lti::entropyFunctor< T > Class Template Reference

Functor which computes entropy for vectors and matrices. More...

#include <ltiEntropyFunctor.h>

Inheritance diagram for lti::entropyFunctor< T >:

[legend]

Collaboration diagram for lti::entropyFunctor< T >:

[legend]

List of all members.

Classes
class	parameters
	the parameters for the class entropyFunctor More...
Public Member Functions
	entropyFunctor ()
	entropyFunctor (const bool &normalize, const bool &rowWise=true)
	entropyFunctor (const entropyFunctor &other)
virtual	~entropyFunctor ()
virtual const char *	getTypeName () const
T	apply (const vector< T > &srcdest) const
T &	apply (const vector< T > &src, T &dest) const
vector< T > &	apply (const matrix< T > &src, vector< T > &dest) const
T &	apply (const matrix< T > &src, T &dest) const
void	entropyOfRows (const matrix< T > &src, vector< T > &dest) const
void	entropyOfColumns (const matrix< T > &src, vector< T > &dest) const
entropyFunctor &	copy (const entropyFunctor &other)
virtual functor *	clone () const
const parameters &	getParameters () const

Detailed Description

template<class T>
class lti::entropyFunctor< T >

Functor which computes entropy for vectors and matrices.

The template parameter T corresponds to the value type of the objects to be evaluated. For example, to calculate the entropy of a vector<double> you will need an entropyFunctor<double>.

This functor assumes that the rows (or columns) of matrix contain probability distributions, i.e. the sum of the rows (column) elements must be 1.0. The entropy for the row will be defined as the sum of p(x)*ln(p(x)) for all x, where p(x) are the elements of the vector, row or column.

Constructor & Destructor Documentation

template<class T >

lti::entropyFunctor< T >::entropyFunctor ( )

default constructor

template<class T >

lti::entropyFunctor< T >::entropyFunctor	(	const bool &	normalize,
		const bool &	rowWise = `true`
	)

constructor with parameter preinitialization

template<class T >

lti::entropyFunctor< T >::entropyFunctor ( const entropyFunctor< T > & other )

copy constructor

Parameters:

other

the object to be copied

template<class T >

virtual lti::entropyFunctor< T >::~entropyFunctor ( ) [virtual]

destructor

Member Function Documentation

template<class T >

T& lti::entropyFunctor< T >::apply	(	const matrix< T > &	src,
		T &	dest
	)			const

computes the entropy of the given matrix.

Consider the matrix as a two-dimensional probability distribution.

The entropy is defined as the sum for all vector elements of p(x,y)*ln(p(x,y)), with p(x,y) = src.at(x,y).

Parameters:

	src	matrix<T> with the source data.
	dest	matrix<T> where the result will be left.

Returns:: a reference to the dest.

template<class T >

vector<T>& lti::entropyFunctor< T >::apply	(	const matrix< T > &	src,
		vector< T > &	dest
	)			const

The result of this function depends on the value of parameters.rowWise.

If this parameter is true, the functor will compute a vector, whose elements contain each the entropy of one column of the matrix (this computes the entropy of the rows, where each row is a data point in n-dimensional space. if rowWise is false, the result vector contains the entropy of the columns of the matrix (each column a data point).

The entropy of a row or column is defined as the sum for all its elements of p(x)*log2(p(x)), with p(x) = elements of the row or column.

Parameters:

	src	matrix<T> with the source data.
	dest	matrix<T> where the result will be left.

Returns:: a reference to the dest.

template<class T >

T& lti::entropyFunctor< T >::apply	(	const vector< T > &	src,
		T &	dest
	)			const

computes the entropy of the given vector.

The entropy is defined as the sum for all vector elements of p(x)*ln(p(x)), with p(x) = srcdest.at(x). Warning: Elements that are less than zero are ignored.

Parameters:

	src	vector<T> with the source data.
	dest	vector<T> where the result will be left.

Returns:: a reference to the dest.

template<class T >

T lti::entropyFunctor< T >::apply ( const vector< T > & srcdest ) const

returns the entropy of the current vector.

The entropy is defined as the sum for all vector elements of p(x)*ln(p(x)), with p(x) = srcdest.at(x). Warning: Elements that are less than zero are ignored.

Parameters:

srcdest

vector<T> with the source data. The result will be left here too.

Returns:: a reference to the srcdest.

template<class T >

virtual functor* lti::entropyFunctor< T >::clone ( ) const [virtual]

returns a pointer to a clone of this functor.

Reimplemented from lti::statisticsFunctor.

template<class T >

entropyFunctor& lti::entropyFunctor< T >::copy ( const entropyFunctor< T > & other )

copy data of "other" functor.

Parameters:

other

the functor to be copied

Returns:: a reference to this functor object

Reimplemented from lti::statisticsFunctor.

template<class T >

void lti::entropyFunctor< T >::entropyOfColumns	(	const matrix< T > &	src,
		vector< T > &	dest
	)			const

The result vector contains the entropy of the columns of the matrix (each column a data point).

Parameters:

	src	matrix<T> with the source data.
	dest	matrix<T> where the result will be left.

template<class T >

void lti::entropyFunctor< T >::entropyOfRows	(	const matrix< T > &	src,
		vector< T > &	dest
	)			const

This function will compute a vector, whose elements contain each the entropy of one column of the matrix (this computes the entropy of the rows, where each row is a data point in n-dimensional space).

Parameters: