ytStat Class Reference

Statistical routines. More...

#include <math/ytStat.h>

Public Member Functions
double	ytStat_dnorm (double x, double m, double sd)
	Calculates the density function of the normal distribution.
void	ytStat_mvmean (const double *X, int n, int p, int N, double *mean)
	Calculates the mean of all the variables.
void	ytStat_mvcov (const double *X, int n, int p, int N, const double *m, double *S)
	Calculates the covariance matrix.
double	ytStat_mvdnorm (const double *x, int p, int N, const double *u, const double *Sinv, double detS)
	Calculates the density of the multivariate normal distribution.
double	ytStat_logmvdnorm (const double *x, int p, int N, const double *u, const double *Sinv, double logDetS)
	Calculates the log of multivariate normal distribution density.

Detailed Description

Statistical routines.

See also

ytMatrix

Member Function Documentation

ytStat_dnorm()

double ytStat_dnorm	(	double	x,
		double	m,
		double	sd
	)

Calculates the density function of the normal distribution.

Parameters

x
m	parameter mean
sd	parameter sd (not variance σ2).

ytStat_logmvdnorm()

double ytStat_logmvdnorm	(	const double *	x,
		int	p,
		int	N,
		const double *	u,
		const double *	Sinv,
		double	logDetS
	)

Calculates the log of multivariate normal distribution density.

See non-log version ytStat_mvdnorm() for details. Note that the determinant of Σ is given by log(|Σ|).

Parameters

x	p input vector.
p	dimension of x.
N	stride of x.
u	mean vector (μ).
Sinv	Upper triangular symmetric matrix of inverse of the p by p covariance matrix (Σ-1).
logDetS	log of determinant of the convariance matrix (log(|Σ|)).

ytStat_mvcov()

void ytStat_mvcov	(	const double *	X,
		int	n,
		int	p,
		int	N,
		const double *	m,
		double *	S
	)

Calculates the covariance matrix.

Parameters

[in]	X	column major n by p input data matrix. Each column represents a set of samples of a variable.
[in]	n	Number of rows (sampls) in X.
[in]	p	Number of columns (variables) in X.
[in]	N	Stride width of X.
[in]	m	p mean vector where the j-th elemente represents the mean of the j -th variable.
[out]	S	p by p upper triangular matrix where element s i,j ( i ≤ j ) represents covariance of the i -th and j -th column.

ytStat_mvdnorm()

double ytStat_mvdnorm	(	const double *	x,
		int	p,
		int	N,
		const double *	u,
		const double *	Sinv,
		double	detS
	)

Calculates the density of the multivariate normal distribution.

The mean vector can be calculated by ytStat_mvmean(), and the covariance matrix can be calculated by ytStat_mvcov().

The inverse of the covariance matrix Σ returned by ytStat_mvcov() can be calculated by ytMatrix_dsytrf() and ytMatrix_dsytri().

Parameters

x	p input vector.
p	dimension of input value x.
u	mean vector (μ).
Sinv	Upper triangular symmetric matrix of inverse of the p by p covariance matrix (Σ-1).
detS	determinant of the covariance matrix (|Σ|).

ytStat_mvmean()

void ytStat_mvmean	(	const double *	X,
		int	n,
		int	p,
		int	N,
		double *	mean
	)

Calculates the mean of all the variables.

Parameters

[in]	X	column major n by p input data matrix. Each column represents a set of samples of a variable.
[in]	n	Number of rows (samples) in X.
[in]	p	Number of columns (variables) in X.
[in]	N	Stride width of X.

---

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

• math/ytStat.c