Binary factor to estimate parameters of zero-mean Gaussian white noise. More...

#include <WhiteNoiseFactor.h>

Inheritance diagram for gtsam::WhiteNoiseFactor:

Static Public Member Functions
static double	f (double z, double u, double p)
	negative log likelihood as a function of mean $\mu$ and precision $\tau$ $f(z, \tau, \mu) = -\log \left( \frac{\sqrt{\tau}}{\sqrt{2\pi}} \exp(-0.5\tau(z-\mu)^2) \right) = \log(\sqrt{2\pi}) - 0.5* \log(\tau) + 0.5\tau(z-\mu)^2$ More...

static HessianFactor::shared_ptr	linearize (double z, double u, double p, Key j1, Key j2)
	linearize returns a Hessianfactor that approximates error Hessian is $0.5f - x^T g + 0.5*x^T G x$ Taylor expansion is $f(x+dx) = f(x) + df(x) dx + 0.5 ddf(x) dx^2$ So f = 2 f(x), g = -df(x), G = ddf(x) More...

Private Types
typedef NonlinearFactor	Base

Private Attributes
Key	meanKey_
	key by which to access mean variable More...

Key	precisionKey_
	key by which to access precision variable More...

double	z_
	Measurement. More...

Standard Constructors
	WhiteNoiseFactor (double z, Key meanKey, Key precisionKey)

Advanced Constructors
	~WhiteNoiseFactor () override
	Destructor. More...

Testable
void	print (const std::string &p="WhiteNoiseFactor", const KeyFormatter &keyFormatter=DefaultKeyFormatter) const override
	Print. More...

Standard Interface
size_t	dim () const override
	get the dimension of the factor (number of rows on linearization) More...

double	error (const Values &x) const override
	Calculate the error of the factor, typically equal to log-likelihood. More...

virtual Vector	unwhitenedError (const Values &x) const

Advanced Interface
boost::shared_ptr< GaussianFactor >	linearize (const Values &x) const override
	linearize returns a Hessianfactor that is an approximation of error(p) More...

Additional Inherited Members
Public Types inherited from gtsam::NonlinearFactor
typedef boost::shared_ptr< This >	shared_ptr

Public Types inherited from gtsam::Factor
typedef KeyVector::const_iterator	const_iterator
	Const iterator over keys. More...

typedef KeyVector::iterator	iterator
	Iterator over keys. More...

Public Member Functions inherited from gtsam::NonlinearFactor
	NonlinearFactor ()

template<typename CONTAINER >
	NonlinearFactor (const CONTAINER &keys)

virtual bool	equals (const NonlinearFactor &f, double tol=1e-9) const

virtual	~NonlinearFactor ()

virtual bool	active (const Values &) const

virtual shared_ptr	clone () const

shared_ptr	rekey (const std::map< Key, Key > &rekey_mapping) const

shared_ptr	rekey (const KeyVector &new_keys) const

Public Member Functions inherited from gtsam::Factor
virtual	~Factor ()=default
	Default destructor. More...

Key	front () const
	First key. More...

Key	back () const
	Last key. More...

const_iterator	find (Key key) const
	find More...

const KeyVector &	keys () const
	Access the factor's involved variable keys. More...

const_iterator	begin () const

const_iterator	end () const

size_t	size () const

virtual void	printKeys (const std::string &s="Factor", const KeyFormatter &formatter=DefaultKeyFormatter) const
	print only keys More...

KeyVector &	keys ()

iterator	begin ()

iterator	end ()

Protected Types inherited from gtsam::NonlinearFactor
typedef Factor	Base

typedef NonlinearFactor	This

Protected Member Functions inherited from gtsam::Factor
	Factor ()

template<typename CONTAINER >
	Factor (const CONTAINER &keys)

template<typename ITERATOR >
	Factor (ITERATOR first, ITERATOR last)

bool	equals (const This &other, double tol=1e-9) const
	check equality More...

Static Protected Member Functions inherited from gtsam::Factor
template<typename CONTAINER >
static Factor	FromKeys (const CONTAINER &keys)

template<typename ITERATOR >
static Factor	FromIterators (ITERATOR first, ITERATOR last)

Protected Attributes inherited from gtsam::Factor
KeyVector	keys_
	The keys involved in this factor. More...

Detailed Description

Binary factor to estimate parameters of zero-mean Gaussian white noise.

This factor uses the mean-precision parameterization.

Takes three template arguments: Key: key type to use for mean Key: key type to use for precision Values: Values type for optimization

Definition at line 39 of file WhiteNoiseFactor.h.

Member Typedef Documentation

typedef NonlinearFactor gtsam::WhiteNoiseFactor::Base

private

Definition at line 48 of file WhiteNoiseFactor.h.

Constructor & Destructor Documentation

gtsam::WhiteNoiseFactor::WhiteNoiseFactor	(	double	z,
		Key	meanKey,
		Key	precisionKey
	)

inline

Construct from measurement

Parameters

z	Measurment value
meanKey	Key for mean variable
precisionKey	Key for precision variable

Definition at line 94 of file WhiteNoiseFactor.h.

gtsam::WhiteNoiseFactor::~WhiteNoiseFactor ( )

inlineoverride

Destructor.

Definition at line 103 of file WhiteNoiseFactor.h.

Member Function Documentation

size_t gtsam::WhiteNoiseFactor::dim ( ) const

inlineoverridevirtual

get the dimension of the factor (number of rows on linearization)

Implements gtsam::NonlinearFactor.

Definition at line 122 of file WhiteNoiseFactor.h.

double gtsam::WhiteNoiseFactor::error ( const Values & x ) const

inlineoverridevirtual

Calculate the error of the factor, typically equal to log-likelihood.

Implements gtsam::NonlinearFactor.

Definition at line 127 of file WhiteNoiseFactor.h.

static double gtsam::WhiteNoiseFactor::f	(	double	z,
		double	u,
		double	p
	)

inlinestatic

negative log likelihood as a function of mean $\mu$ and precision $\tau$

$f(z, \tau, \mu) = -\log \left( \frac{\sqrt{\tau}}{\sqrt{2\pi}} \exp(-0.5\tau(z-\mu)^2) \right) = \log(\sqrt{2\pi}) - 0.5* \log(\tau) + 0.5\tau(z-\mu)^2$

Definition at line 59 of file WhiteNoiseFactor.h.

static HessianFactor::shared_ptr gtsam::WhiteNoiseFactor::linearize	(	double	z,
		double	u,
		double	p,
		Key	j1,
		Key	j2
	)

inlinestatic

linearize returns a Hessianfactor that approximates error Hessian is

$0.5f - x^T g + 0.5*x^T G x$

Taylor expansion is

$f(x+dx) = f(x) + df(x) dx + 0.5 ddf(x) dx^2$

So f = 2 f(x), g = -df(x), G = ddf(x)

Definition at line 73 of file WhiteNoiseFactor.h.

boost::shared_ptr<GaussianFactor> gtsam::WhiteNoiseFactor::linearize ( const Values & x ) const

inlineoverridevirtual

linearize returns a Hessianfactor that is an approximation of error(p)

Implements gtsam::NonlinearFactor.

Definition at line 156 of file WhiteNoiseFactor.h.

void gtsam::WhiteNoiseFactor::print	(	const std::string &	p = `"WhiteNoiseFactor"`,
		const KeyFormatter &	keyFormatter = `DefaultKeyFormatter`
	)		const

inlineoverridevirtual

Print.

Reimplemented from gtsam::NonlinearFactor.

Definition at line 111 of file WhiteNoiseFactor.h.

virtual Vector gtsam::WhiteNoiseFactor::unwhitenedError ( const Values & x ) const

inlinevirtual

Vector of errors "unwhitened" does not make sense for this factor What is meant typically is only "e" above Here we shoehorn sqrt(2*error(p)) TODO: Where is this used? should disappear.

Definition at line 138 of file WhiteNoiseFactor.h.

Member Data Documentation

Key gtsam::WhiteNoiseFactor::meanKey_

private

key by which to access mean variable

Definition at line 45 of file WhiteNoiseFactor.h.

Key gtsam::WhiteNoiseFactor::precisionKey_

private

key by which to access precision variable

Definition at line 46 of file WhiteNoiseFactor.h.

double gtsam::WhiteNoiseFactor::z_

private

Measurement.

Definition at line 43 of file WhiteNoiseFactor.h.

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

WhiteNoiseFactor.h

Static Public Member Functions

Private Types

Private Attributes

Standard Constructors

Advanced Constructors

Testable

Standard Interface

Advanced Interface

Additional Inherited Members

Detailed Description

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation