Class for computing variable projection costs and derivatives. Right now, it is only available for Gaussian noise. By the way, this is not thread safe at all so be sure to intantiate one per thread. More...

#include <animaGaussianMCMVariableProjectionCost.h>

Inheritance diagram for anima::GaussianMCMVariableProjectionCost:

Collaboration diagram for anima::GaussianMCMVariableProjectionCost:

Public Types
typedef itk::SmartPointer< const Self >	ConstPointer

using	LECalculatorPointer = LECalculatorType::Pointer

using	LECalculatorType = anima::LogEuclideanTensorCalculator< double >

typedef itk::SmartPointer< Self >	Pointer

typedef GaussianMCMVariableProjectionCost	Self

typedef anima::BaseMCMCost	Superclass

Public Types inherited from anima::BaseMCMCost
typedef itk::SmartPointer< const Self >	ConstPointer

typedef itk::Array2D< double >	DerivativeMatrixType

typedef itk::Array< double >	DerivativeType

typedef MCMType::ListType	ListType

typedef MCMType::Pointer	MCMPointer

typedef anima::MultiCompartmentModel	MCMType

typedef itk::Array< double >	MeasureType

typedef itk::OptimizerParameters< double >	ParametersType

typedef itk::SmartPointer< Self >	Pointer

typedef BaseMCMCost	Self

typedef itk::LightObject	Superclass

typedef MCMType::Vector3DType	Vector3DType

Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const

double	GetCurrentCostValue () ITK_OVERRIDE
	For the current set of parameters, compute the cost function value, requires GetValues to be called first. More...

void	GetCurrentDerivative (DerivativeMatrixType &derivativeMatrix, DerivativeType &derivative) ITK_OVERRIDE
	Get cost function derivatives from a derivative matrix obtained from GetDerivativeMatrix. More...

void	GetDerivativeMatrix (const ParametersType &parameters, DerivativeMatrixType &derivative) ITK_OVERRIDE
	Get residual derivatives for a given set of parameters, returns a matrix of residuals derivatives. More...

virtual const char *	GetNameOfClass () const

std::vector< double > &	GetOptimalWeights ()

MeasureType	GetValues (const ParametersType &parameters) ITK_OVERRIDE
	Get residual values for a given set of parameters, returns a vector of residuals. More...

Public Member Functions inherited from anima::BaseMCMCost
MCMPointer &	GetMCMStructure ()

unsigned int	GetNumberOfObservations () const

unsigned int	GetNumberOfParameters () const
	Returns number of optimized parameters. More...

virtual double	GetSigmaSquare ()

void	SetBigDelta (double val)

void	SetGradients (std::vector< Vector3DType > &value)

void	SetGradientStrengths (ListType &value)

void	SetMCMStructure (MCMType *model)

void	SetObservedSignals (ListType &value)

void	SetSmallDelta (double val)

Static Public Member Functions
static Pointer	New ()

Protected Member Functions
bool	CheckBoundaryConditions ()

	GaussianMCMVariableProjectionCost ()

void	PrepareDataForDerivative ()

void	PrepareDataForLLS ()

void	SolveLinearLeastSquares ()
	Computes maximum likelihood estimates of weights. More...

virtual	~GaussianMCMVariableProjectionCost () ITK_OVERRIDE

Protected Member Functions inherited from anima::BaseMCMCost
	BaseMCMCost ()

virtual	~BaseMCMCost ()

Additional Inherited Members
Protected Attributes inherited from anima::BaseMCMCost
double	m_BigDelta

std::vector< Vector3DType >	m_Gradients

ListType	m_GradientStrengths

MCMPointer	m_MCMStructure

ListType	m_ObservedSignals

std::vector< double >	m_PredictedSignals

double	m_SigmaSquare

double	m_SmallDelta

ListType	m_TestedParameters

Detailed Description

Class for computing variable projection costs and derivatives. Right now, it is only available for Gaussian noise. By the way, this is not thread safe at all so be sure to intantiate one per thread.

Definition at line 20 of file animaGaussianMCMVariableProjectionCost.h.