gplib::mv_gauss Class Reference

#include <mvgauss.hpp>

Collaboration diagram for gplib::mv_gauss:

Classes
struct	implementation

Public Member Functions
	mv_gauss ()
	Constructor. More...
	mv_gauss (const mv_gauss &other)
	Constructor, requieres another Multivariate Gaussian distribution to be copied. More...
	mv_gauss (const arma::vec &mean, const arma::mat &cov)
	Constructor, requieres a arma vector of means and a matrix of covariances. More...
	~mv_gauss ()
	Destructor. More...
void	set_mean (const arma::vec &mean)
	Sets the mean vector to the Gaussian distribution. More...
void	set_cov (const arma::mat &cov)
	Sets the covariance matrix to the Gaussian distribution. More...
arma::vec	get_mean () const
	Gets the mean vector. More...
arma::mat	get_cov () const
	Gets the covanriance matrix. More...
arma::mat	get_cov_inv () const
	Returns the inverse of the covanriance matrix. More...
arma::mat	get_cov_chol () const
	Returns the cholesky decomposition of the covanriance matrix. More...
size_t	dimension () const
	Returns the dimensionality of the Gaussian ditribution. More...
arma::mat	sample (int n_samples) const
	Returns n_samples samples in a matrix with n_samples rows and D columns. More...
double	log_density (const arma::vec &x) const
	Returns the logartihm density of the Gaussian distribution. More...
double	density (const arma::vec &x) const
	Returns the density of the Gaussian ditribution. More...
mv_gauss	marginalize_hidden (const std::vector< bool > &observed) const
	Returns the marginal distribution after integrating out the non observed variables. More...
mv_gauss	conditional (const arma::vec &observation, const std::vector< bool > &observed) const
	Returns the conditional distribution of the hidden variables given the an observation of the observed variables. More...
mv_gauss	operator= (const mv_gauss &other)
	Overload of the operand = to work with Multivariate Gaussian class. More...

Private Attributes
implementation *	pimpl

Constructor & Destructor Documentation

gplib::mv_gauss::mv_gauss

(

)

Constructor.

gplib::mv_gauss::mv_gauss

(

const mv_gauss &

other

)

Constructor, requieres another Multivariate Gaussian distribution to be copied.

Parameters

other

: Multivariate Gaussian.

gplib::mv_gauss::mv_gauss	(	const arma::vec &	mean,
		const arma::mat &	cov
	)

Constructor, requieres a arma vector of means and a matrix of covariances.

Parameters

mean	: Vector of means.
cov	: Matrix of covariance.

gplib::mv_gauss::~mv_gauss

(

)

Destructor.

Member Function Documentation

mv_gauss gplib::mv_gauss::conditional	(	const arma::vec &	observation,
		const std::vector< bool > &	observed
	)		const

Returns the conditional distribution of the hidden variables given the an observation of the observed variables.

Only the values for which the observed vector is true are considered on vector observation.

Parameters

observation	: vector, indicates the observed values.
observed	: boolean vector, indicates which values are observed.

double gplib::mv_gauss::density

(

const arma::vec &

x

)

const

Returns the density of the Gaussian ditribution.

Parameters

x	: Vector of random variables.

size_t gplib::mv_gauss::dimension

(

)

const

Returns the dimensionality of the Gaussian ditribution.

mat gplib::mv_gauss::get_cov

(

)

const

Gets the covanriance matrix.

mat gplib::mv_gauss::get_cov_chol

(

)

const

Returns the cholesky decomposition of the covanriance matrix.

mat gplib::mv_gauss::get_cov_inv

(

)

const

Returns the inverse of the covanriance matrix.

vec gplib::mv_gauss::get_mean

(

)

const

Gets the mean vector.

double gplib::mv_gauss::log_density

(

const arma::vec &

x

)

const

Returns the logartihm density of the Gaussian distribution.

Parameters

x	: Vector of random variables.

mv_gauss gplib::mv_gauss::marginalize_hidden

(

const std::vector< bool > &

observed

)

const

Returns the marginal distribution after integrating out the non observed variables.

If the value of observed[i] is true then the variable is assumed observed, otherwise it is integrated out.

Parameters

observed

: boolean vector which indicates which values are observed.

mv_gauss gplib::mv_gauss::operator=

(

const mv_gauss &

other

)

Overload of the operand = to work with Multivariate Gaussian class.

Parameters

other

: Gaussian distribution to be set.

mat gplib::mv_gauss::sample

(

int

n_samples

)

const

Returns n_samples samples in a matrix with n_samples rows and D columns.

Where D is the dimensionality of the Gaussian distribution.

Parameters

n_samples

: number of samples.

void gplib::mv_gauss::set_cov

(

const arma::mat &

cov

)

Sets the covariance matrix to the Gaussian distribution.

Parameters

cov	: Matrix of covanriance.

void gplib::mv_gauss::set_mean

(

const arma::vec &

mean

)

Sets the mean vector to the Gaussian distribution.

Parameters

mean	: Vector of means.

Member Data Documentation

implementation* gplib::mv_gauss::pimpl

private

---

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

• src/gplib/mvgauss.hpp
• src/gplib/mvgauss.cc