kjb::MV_gaussian_distribution Class Reference

Multivariate Gaussian (normal) distribution. More...

#include <prob_distribution.h>

Public Member Functions
	MV_gaussian_distribution (int d)
	Constructs a d-dimensinoal multivariate Gaussian with a mean of zero and a covariance matrix of identity. More...
	MV_gaussian_distribution (const Vector &mu, const Matrix &sigma)
	Constructs a multivariate Gaussian with a mean of mu and a covariance matrix of sigma. sigma must be positive-definite. More...
	MV_gaussian_distribution (const Vector &mu, const Vector &sigma_diag)
	Constructs a multivariate Gaussian with a mean of mu and a covariance matrix diag(sigma_diag). More...
const Vector &	get_mean () const
	Gets the mean of this distribution. More...
const Matrix &	get_covariance_matrix () const
	Gets the covariance matrix of this distribution. More...
void	set_mean (const Vector &m)
	Gets the mean of this distribution. More...
void	set_covariance_matrix (const Matrix &S)
	Gets the covariance matrix of this distribution. More...
void	set_covariance_matrix (const Matrix &S, const Vector &m)
	Gets the covariance matrix of this distribution. More...
int	get_dimension () const
	Gets the dimension of this distribution. More...
MV_gaussian_conditional_distribution	conditional (int i) const
	Conditional distribution of the ith variable given the rest. More...

Friends
double	pdf (const MV_gaussian_distribution &, const Vector &)
	Computes the joint PDF of a multivariate Gaussian at x. More...
double	log_pdf (const MV_gaussian_distribution &, const Vector &)
	Computes the log PDF a multivariate normal distribution at x. More...
Vector	sample (const MV_gaussian_distribution &)
	Sample from a multivariate normal distribution. More...

Detailed Description

Multivariate Gaussian (normal) distribution.

This class implements a multivariate Gaussian distribution.

Constructor & Destructor Documentation

kjb::MV_gaussian_distribution::MV_gaussian_distribution ( int  d )

inline

Constructs a d-dimensinoal multivariate Gaussian with a mean of zero and a covariance matrix of identity.

Parameters

d	The dimension of the distribution.

kjb::MV_gaussian_distribution::MV_gaussian_distribution ( const Vector &  mu, const Matrix &  sigma  )

inline

Constructs a multivariate Gaussian with a mean of mu and a covariance matrix of sigma. sigma must be positive-definite.

Parameters

mu	The mean of the distribution.
sigma	The covariance matrix of the distribution.

kjb::MV_gaussian_distribution::MV_gaussian_distribution ( const Vector &  mu, const Vector &  sigma_diag  )

inline

Constructs a multivariate Gaussian with a mean of mu and a covariance matrix diag(sigma_diag).

Parameters

mu	The mean of the distribution.
sigma_diag	The diagonal of the covariance matrix.

Member Function Documentation

MV_gaussian_conditional_distribution kjb::MV_gaussian_distribution::conditional

(

int

i

)

const

Conditional distribution of the ith variable given the rest.

const Matrix& kjb::MV_gaussian_distribution::get_covariance_matrix ( ) const

inline

Gets the covariance matrix of this distribution.

int kjb::MV_gaussian_distribution::get_dimension ( ) const

inline

Gets the dimension of this distribution.

const Vector& kjb::MV_gaussian_distribution::get_mean ( ) const

inline

Gets the mean of this distribution.

void kjb::MV_gaussian_distribution::set_covariance_matrix ( const Matrix &  S )

inline

Gets the covariance matrix of this distribution.

void kjb::MV_gaussian_distribution::set_covariance_matrix ( const Matrix &  S, const Vector &  m  )

inline

Gets the covariance matrix of this distribution.

void kjb::MV_gaussian_distribution::set_mean ( const Vector &  m )

inline

Gets the mean of this distribution.

Friends And Related Function Documentation

double log_pdf ( const MV_gaussian_distribution &  , const Vector &    )

friend

Computes the log PDF a multivariate normal distribution at x.

This is a specialization of the generic joint_log_pdf function. It is specialized to avoid the computation of the log of the exponential.

double pdf ( const MV_gaussian_distribution &  , const Vector &    )

friend

Computes the joint PDF of a multivariate Gaussian at x.

Vector sample ( const MV_gaussian_distribution &  )

friend

Sample from a multivariate normal distribution.

---

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

• /work/kobus/src/lib/prob_cpp/prob_distribution.h
• /work/kobus/src/lib/prob_cpp/prob_distribution.cpp