximumNumberOfSamplingAttempts()

◆ SetSelfHessian()

◆ StartOptimization()

Field Documentation

◆ elxOverrideGetSelfMacro

◆ m_AutomaticParameterEstimation

◆ m_AutomaticParameterEstimationDone

◆ m_CurrentNumberOfSamplingAttempts

◆ m_MaximumNumberOfSamplingAttempts

◆ m_NumberOfGradientMeasurements

◆ m_NumberOfSamplesForExactGradient

◆ m_PreconditionMatrixSet

◆ m_PreviousErrorAtIteration

◆ m_RandomGenerator

◆ m_SettingsVector

◆ m_SigmoidScaleFactor