Detailed Description

This class acts as a common interface to the different interfaces (see CParticleFilter::TParticleFilterAlgorithm) any bayes::CParticleFilterCapable class can implement: it is the invoker of particle filter algorithms.

The particle filter is executed on a probability density function (PDF) described by a CParticleFilterCapable object, passed in the constructor or alternatively through the CParticleFilter::executeOn method.

For a complete example and further details, see the Particle Filter tutorial.

The basic SIR algorithm (pfStandardProposal) consists of:

Execute a prediction with the given "action".
Update the weights of the particles using the likelihood of the "observation".
Normalize weights.
Perform resampling if the ESS is below the threshold options.BETA.

See also: mrpt::poses::CPoseParticlesPDF

Definition at line 40 of file CParticleFilter.h.

#include <mrpt/bayes/CParticleFilter.h>

Inheritance diagram for mrpt::bayes::CParticleFilter:

Classes
struct	TParticleFilterOptions
	The configuration of a particle filter. More...

struct	TParticleFilterStats
	Statistics for being returned from the "execute" method. More...

Public Types
enum	TParticleFilterAlgorithm { pfStandardProposal = 0 , pfAuxiliaryPFStandard , pfOptimalProposal , pfAuxiliaryPFOptimal }
	Defines different types of particle filter algorithms. More...

enum	TParticleResamplingAlgorithm { prMultinomial = 0 , prResidual , prStratified , prSystematic }
	Defines the different resampling algorithms. More...

Public Member Functions
	CParticleFilter ()
	Default constructor. More...

virtual	~CParticleFilter ()

void	executeOn (CParticleFilterCapable &obj, const mrpt::obs::CActionCollection action, const mrpt::obs::CSensoryFrame observation, TParticleFilterStats *stats=NULL)
	Executes a complete prediction + update step of the selected particle filtering algorithm. More...

Static Public Member Functions
static void	printf_debug (const char *frmt,...)
	Sends a formated text to "debugOut" if not NULL, or to cout otherwise. More...

Public Attributes
CParticleFilter::TParticleFilterOptions	m_options
	The options to be used in the PF, must be set before executing any step of the particle filter. More...

Member Enumeration Documentation

◆ TParticleFilterAlgorithm

enum mrpt::bayes::CParticleFilter::TParticleFilterAlgorithm

Defines different types of particle filter algorithms.

The defined SIR implementations are:

pfStandardProposal: Standard proposal distribution + weights according to likelihood function.
pfAuxiliaryPFStandard: An auxiliary PF using the standard proposal distribution.
pfOptimalProposal: Use the optimal proposal distribution (where available!, usually this will perform approximations)
pfAuxiliaryPFOptimal: Use the optimal proposal and a auxiliary particle filter (see paper).

See the theoretical discussion in resampling schemes.

Enumerator
pfStandardProposal
pfAuxiliaryPFStandard
pfOptimalProposal
pfAuxiliaryPFOptimal

Definition at line 53 of file CParticleFilter.h.

◆ TParticleResamplingAlgorithm

enum mrpt::bayes::CParticleFilter::TParticleResamplingAlgorithm

Defines the different resampling algorithms.

The implemented resampling methods are:

prMultinomial (Default): Uses standard select with replacement (draws M random uniform numbers)
prResidual: The residual or "remainder" method.
prStratified: The stratified resampling, where a uniform sample is drawn for each of M subdivisions of the range (0,1].
prSystematic: A single uniform sample is drawn in the range (0,1/M].

See the theoretical discussion in resampling schemes.

Enumerator
prMultinomial
prResidual
prStratified
prSystematic

Definition at line 70 of file CParticleFilter.h.

Constructor & Destructor Documentation

◆ CParticleFilter()

mrpt::bayes::CParticleFilter::CParticleFilter ( )

Default constructor.

After creating the PF object, set the options in CParticleFilter::m_options, then execute steps through CParticleFilter::executeOn.

◆ ~CParticleFilter()

virtual mrpt::bayes::CParticleFilter::~CParticleFilter ( )

inlinevirtual

Definition at line 129 of file CParticleFilter.h.

Member Function Documentation

◆ executeOn()

void mrpt::bayes::CParticleFilter::executeOn	(	CParticleFilterCapable &	obj,
		const mrpt::obs::CActionCollection *	action,
		const mrpt::obs::CSensoryFrame *	observation,
		TParticleFilterStats *	stats = `NULL`
	)

Executes a complete prediction + update step of the selected particle filtering algorithm.

The member CParticleFilter::m_options must be set before calling this to settle the algorithm parameters.

Parameters

obj	The object representing the probability distribution function (PDF) which apply the particle filter algorithm to.
action	A pointer to an action in the form of a CActionCollection, or NULL if there is no action.
observation	A pointer to observations in the form of a CSensoryFrame, or NULL if there is no observation.
stats	An output structure for gathering statistics of the particle filter execution, or set to NULL if you do not need it (see CParticleFilter::TParticleFilterStats).

See also: CParticleFilterCapable, executeOn

◆ printf_debug()

static void mrpt::utils::CDebugOutputCapable::printf_debug	(	const char *	frmt,
			...
	)

staticinherited

Sends a formated text to "debugOut" if not NULL, or to cout otherwise.

Referenced by mrpt::math::CLevenbergMarquardtTempl< VECTORTYPE, USERPARAM >::execute().

Member Data Documentation

◆ m_options

CParticleFilter::TParticleFilterOptions mrpt::bayes::CParticleFilter::m_options

The options to be used in the PF, must be set before executing any step of the particle filter.

Definition at line 150 of file CParticleFilter.h.