plLearnDistribVector Class Reference

This class permits to learn a set of conditional and non-conditional distributions in the same time. More...

#include <plLearn.h>

Inheritance diagram for plLearnDistribVector:

[legend]Collaboration diagram for plLearnDistribVector:

[legend]List of all members.

Public Member Functions
	plLearnDistribVector (const vector< plLearnObject * > &learn_objects, const plVariablesConjunction &vars, const vector< bool > &known_range)
	Constructs an objects to learn the vector {learn_objects} of learning objects (corresponding to a set of distributions).
	plLearnDistribVector (const vector< plLearnObject * > &learn_objects, const plVariablesConjunction &vars)
	The same as above the constructor, but {known_range} is assumed to be {true} as value for all variables.
virtual	~plLearnDistribVector ()
	Destructor.
void	reset ()
	Resets all learning objects.
void	internal_addPoint (const plDataValues &point, double weight=1.0)
	Adds a point {point} with a given weight {weight} and updates the statistics.
template<class filterArrayT>
void	addFilterNewPoint (const plValues &point, filterArrayT filter, double weight=1.0)
	Adds a point {point} (represented as a {plValue}) with a given weight {weight} and updates the statistics.
template<class arrayT, class filterArrayT>
void	addFilterNewPoint (arrayT *point, filterArrayT filter, double weight=1.0)
	Same as above, but {point} is represented as a C array.
template<class arrayT, class filterArrayT>
void	addFilterNewPoint (const vector< arrayT > &point, filterArrayT filter, double weight=1.0)
	Same as above, but {point} is represented as an STL vector.
template<class singleT, class filterArrayT>
void	addFilterNewPoint (singleT point, filterArrayT filter, double weight=1.0)
	Same as above, but {point} is a scalar value.
template<class filterArrayT>
void	addFilterPoint (const plDataValues &point, filterArrayT distrib_to_update, double weight=1.0)
	Same as above, but {point} is represented as an {plDataValues}.

---

Detailed Description

This class permits to learn a set of conditional and non-conditional distributions in the same time.

Definition at line 357 of file plLearn.h.

---

Constructor & Destructor Documentation

plLearnDistribVector::plLearnDistribVector (  const vector< plLearnObject * > &  learn_objects, const plVariablesConjunction &  vars, const vector< bool > &  known_range )

Constructs an objects to learn the vector {learn_objects} of learning objects (corresponding to a set of distributions). {vars} is the set of variables on which learning is performed and {known_range} is a vector of boolean to say, for each variable, if its range is known or not.

---

Member Function Documentation

template<class singleT, class filterArrayT> void plLearnDistribVector::addFilterNewPoint (  singleT  point, filterArrayT  filter, double  weight = 1.0 )  [inline]

Same as above, but {point} is a scalar value. ATTENTION: this method can only be used for one-dimensional cases. Definition at line 419 of file plLearn.h. References addFilterPoint().

template<class filterArrayT> void plLearnDistribVector::addFilterNewPoint (  const plValues &  point, filterArrayT  filter, double  weight = 1.0 )  [inline]

Adds a point {point} (represented as a {plValue}) with a given weight {weight} and updates the statistics. {filter} is an array of boolean values to say, for each learning object, if it will be updated or not. Definition at line 391 of file plLearn.h. References addFilterPoint().

---

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

• plLearn.h

---