stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, ParentT > Class Template Reference

A base class for sums of 'generalised' objective functions, i.e. objective functions for which at least a 'gradient' is defined. More...

#include "stir/recon_buildblock/SumOfGeneralisedObjectiveFunctions.h"

Inheritance diagram for stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, ParentT >:

Public Member Functions
template<typename IterT >
	SumOfGeneralisedObjectiveFunctions (IterT begin, IterT end)
template<typename IterT >
void	set_functions (IterT begin, IterT end)
virtual Succeeded	set_up (shared_ptr< TargetT > const &target_sptr)
	Has to be called before using this object. More...
virtual void	compute_sub_gradient_without_penalty (TargetT &gradient, const TargetT &current_estimate, const int subset_num)
	This computes the gradient of the unregularised objective function at the current_estimate. More...
virtual double	actual_compute_objective_function_without_penalty (const TargetT &current_estimate, const int subset_num)
virtual int	set_num_subsets (const int num_subsets)
	Attempts to change the number of subsets. More...

Protected Types
typedef std::vector< ObjFuncT >	_functions_type
typedef _functions_type::iterator	_functions_iterator_type
typedef _functions_type::const_iterator	_functions_const_iterator_type

Protected Member Functions
virtual bool	actual_subsets_are_approximately_balanced (std::string &warning_message) const
	Implementation of function that checks subset balancing. More...

Protected Attributes
_functions_type	_functions

Detailed Description

template<typename ObjFuncT, typename TargetT, typename ParentT = GeneralisedObjectiveFunction<TargetT>>
class stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, ParentT >

A base class for sums of 'generalised' objective functions, i.e. objective functions for which at least a 'gradient' is defined.

Todo:

document why use of ParentT template

doc subsets

Member Function Documentation

set_up()

template<typename ObjFuncT , typename TargetT, typename Parent >

Succeeded stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, Parent >::set_up ( shared_ptr< TargetT > const &  target_sptr )

inlinevirtual

Has to be called before using this object.

Will call set_up() for all terms in the sum, but will stop as soon as one set_up() fails.

compute_sub_gradient_without_penalty()

template<typename ObjFuncT , typename TargetT, typename Parent >

void stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, Parent >::compute_sub_gradient_without_penalty ( TargetT &  gradient, const TargetT &  current_estimate, const int  subset_num  )

inlinevirtual

This computes the gradient of the unregularised objective function at the current_estimate.

It is computed as the sum of the subgradients for each term, depending on the subset scheme.

Referenced by stir::SumOfGeneralisedObjectiveFunctions< PoissonLogLikelihoodWithLinearModelForMeanAndProjData< TargetT >, TargetT, PoissonLogLikelihoodWithLinearModelForMean< TargetT > >::compute_sub_gradient_without_penalty().

set_num_subsets()

template<typename ObjFuncT , typename TargetT , typename Parent >

int stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, Parent >::set_num_subsets ( const int  num_subsets )

inlinevirtual

Attempts to change the number of subsets.

Returns

The number of subsets that will be used later, which is not guaranteed to be what you asked for.

Reimplemented in stir::PoissonLogLikelihoodWithLinearModelForMeanAndGatedProjDataWithMotion< TargetT >.

actual_subsets_are_approximately_balanced()

template<typename ObjFuncT , typename TargetT , typename Parent >

bool stir::SumOfGeneralisedObjectiveFunctions< ObjFuncT, TargetT, Parent >::actual_subsets_are_approximately_balanced ( std::string &  warning_message ) const

inlineprotectedvirtual

Implementation of function that checks subset balancing.

Todo:

doc subset

Reimplemented in stir::PoissonLogLikelihoodWithLinearModelForMeanAndGatedProjDataWithMotion< TargetT >.

---

The documentation for this class was generated from the following files:

• /home/sirfuser/devel/STIRdistrib/STIR/src/include/stir/recon_buildblock/SumOfGeneralisedObjectiveFunctions.h
• /home/sirfuser/devel/STIRdistrib/STIR/src/include/stir/recon_buildblock/SumOfGeneralisedObjectiveFunctions.inl