StratifiedParticleResampler Class Reference

Inheritance diagram for StratifiedParticleResampler:

[legend]

List of all members.

---

Detailed Description

Stratified particle resampler.

Author:

Lawrence Murray <lawrence@indii.org>

Version:

Rev

Date:

Date

Produces a new approximation of a weighted sample set using a set of equally weighted samples, possibly with duplication. Based on the deterministic stratified resampling scheme of (Kitagawa 1996), without sorting.

References

Kitagawa, G. Monte Carlo Filter and Smoother for Non-Gaussian Nonlinear State Space Models. Journal of Computational and Graphical Statistics, 1996, 5, 1-25.

Definition at line 28 of file StratifiedParticleResampler.hpp.

Public Member Functions
	StratifiedParticleResampler (const unsigned int P=0)
	Constructor.
virtual	~StratifiedParticleResampler ()
	Destructor.
void	setNumParticles (const unsigned int P=0)
	Set the number of particles to resample.
virtual indii::ml::aux::DiracMixturePdf	resample (indii::ml::aux::DiracMixturePdf &p)
	Resample the distribution.

---

Constructor & Destructor Documentation

StratifiedParticleResampler

(

const unsigned int

P = 0

)

Constructor.

Parameters:

P

Number of particles to resample from each distribution supplied to resample(). If zero, will resample the same number of particles as the distribution supplied.

Definition at line 11 of file StratifiedParticleResampler.cpp.

~StratifiedParticleResampler

(

)

[virtual]

Destructor.

Definition at line 16 of file StratifiedParticleResampler.cpp.

---

Member Function Documentation

void setNumParticles

(

const unsigned int

P = 0

)

Set the number of particles to resample.

Parameters:

P

Number of particles to resample from each distribution supplied to resample(). If zero, will resample the same number of particles as the distribution supplied.

Definition at line 20 of file StratifiedParticleResampler.cpp.

aux::DiracMixturePdf resample

(

indii::ml::aux::DiracMixturePdf &

p

)

[virtual]

Resample the distribution.

This produces a new approximation of the same distribution using a set of equally weighted sample points. Sample points are selected using the deterministic resampling method given in the appendix of Kitagawa (1996).

Returns:

The resampled distribution.

Implements ParticleResampler.

Definition at line 24 of file StratifiedParticleResampler.cpp.

---