This class represents a discrete probability distribution (or probability mass function). Currently, this PMF can only take strings as values. More...

#include <PMF.h>

Inheritance diagram for pbl::PMF:

[legend]

List of all members.

Classes
struct	PMFStruct
Public Member Functions
PMF *	clone () const
	Creates a clone of the object. The clone method is cheap since it only copies a pointer. A deep clone will only be created if the original object is modified.
double	getDensity (const arma::vec &v) const
int	getDomainSize () const
	Returns the domain size of this distribution.
bool	getExpectedValue (std::string &v) const
	Returns in parameter v the expected value for this distribution, i.e., the value with the highest probability.
double	getLikelihood (const PDF &pdf) const
double	getLikelihood (const PMF &pmf) const
double	getMaxDensity () const
std::string	getMostProbableValue () const
void	getProbabilities (std::vector< double > &probabilities) const
	Returns all probabilities of the known values.
double	getProbability (const std::string &value) const
	Returns the probability of the given value.
double	getProbability (const std::string &value, int domain_size) const
	Returns the probability of the given value, for a given domain_size. The domain size determines the probability if the value is unknown, in which case a uniform distribution over all unknown values is assumed.
double	getProbabilityUnknown () const
double	getProbabilityUnknown (int domain_size) const
void	getValues (std::vector< std::string > &values) const
	Returns a vector of values for which a probability is specified.
void	normalize ()
PMF &	operator= (const PMF &other)
	Assignment operator. The operation is cheap since it only copies a pointer. A deep clone will only be created if the original object is modified.
	PMF (int domain_size=-1)
	Constructs a discrete probability distribution. The optional parameter domain size states the number of possible values of the random variable underlying this distribution, and determines the probability of unknown values if the probabilities of all unknown values do not sum up to one.
	PMF (const PMF &pmf)
	Copy constructor.
void	setDomainSize (int domain_size)
	Sets the domain size of this discrete distribution.
void	setExact (const std::string &value)
	Set the probability of the given value to 1. All other values are given a probability of 0.
void	setProbability (const std::string &value, double p)
	Set the probability of a given value.
std::string	toString (const std::string &indent="") const
	Represents the PMF as a string for easier console output.
void	update (const pbl::PMF &pmf)
virtual	~PMF ()
	Destructor.
Protected Member Functions
void	cloneStruct ()
Protected Attributes
PMFStruct *	ptr_

Detailed Description

This class represents a discrete probability distribution (or probability mass function). Currently, this PMF can only take strings as values.

Author:: Sjoerd van den Dries

Date:: December, 2012

Version:: 1.0

Definition at line 52 of file PMF.h.

Constructor & Destructor Documentation

PMF::PMF ( int domain_size = -1 )

Constructs a discrete probability distribution. The optional parameter domain size states the number of possible values of the random variable underlying this distribution, and determines the probability of unknown values if the probabilities of all unknown values do not sum up to one.

Parameters:

domain_size The number of values the random variable of this pmf can take

Definition at line 41 of file PMF.cpp.

PMF::PMF ( const PMF & pmf )

Copy constructor.

Definition at line 44 of file PMF.cpp.

PMF::~PMF ( ) [virtual]

Destructor.

Definition at line 50 of file PMF.cpp.

Member Function Documentation

PMF * PMF::clone ( ) const [virtual]

Creates a clone of the object. The clone method is cheap since it only copies a pointer. A deep clone will only be created if the original object is modified.

Implements pbl::PDF.

Definition at line 75 of file PMF.cpp.

void PMF::cloneStruct ( ) [protected]

Definition at line 79 of file PMF.cpp.

double PMF::getDensity ( const arma::vec & v ) const

Definition at line 283 of file PMF.cpp.

int PMF::getDomainSize ( ) const

Returns the domain size of this distribution.

Definition at line 252 of file PMF.cpp.

bool PMF::getExpectedValue ( std::string & v ) const [virtual]

Returns in parameter v the expected value for this distribution, i.e., the value with the highest probability.

Returns:: Whether an expected value was found

Reimplemented from pbl::PDF.

Definition at line 136 of file PMF.cpp.

double PMF::getLikelihood ( const PDF & pdf ) const [virtual]

Implements pbl::PDF.

Definition at line 153 of file PMF.cpp.

double PMF::getLikelihood ( const PMF & pmf ) const

Definition at line 160 of file PMF.cpp.

double PMF::getMaxDensity ( ) const [virtual]

Implements pbl::PDF.

Definition at line 288 of file PMF.cpp.

std::string PMF::getMostProbableValue ( ) const

Definition at line 311 of file PMF.cpp.

void PMF::getProbabilities ( std::vector< double > & probabilities ) const

Returns all probabilities of the known values.

Definition at line 130 of file PMF.cpp.

double PMF::getProbability ( const std::string & value ) const

Returns the probability of the given value.

Returns:: The probability of the given value

Definition at line 86 of file PMF.cpp.

double PMF::getProbability	(	const std::string &	value,
		int	domain_size
	)		const

Returns the probability of the given value, for a given domain_size. The domain size determines the probability if the value is unknown, in which case a uniform distribution over all unknown values is assumed.

Returns:: The probability of the given value

Definition at line 90 of file PMF.cpp.

double PMF::getProbabilityUnknown ( ) const

Definition at line 256 of file PMF.cpp.

double PMF::getProbabilityUnknown ( int domain_size ) const

Definition at line 260 of file PMF.cpp.

void PMF::getValues ( std::vector< std::string > & values ) const

Returns a vector of values for which a probability is specified.

Definition at line 124 of file PMF.cpp.

void PMF::normalize ( )

Definition at line 267 of file PMF.cpp.

PMF & PMF::operator= ( const PMF & other )

Assignment operator. The operation is cheap since it only copies a pointer. A deep clone will only be created if the original object is modified.

Definition at line 58 of file PMF.cpp.

void PMF::setDomainSize ( int domain_size )

Sets the domain size of this discrete distribution.

Definition at line 247 of file PMF.cpp.

void PMF::setExact ( const std::string & value )

Set the probability of the given value to 1. All other values are given a probability of 0.

Definition at line 113 of file PMF.cpp.

void PMF::setProbability	(	const std::string &	value,
		double	p
	)

Set the probability of a given value.

Definition at line 100 of file PMF.cpp.

std::string PMF::toString ( const std::string & indent = "" ) const [virtual]

Represents the PMF as a string for easier console output.

Note:: Should be changed into stream operator <<

Returns:: The Gaussian as string

Implements pbl::PDF.

Definition at line 293 of file PMF.cpp.

void PMF::update ( const pbl::PMF & other )

Todo:: : make this implementation more efficient (no need for O(log n) look-ups)

Definition at line 214 of file PMF.cpp.

Member Data Documentation

PMFStruct* pbl::PMF::ptr_ [protected]

Definition at line 186 of file PMF.h.

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

Classes

Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation