A base class for the bootstrap particle filter. More...

#include <bootstrap_filter.h>

Inheritance diagram for pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >:

Collaboration diagram for pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >:

Public Member Functions
	BSFilter (const unsigned int &rs=1)
	The constructor. More...

virtual	~BSFilter ()
	The (virtual) destructor.

float_t	getLogCondLike () const
	Returns the most recent (log-) conditional likelihood. More...

void	filter (const osv &data, const std::vector< std::function< const Mat(const ssv &)> > &fs=std::vector< std::function< const Mat(const ssv &)> >())
	updates filtering distribution on a new datapoint. Optionally stores expectations of functionals. More...

auto	getExpectations () const -> std::vector< Mat >
	return all stored expectations (taken with respect to $p(x_t\|y_{1:t})$ More...

virtual float_t	logMuEv (const ssv &x1)=0
	Calculate muEv or logmuEv. More...

virtual ssv	q1Samp (const osv &y1)=0
	Samples from time 1 proposal. More...

virtual float_t	logQ1Ev (const ssv &x1, const osv &y1)=0
	Calculate q1Ev or log q1Ev. More...

virtual float_t	logGEv (const osv &yt, const ssv &xt)=0
	Calculate gEv or logGEv. More...

virtual ssv	fSamp (const ssv &xtm1)=0
	Sample from the state transition distribution. More...

Public Member Functions inherited from pf::bases::pf_base< float_t, dimy, dimx >
virtual void	filter (const obs_sized_vec &data, const func_vec &fs=func_vec())=0
	the filtering function that must be defined More...

virtual float_t	getLogCondLike () const=0
	the getter method that must be defined (for conditional log-likelihood) More...

virtual	~pf_base ()
	virtual destructor

Protected Attributes
arrayStates	m_particles
	particle samples

arrayFloat	m_logUnNormWeights
	particle unnormalized weights

unsigned int	m_now
	time point

float_t	m_logLastCondLike
	log p(y_t\|y_{1:t-1}) or log p(y1)

resamp_t	m_resampler
	resampler object

std::vector< Mat >	m_expectations
	expectations E[h(x_t) \| y_{1:t}] for user defined "h"s

unsigned int	m_resampSched
	resampling schedule (e.g. resample every __ time points)

Private Types
using	ssv = Eigen::Matrix< float_t, dimx, 1 >

using	osv = Eigen::Matrix< float_t, dimy, 1 >

using	Mat = Eigen::Matrix< float_t, Eigen::Dynamic, Eigen::Dynamic >

using	arrayStates = std::array< ssv, nparts >

using	arrayFloat = std::array< float_t, nparts >

Additional Inherited Members
Public Types inherited from pf::bases::pf_base< float_t, dimy, dimx >
using	float_type = float_t

using	obs_sized_vec = Eigen::Matrix< float_t, dimobs, 1 >

using	state_sized_vec = Eigen::Matrix< float_t, dimstate, 1 >

using	dynamic_matrix = Eigen::Matrix< float_t, Eigen::Dynamic, Eigen::Dynamic >

using	func = std::function< const dynamic_matrix(const state_sized_vec &)>

using	func_vec = std::vector< func >

Static Public Attributes inherited from pf::bases::pf_base< float_t, dimy, dimx >
static constexpr unsigned int	dim_obs

static constexpr unsigned int	dim_state

Detailed Description

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t, typename float_t, bool debug = false>
class pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >

A base class for the bootstrap particle filter.

Author: taylor

Member Typedef Documentation

◆ arrayFloat

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

using pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::arrayFloat = std::array<float_t, nparts>

private

type alias for array of floating points

◆ arrayStates

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

using pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::arrayStates = std::array<ssv, nparts>

private

type alias for linear algebra stuff

◆ Mat

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

using pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::Mat = Eigen::Matrix<float_t, Eigen::Dynamic, Eigen::Dynamic>

private

type alias for dynamically sized matrix

◆ osv

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

using pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::osv = Eigen::Matrix<float_t, dimy, 1>

private

"obs size vector" type alias for linear algebra stuff

◆ ssv

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

using pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::ssv = Eigen::Matrix<float_t, dimx, 1>

private

"state size vector" type alias for linear algebra stuff

Constructor & Destructor Documentation

◆ BSFilter()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug>

BSFilter::BSFilter ( const unsigned int & rs = 1 )

The constructor.

Parameters

rs	the resampling schedule (e.g. every rs time point)

Member Function Documentation

◆ filter()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug>

void BSFilter::filter	(	const osv &	data,
		const std::vector< std::function< const Mat(const ssv &)> > &	fs = `std::vector<std::function<const Mat(const ssv&)> >()`
	)

updates filtering distribution on a new datapoint. Optionally stores expectations of functionals.

Parameters

data	the most recent data point
fs	a vector of functions if you want to calculate expectations.

◆ fSamp()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

virtual ssv pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::fSamp ( const ssv & xtm1 )

pure virtual

Sample from the state transition distribution.

Parameters

xtm1	is a const Vec& describing the time t-1 state

Returns: the sample as a Vec

◆ getExpectations()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug>

auto BSFilter::getExpectations

return all stored expectations (taken with respect to $p(x_t|y_{1:t})$

Returns: return a std::vector<Mat> of expectations. How many depends on how many callbacks you gave to

◆ getLogCondLike()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug>

float_t BSFilter::getLogCondLike

Returns the most recent (log-) conditional likelihood.

Returns: log p(y_t | y_{1:t-1})

◆ logGEv()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

virtual float_t pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::logGEv	(	const osv &	yt,
		const ssv &	xt
	)

pure virtual

Calculate gEv or logGEv.

Parameters

yt	is a const Vec& describing the time t datum
xt	is a const Vec& describing the time t state

Returns: the density or log-density evaluation

◆ logMuEv()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

virtual float_t pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::logMuEv ( const ssv & x1 )

pure virtual

Calculate muEv or logmuEv.

Parameters

x1	is a const Vec& describing the state sample

Returns: the density or log-density evaluation

◆ logQ1Ev()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

virtual float_t pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::logQ1Ev	(	const ssv &	x1,
		const osv &	y1
	)

pure virtual

Calculate q1Ev or log q1Ev.

Parameters

x1	is a const Vec& describing the time 1 state sample
y1	is a const Vec& describing the time 1 datum

Returns: the density or log-density evaluation

◆ q1Samp()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t , typename float_t , bool debug = false>

virtual ssv pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >::q1Samp ( const osv & y1 )

pure virtual

Samples from time 1 proposal.

Parameters

y1	is a const Vec& representing the first observed datum

Returns: the sample as a Vec

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

include/pf/bootstrap_filter.h

Public Member Functions

Protected Attributes

Private Types

Additional Inherited Members

Detailed Description

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t, typename float_t, bool debug = false> class pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >

Member Typedef Documentation

◆ arrayFloat

◆ arrayStates

◆ Mat

◆ osv

◆ ssv

Constructor & Destructor Documentation

◆ BSFilter()

Member Function Documentation

◆ filter()

◆ fSamp()

◆ getExpectations()

◆ getLogCondLike()

◆ logGEv()

◆ logMuEv()

◆ logQ1Ev()

◆ q1Samp()

template<size_t nparts, size_t dimx, size_t dimy, typename resamp_t, typename float_t, bool debug = false>
class pf::filters::BSFilter< nparts, dimx, dimy, resamp_t, float_t, debug >