lti::serialVectorStats< T > Class Template Reference

Functor which computes the statistics of a series of vectors. More...

#include <ltiSerialVectorStats.h>

Inheritance diagram for lti::serialVectorStats< T >:

[legend]

Collaboration diagram for lti::serialVectorStats< T >:

[legend]

List of all members.

Classes
class	parameters
	the parameters for the class serialVectorStats More...
Public Member Functions
	serialVectorStats ()
	serialVectorStats (const serialVectorStats &other)
virtual	~serialVectorStats ()
virtual const char *	getTypeName () const
void	apply (vector< T > &mean, vector< T > &variance) const
void	apply (vector< T > &mean, matrix< T > &covariance) const
void	considerRows (const matrix< T > &element)
void	consider (const vector< T > &element)
void	consider (const vector< T > &element, const int &n)
double	getN () const
void	setN (double newN)
void	getMean (vector< T > &mean) const
void	getVariance (vector< T > &var) const
void	getCovariance (matrix< T > &covar) const
void	getSumOfVectors (vector< T > &s) const
void	getSumOfVecOuterProd (matrix< sqrType > &s) const
const vector< T > &	getSumOfVectors () const
const matrix< sqrType > &	getSumOfVecOuterProd () const
void	reset ()
serialVectorStats &	copy (const serialVectorStats &other)
serialVectorStats &	operator= (const serialVectorStats &other)
virtual functor *	clone () const
const parameters &	getParameters () const
virtual bool	read (ioHandler &handler, const bool complete=true)
virtual bool	write (ioHandler &handler, const bool complete=true) const
Protected Member Functions
void	outer (const vector< T > &a, const vector< T > &b, matrix< sqrType > &dest) const
void	outerAcc (const vector< T > &a, const vector< T > &b, matrix< sqrType > &dest) const
void	outerAcc (const vector< T > &a, const vector< T > &b, const int n, matrix< sqrType > &dest) const
void	diagonal (const matrix< sqrType > &mat, vector< T > &diag) const

Detailed Description

template<class T>
class lti::serialVectorStats< T >

Functor which computes the statistics of a series of vectors.

It delivers the mean vector and the covariance matrix, or alternatively the variance vector under the assumption, that the components of the vectors are independent from each other.

If the empirical computation type is selected, the variances resp. covariances will be normalized with (N-1). Otherwise with N.

An example (maximum-likelihood case, i.e. normalizing with N):

Assume we have three vectors {3,4},{1,3},{2,2}

The mean vector is {2,3}

The covariance matrix, which is always symetric, is:

 | 2/3  1/3  |
 | 1/3  2/3  |

The variance vector is {2/3,2/3}, i.e. the diagonal elements of the covariance matrix.

Constructor & Destructor Documentation

template<class T>

lti::serialVectorStats< T >::serialVectorStats ( )

default constructor

template<class T>

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

copy constructor

Parameters:

other

the object to be copied

template<class T>

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

destructor

Member Function Documentation

template<class T>

void lti::serialVectorStats< T >::apply	(	vector< T > &	mean,
		matrix< T > &	covariance
	)			const

returns mean vector and covariance matrix of the data vectors considered so far.

Parameters:

	mean	the result, where the mean will be put
	covariance	the matrix, where the covariances will be put

template<class T>

void lti::serialVectorStats< T >::apply	(	vector< T > &	mean,
		vector< T > &	variance
	)			const

returns mean and variance vectors of the data vectors considered so far.

Parameters:

	mean	the result, where the mean will be put
	variance	the result, where the variance will be put

template<class T>

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

returns a pointer to a clone of this functor.

Reimplemented from lti::statisticsFunctor.

template<class T>

void lti::serialVectorStats< T >::consider	(	const vector< T > &	element,
		const int &	n
	)

This function considers the given vector for computing the mean and variance, as if it were presented n times.

template<class T>

void lti::serialVectorStats< T >::consider ( const vector< T > & element )

This function considers the given vector for computing the mean and variance.

template<class T>

void lti::serialVectorStats< T >::considerRows ( const matrix< T > & element )

This function considers the row vectors of the given matrix for computing the mean and variance.

template<class T>

serialVectorStats& lti::serialVectorStats< T >::copy ( const serialVectorStats< 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::serialVectorStats< T >::diagonal	(	const matrix< sqrType > &	mat,
		vector< T > &	diag
	)			const `[protected]`

get diagonal vector of the given matrix, but casted into the lower type

template<class T>

void lti::serialVectorStats< T >::getCovariance ( matrix< T > & covar ) const

The covariance of all vectors that have been considered so far.

template<class T>

void lti::serialVectorStats< T >::getMean ( vector< T > & mean ) const

The mean of all vectors that have been considered so far.

template<class T>

double lti::serialVectorStats< T >::getN ( ) const

return the number of elements considered so far

template<class T>

const parameters& lti::serialVectorStats< T >::getParameters ( ) const

returns used parameters

Reimplemented from lti::statisticsFunctor.

template<class T>

const matrix<sqrType>& lti::serialVectorStats< T >::getSumOfVecOuterProd ( ) const

The sum of the outer products of all vectors that have been considered so far.

(i.e. sum(vec*vecTransposed))

Returns:: a read-only reference to the internal data

template<class T>

void lti::serialVectorStats< T >::getSumOfVecOuterProd ( matrix< sqrType > & s ) const

The sum of the squares of all vectors that have been considered so far.

The values will be copied into the given vector.

template<class T>

const vector<T>& lti::serialVectorStats< T >::getSumOfVectors ( ) const

The sum of all vectors that have been considered so far.

Returns:: a read-only reference to the internal data

template<class T>

void lti::serialVectorStats< T >::getSumOfVectors ( vector< T > & s ) const

The sum of all vectors that have been considered so far.

The values will be copied into the given vector.

template<class T>

virtual const char* lti::serialVectorStats< T >::getTypeName ( ) const [virtual]

returns the name of this type ("serialVectorStats")

Reimplemented from lti::statisticsFunctor.

template<class T>

void lti::serialVectorStats< T >::getVariance ( vector< T > & var ) const

The variance of all vectors that have been considered so far.

template<class T>

serialVectorStats& lti::serialVectorStats< T >::operator= ( const serialVectorStats< T > & other )

Alias for copy.

Reimplemented from lti::functor.

template<class T>

void lti::serialVectorStats< T >::outer	(	const vector< T > &	a,
		const vector< T > &	b,
		matrix< sqrType > &	dest
	)			const `[protected]`

compute the outer product of two vector<T> and leave the result in a matrix<sqrType>

template<class T>

void lti::serialVectorStats< T >::outerAcc	(	const vector< T > &	a,
		const vector< T > &	b,
		const int	n,
		matrix< sqrType > &	dest
	)			const `[protected]`

accumulate n times the outer product of a and b in the matrix dest The matrix dest MUST have the correct size before calling this method.

template<class T>

void lti::serialVectorStats< T >::outerAcc	(	const vector< T > &	a,
		const vector< T > &	b,
		matrix< sqrType > &	dest
	)			const `[protected]`

accumulate the outer product of a and b in the matrix dest The matrix dest MUST have the correct size before calling this method.

template<class T>

virtual bool lti::serialVectorStats< T >::read	(	ioHandler &	handler,
		const bool	complete = `true`
	)			`[virtual]`

Reads this functor from the given handler.

Reimplemented from lti::functor.

template<class T>

void lti::serialVectorStats< T >::reset ( )

Reset the series, i.e.

discard all considered elements.

template<class T>

void lti::serialVectorStats< T >::setN ( double newN )

set the number of elements considered so far

template<class T>

virtual bool lti::serialVectorStats< T >::write	(	ioHandler &	handler,
		const bool	complete = `true`
	)			const `[virtual]`

Writes this functor to the given handler.

Reimplemented from lti::functor.

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

ltiSerialVectorStats.h

lti::serialVectorStats< T > Class Template Reference

Classes

Public Member Functions

Protected Member Functions

Detailed Description

template<class T> class lti::serialVectorStats< T >

Constructor & Destructor Documentation

Member Function Documentation

template<class T>
class lti::serialVectorStats< T >