#include <SamplerMethod.h>

Inheritance diagram for SamplerMethod:

[legend]

Collaboration diagram for SamplerMethod:

Public Types
Motion Planning Types
typedef MPBaseObject::GroupCfgType	GroupCfgType

Local Types
typedef std::vector< Cfg >::iterator	InputIterator

typedef std::back_insert_iterator< std::vector< Cfg > >	OutputIterator

typedef std::vector< GroupCfgType >::iterator	GroupInputIterator

typedef std::back_insert_iterator< std::vector< GroupCfgType > >	GroupOutputIterator

typedef std::map< Robot , const Boundary >	BoundaryMap
	A map from robots to sampling boundaries. More...

Public Types inherited from MPBaseObject
typedef DefaultWeight< Cfg >	WeightType

typedef GenericStateGraph< Cfg, WeightType >	RoadmapType

typedef GroupCfg< RoadmapType >	GroupCfgType

typedef GroupLocalPlan< RoadmapType >	GroupWeightType

typedef GroupRoadmap< GroupCfgType, GroupWeightType >	GroupRoadmapType

Public Member Functions
Construction
	SamplerMethod ()=default

	SamplerMethod (XMLNode &_node)

virtual	~SamplerMethod ()=0

Individual Configuration Sampling
virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary *const _boundary, OutputIterator _valid, OutputIterator _invalid)

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary *const _boundary, OutputIterator _valid)
	This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts. More...

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary const _robotBoundary, const Boundary const _eeBoundary, OutputIterator _valid, OutputIterator _invalid)

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary const _robotBoundary, const Boundary const _eeBoundary, OutputIterator _valid)

virtual void	Filter (InputIterator _first, InputIterator _last, size_t _maxAttempts, const Boundary *const _boundary, OutputIterator _valid, OutputIterator _invalid)

void	Filter (InputIterator _first, InputIterator _last, size_t _maxAttempts, const Boundary *const _boundary, OutputIterator _valid)

Group Configuration Sampling
virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary *const _boundary, GroupOutputIterator _valid, GroupOutputIterator _invalid)

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const Boundary *const _boundary, GroupOutputIterator _valid)

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const BoundaryMap &_boundary, GroupOutputIterator _valid, GroupOutputIterator _invalid)

virtual void	Sample (size_t _numNodes, size_t _maxAttempts, const BoundaryMap &_boundary, GroupOutputIterator _valid)

virtual void	Filter (GroupInputIterator _first, GroupInputIterator _last, size_t _maxAttempts, const Boundary *const _boundary, GroupOutputIterator _valid, GroupOutputIterator _invalid)

void	Filter (GroupInputIterator _first, GroupInputIterator _last, size_t _maxAttempts, const Boundary *const _boundary, GroupOutputIterator _valid)
	This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts. More...

Public Member Functions inherited from MPBaseObject
	MPBaseObject (const std::string &_label="", const std::string &_name="", bool _debug=false)

	MPBaseObject (XMLNode &_node)

virtual	~MPBaseObject ()

virtual void	Print (std::ostream &_os) const

virtual void	Initialize ()

const std::string &	GetName () const
	Get the class name for this object. More...

const std::string &	GetLabel () const
	Get the unique label for this object. More...

std::string	GetNameAndLabel () const
	Get the unique string identifier for this object "m_name::m_label". More...

void	SetLabel (const std::string &)
	Set the unique label for this object. More...

void	SetMPLibrary (MPLibrary *) noexcept
	Set the owning MPLibrary. More...

MPLibrary *	GetMPLibrary () const noexcept
	Get the owning MPLibrary. More...

bool	IsRunning () const noexcept
	Check the library's running flag. More...

MPProblem *	GetMPProblem () const noexcept
	Get the library's current MPProblem. More...

Environment *	GetEnvironment () const noexcept
	Get the current environment. More...

MPTask *	GetTask () const noexcept
	Get the current task. More...

GroupTask *	GetGroupTask () const noexcept
	Get the current group task. More...

MPSolutionType *	GetMPSolution () const noexcept

RoadmapType *	GetRoadmap (Robot *const _r=nullptr) const noexcept
	Get the current free-space roadmap. More...

GroupRoadmapType *	GetGroupRoadmap (RobotGroup *const _g=nullptr) const noexcept
	Get the current free-space group roadmap. More...

RoadmapType *	GetBlockRoadmap (Robot *const _r=nullptr) const noexcept
	Get the current obstacle-space roadmap. More...

Path *	GetPath (Robot *const _r=nullptr) const noexcept

GroupPath *	GetGroupPath (RobotGroup *const _g=nullptr) const noexcept
	Get the current best group path. More...

StatClass *	GetStatClass () const noexcept
	Get the current StatClass. More...

LocalObstacleMap *	GetLocalObstacleMap () const noexcept
	Get the local obstacle map. More...

Protected Member Functions
Sampler Rule
The default implementations of these throw exceptions so that derived classes can avoid implementing unsupported functions (as opposed to making these pure virtual).
virtual bool	Sampler (Cfg &_cfg, const Boundary *const _boundary, std::vector< Cfg > &_valid, std::vector< Cfg > &_invalid)

virtual bool	Sampler (Cfg &_cfg, const Boundary const _robotBoundary, const Boundary const _eeBoundary, std::vector< Cfg > &_valid, std::vector< Cfg > &_invalid)

virtual bool	Sampler (GroupCfgType &_cfg, const Boundary *const _boundary, std::vector< GroupCfgType > &_valid, std::vector< GroupCfgType > &_invalid)

virtual bool	Sampler (GroupCfgType &_cfg, const BoundaryMap &_boundaryMap, std::vector< GroupCfgType > &_valid, std::vector< GroupCfgType > &_invalid)

Protected Member Functions inherited from MPBaseObject
void	SetName (const std::string &_s)

const std::string &	GetBaseFilename () const

Friends
class	MixSampler

Additional Inherited Members
Protected Attributes inherited from MPBaseObject
bool	m_debug
	Print debug info? More...

Detailed Description

Base algorithm abstraction for Samplers.

SamplerMethod has three sets of important functions. The first two are the various public methods in the base class, Sample and Filter, and third is the private virtual function which the derived classes overload, Sampler.

Sample is called in various ways but they break down into two forms: desired number and input configurations. When specifying a desired number n of configurations the sampler attempts a attempts per desired sample. The output is placed on an output iterator.

@usage

size_t num, attempts;
Boundary* bounds;
std::vector<Cfg> valid;
auto s = this->GetSampler(m_sLabel);
s->Sample(num, attempts, bounds, std::back_inserter(valid));

Filter sends a list of input configurations to apply the sampler rule to.

@usage

std::vector<Cfg> input, valid;
size_t attempts;
Boundary* bounds;
auto s = this->GetSampler(m_sLabel);
s->Filter(input.begin(), input.end(), attempts, bounds,
    std::back_inserter(valid));

Todo:: The present implementation of Sample temporarily saves invalid configurations whether we will use them or not. Avoiding this extraneous retention will likely have good performance benefits in sampling-intensive applications such as proteins and manipulators where the success rate is very low.

Todo:: The present layout of this class is confusing since the base method is actually generating uniform random samples which are filtered by the derived class. This does not work for many derived classes as evidenced by their overriding of Sample rather than Sampler: we should rework the design so that derived classes generate and filter their own samples. The base class can then implement its functions in terms of the generate and filter helpers.

Examples: Samplers_UseCase.cpp.

Member Typedef Documentation

◆ BoundaryMap

typedef std::map<Robot*, const Boundary*> SamplerMethod::BoundaryMap

A map from robots to sampling boundaries.

◆ GroupCfgType

typedef MPBaseObject::GroupCfgType SamplerMethod::GroupCfgType

◆ GroupInputIterator

typedef std::vector<GroupCfgType>::iterator SamplerMethod::GroupInputIterator

◆ GroupOutputIterator

typedef std::back_insert_iterator<std::vector<GroupCfgType> > SamplerMethod::GroupOutputIterator

◆ InputIterator

typedef std::vector<Cfg>::iterator SamplerMethod::InputIterator

◆ OutputIterator

typedef std::back_insert_iterator<std::vector<Cfg> > SamplerMethod::OutputIterator

Constructor & Destructor Documentation

◆ SamplerMethod() [1/2]

SamplerMethod::SamplerMethod ( )

default

◆ SamplerMethod() [2/2]

SamplerMethod::SamplerMethod ( XMLNode & _node )

◆ ~SamplerMethod()

SamplerMethod::~SamplerMethod ( )

pure virtualdefault

Member Function Documentation

◆ Filter() [1/4]

void SamplerMethod::Filter	(	GroupInputIterator	_first,
		GroupInputIterator	_last,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		GroupOutputIterator	_valid
	)

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Filter() [2/4]

void SamplerMethod::Filter	(	GroupInputIterator	_first,
		GroupInputIterator	_last,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		GroupOutputIterator	_valid,
		GroupOutputIterator	_invalid
	)

virtual

Apply the sampler rule to a set of existing configurations. The output will generally be a filtered or perturbed version of the input set.

Parameters

_first	An iterator to the beginning of a list of input configurations.
_last	An iterator to the end of a list of input configurations.
_maxAttempts	The maximum number of attempts to successfully apply the sampler rule to each input.
_boundary	The sampling boundary to use.
_valid	An iterator to storage for the output configurations.
_invalid	An (optional) iterator to storage for failed attempts.

◆ Filter() [3/4]

void SamplerMethod::Filter	(	InputIterator	_first,
		InputIterator	_last,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		OutputIterator	_valid
	)

This version does not return invalid samples. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Filter() [4/4]

void SamplerMethod::Filter	(	InputIterator	_first,
		InputIterator	_last,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		OutputIterator	_valid,
		OutputIterator	_invalid
	)

virtual

Apply the sampler rule to a set of existing configurations. The output will generally be a filtered or perturbed version of the input set.

Parameters

_first	An iterator to the beginning of a list of input configurations.
_last	An iterator to the end of a list of input configurations.
_maxAttempts	The maximum number of attempts to successfully apply the sampler rule to each input.
_boundary	The sampling boundary to use.
_valid	An iterator to storage for the output configurations.
_invalid	An (optional) iterator to storage for failed attempts.

Examples: Samplers_UseCase.cpp.

◆ Sample() [1/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		GroupOutputIterator	_valid
	)

virtual

This version does not return invalid samples. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Sample() [2/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		GroupOutputIterator	_valid,
		GroupOutputIterator	_invalid
	)

virtual

Try to sample a set number of new configurations from a single boundary.

Parameters

_numNodes	The number of samples desired.
_maxAttempts	The maximum number of attempts for each sample.
_boundary	The boundary to sample from.
_valid	An iterator to storage for the new configurations.
_invalid	An (optional) iterator to storage for failed attempts.

◆ Sample() [3/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		OutputIterator	_valid
	)

virtual

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Sample() [4/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_boundary,
		OutputIterator	_valid,
		OutputIterator	_invalid
	)

virtual

Try to sample a set number of new configurations from a given boundary.

Parameters

_numNodes	The number of samples desired.
_maxAttempts	The maximum number of attempts for each sample.
_boundary	The boundary to sample from.
_valid	An iterator to storage for the new configurations.
_invalid	An (optional) iterator to storage for failed attempts.

Examples: Samplers_UseCase.cpp.

◆ Sample() [5/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_robotBoundary,
		const Boundary *const	_eeBoundary,
		OutputIterator	_valid
	)

virtual

This version does not return invalid samples. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Sample() [6/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const Boundary *const	_robotBoundary,
		const Boundary *const	_eeBoundary,
		OutputIterator	_valid,
		OutputIterator	_invalid
	)

virtual

Try to sample a set number of new configurations from a given boundary (for the entire robot), and additionally satisfying a constraint boundary for the end-effector (robot is presumed to have only one EE).

Parameters

_robotBoundary	The boundary for the entire robot.
_eeBoundary	The boundary for the end-effector.

◆ Sample() [7/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const BoundaryMap &	_boundary,
		GroupOutputIterator	_valid
	)

virtual

This version does not return invalid samples. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Sample() [8/8]

void SamplerMethod::Sample	(	size_t	_numNodes,
		size_t	_maxAttempts,
		const BoundaryMap &	_boundary,
		GroupOutputIterator	_valid,
		GroupOutputIterator	_invalid
	)

virtual

Try to sample a set number of new configurations from a boundary for each robot.

Parameters

_numNodes	The number of samples desired.
_maxAttempts	The maximum number of attempts for each sample.
_boundaryMap	A map from robot to sampling boundary. Any robots which are not found in the map will use the environment boundary.
_valid	An iterator to storage for the new configurations.
_invalid	An (optional) iterator to storage for failed attempts.

◆ Sampler() [1/4]

bool SamplerMethod::Sampler	(	Cfg &	_cfg,
		const Boundary *const	_boundary,
		std::vector< Cfg > &	_valid,
		std::vector< Cfg > &	_invalid
	)

protectedvirtual

Takes a single input configuration and applies the sampler rule to generate one or more output configurations.

Parameters

_cfg	The input configuration.
_boundary	The sampling boundary.
_valid	The resulting output configurations.
_invalid	The (optional) return for failed attempts.

Returns: True if a valid configuration was generated, false otherwise.

Reimplemented in BridgeTestSampler, MixSampler, GaussianSampler, UniformRandomSampler, UniformObstacleBasedSampler, and ObstacleBasedSampler.

◆ Sampler() [2/4]

bool SamplerMethod::Sampler	(	Cfg &	_cfg,
		const Boundary *const	_robotBoundary,
		const Boundary *const	_eeBoundary,
		std::vector< Cfg > &	_valid,
		std::vector< Cfg > &	_invalid
	)

protectedvirtual

This version also specifies a boundary for the end-effector. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

◆ Sampler() [3/4]

bool SamplerMethod::Sampler	(	GroupCfgType &	_cfg,
		const Boundary *const	_boundary,
		std::vector< GroupCfgType > &	_valid,
		std::vector< GroupCfgType > &	_invalid
	)

protectedvirtual

Takes a single input configuration and applies the sampler rule to generate one or more output configurations.

Parameters

_cfg	The input group configuration.
_boundary	The sampling boundary.
_valid	The resulting output configurations.
_invalid	The (optional) return for failed attempts.

Returns: True if a valid configuration was generated, false otherwise.

Reimplemented in BridgeTestSampler, UniformRandomSampler, UniformObstacleBasedSampler, and ObstacleBasedSampler.

◆ Sampler() [4/4]

bool SamplerMethod::Sampler	(	GroupCfgType &	_cfg,
		const BoundaryMap &	_boundaryMap,
		std::vector< GroupCfgType > &	_valid,
		std::vector< GroupCfgType > &	_invalid
	)

protectedvirtual

This version specifies a (possibly different) boundary for each robot. Robots which are not in the boundary map will use the environment boundary. This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

Reimplemented in UniformRandomSampler, BridgeTestSampler, and ObstacleBasedSampler.

Friends And Related Function Documentation

◆ MixSampler

friend class MixSampler

friend

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

/opt/actions-runner/_work/open-ppl/open-ppl/src/MPLibrary/Samplers/SamplerMethod.h
/opt/actions-runner/_work/open-ppl/open-ppl/src/MPLibrary/Samplers/SamplerMethod.cpp

Public Types

Public Member Functions

Protected Member Functions

Friends

Additional Inherited Members

Detailed Description

Member Typedef Documentation

◆ BoundaryMap

◆ GroupCfgType

◆ GroupInputIterator

◆ GroupOutputIterator

◆ InputIterator

◆ OutputIterator

Constructor & Destructor Documentation

◆ SamplerMethod() [1/2]

◆ SamplerMethod() [2/2]

◆ ~SamplerMethod()

Member Function Documentation

◆ Filter() [1/4]

◆ Filter() [2/4]

◆ Filter() [3/4]

◆ Filter() [4/4]

◆ Sample() [1/8]

◆ Sample() [2/8]

◆ Sample() [3/8]

◆ Sample() [4/8]

◆ Sample() [5/8]

◆ Sample() [6/8]

◆ Sample() [7/8]

◆ Sample() [8/8]

◆ Sampler() [1/4]

◆ Sampler() [2/4]

◆ Sampler() [3/4]

◆ Sampler() [4/4]

Friends And Related Function Documentation

◆ MixSampler