meta::stats Namespace Reference

Probability distributions and other statistics functions. More...

Classes
class	dirichlet
	Represents a Dirichlet distribution. More...
class	multinomial
	Represents a multinomial/categorical distribution. More...

Functions
template<class T >
multinomial< T >	operator+ (const multinomial< T > &lhs, const multinomial< T > &rhs)
template<class T >
multinomial< T >	operator+ (multinomial< T > &&lhs, const multinomial< T > &rhs)
template<class T >
multinomial< T >	operator+ (const multinomial< T > &lhs, multinomial< T > &&rhs)
template<class T >
multinomial< T >	operator+ (multinomial< T > &&lhs, multinomial< T > &&rhs)
template<class Dist , class Fun >
double	expected_value (Dist &&dist, Fun &&fun)
	Computation for \(E_d[f(x)]\) where \(d\) is specified by the dist parameter and \(f(x)\) is the fun parameter. More...
template<class Dist >
double	entropy (Dist &&dist)
	Computes the entropy \(H(X) = - \sum_{x \in X} p(x) \log_2 p(x)\). More...

Detailed Description

Probability distributions and other statistics functions.

Function Documentation

template<class Dist , class Fun >

double meta::stats::expected_value	(	Dist &&	dist,
		Fun &&	fun
	)

Computation for \(E_d[f(x)]\) where \(d\) is specified by the dist parameter and \(f(x)\) is the fun parameter.

dist must be a distribution over input type accepted by f.

Parameters

dist	The distribution the expectation should be calculated against
fun	The function to compute the expectation of

Returns

the expected value of the function under the given distribution

template<class Dist >

double meta::stats::entropy

(

Dist &&

dist

)

Computes the entropy \(H(X) = - \sum_{x \in X} p(x) \log_2 p(x)\).

Parameters

dist	The distribution to compute the entropy over

Returns

the entropy of the supplied distribution