Main Page | Class Hierarchy | Alphabetical List | Class List | Directories | File List | Class Members

plLearnUnKLaplace Class Reference

Class for learning Laplace's parametrical form for variables wich definition set is unknown. More...

#include <plLearnUnKLaplace.h>

Inheritance diagram for plLearnUnKLaplace:

Inheritance graph
[legend]
Collaboration diagram for plLearnUnKLaplace:

Collaboration graph
[legend]
List of all members.

Public Member Functions

 plLearnUnKLaplace (const plLearnUnKLaplace &plLUKH)
 The copy-constructor.
 plLearnUnKLaplace (const plVariablesConjunction &vars, map< plDataValues, plLearnFrequence > *const initial, bool copie=false)
 init is a pointer on an map of initial values of frequencies.
 plLearnUnKLaplace (const plVariablesConjunction &vars, const map< plDataValues, plLearnFrequence > &initial)
 An other way to specifie that the map must be copied.
void get_plProbValue (map< plDataValues, plProbValue > &prob_map_ref) const
 If the given map is empty : return a map of plProbValues with the actual learn probabilities.
void get_plProbValue (map< plDataValues, plProbValue > &prob_map_ref, plLearnClassesTotal cardinality) const
 If the given map is empty : return a map of plProbValues with the actual learn probabilities.
void get_plProbValue (map< plDataValues, plProbValue > &prob_map_ref, plLearnClassesTotal cardinality, double alpha) const
 If the given map is empty : return a map of plProbValues with the actual learn probabilities.
void select_plProbValue (vector< plProbValue > &prob_ref, const vector< plValues > &selected_values) const
 The selected values are given in a vector of plVlaues and a vector of plProbValues of this range id filled with the actual learn probabilities.
plProbValue * select_plProbValue (const vector< plValues > &selected_values) const
 The selected values are given in a vector of plVlaues and an array of plProbValues of this range with the actual learn probabilities is returned.
void select_plProbValue (vector< plProbValue > &prob_ref, const vector< plValues > &selected_values, double alpha) const
 The selected values are given in a vector of plVlaues and a vector of plProbValues of this range id filled with the actual learn probabilities.
plProbValue * select_plProbValue (const vector< plValues > &selected_values, double alpha) const
 The selected values are given in a vector of plVlaues and an array of plProbValues of this range with the actual learn probabilities is returned.
void get_plProbValue (vector< plProbValue > &prob_ref, const plValues &min, const plValues &max) const
 The boundaries are given in min and max and a table for this range is filled with the actual probabilites.
void get_plProbValue (vector< plProbValue > &prob_ref, const plValues &min, const plValues &max, double alpha) const
 The boundaries are given in min and max and a table for this range is filled with the actual probabilites for the Lidstone form : p(i) = ( alpha+freq(i) ) / (alpha*cardinality+_total_weight) .
plProbValue * get_plProbValue (const plValues &min, const plValues &max) const
 The boundaries are given in min and max and an array of plProbValues of this range with the actual learn probabilities is returned.
plProbValue * get_plProbValue (const plValues &min, const plValues &max, double alpha) const
 The boundaries are given in min and max and an array of plProbValues of this range with the actual learn probabilities is returned.
plKernel get_distribution (const void *parameters=NULL) const
 Returns the probability table corresponding to the learnt distribution.

Detailed Description

Class for learning Laplace's parametrical form for variables wich definition set is unknown.

Definition at line 53 of file plLearnUnKLaplace.h.

Constructor & Destructor Documentation

plLearnUnKLaplace::plLearnUnKLaplace (  const plVariablesConjunction &  vars,
map< plDataValues, plLearnFrequence > *const   initial,
bool  copie = false
) 
 

init is a pointer on an map of initial values of frequencies.

The map is copied when the flag is on true. If the flag copied is positionned on false the object access the map passed when the initialization. By default the map is not copied.

Member Function Documentation

plProbValue* plLearnUnKLaplace::get_plProbValue (  const plValues &  min,
const plValues &  max,
double  alpha
)  const
 

The boundaries are given in min and max and an array of plProbValues of this range with the actual learn probabilities is returned.

The result is normalized. The distribution is of the Lidstone form : p(i) = ( alpha+freq(i) ) / (alpha*cardinality+_total_weight)

plProbValue* plLearnUnKLaplace::get_plProbValue (  const plValues &  min,
const plValues &  max
)  const [virtual]
 

The boundaries are given in min and max and an array of plProbValues of this range with the actual learn probabilities is returned.

The result is normalized.

Reimplemented from plLearnUnKHistogram.

void plLearnUnKLaplace::get_plProbValue (  map< plDataValues, plProbValue > &  prob_map_ref,
plLearnClassesTotal  cardinality,
double  alpha
)  const
 

If the given map is empty : return a map of plProbValues with the actual learn probabilities.

The sum of the inserted plProbValues is not normalized. It misses : cardinality-histo.size() points with proba : (cardinality-histo.size())/(cardinality+_total_weight). The cardinality must be given. The parameter alpha allow you to fit the weight of the data : the parametrical form is : p(i) = ( alpha+freq(i) ) / (alpha*cardinality+_total_weight) It is the Lidstone Low

void plLearnUnKLaplace::get_plProbValue (  map< plDataValues, plProbValue > &  prob_map_ref,
plLearnClassesTotal  cardinality
)  const
 

If the given map is empty : return a map of plProbValues with the actual learn probabilities.

The sum of the inserted plProbValues is not normalized. It misses : cardinality-histo.size() points with proba : (cardinality-histo.size())/(cardinality+_total_weight). The cardinality is given.

void plLearnUnKLaplace::get_plProbValue (  map< plDataValues, plProbValue > &  prob_map_ref  )  const [virtual]
 

If the given map is empty : return a map of plProbValues with the actual learn probabilities.

The sum of the inserted plProbValues is normalized. The cardinality is exactly the size of the map histo.

Reimplemented from plLearnUnKHistogram.

Referenced by get_distribution().

plProbValue* plLearnUnKLaplace::select_plProbValue (  const vector< plValues > &  selected_values,
double  alpha
)  const
 

The selected values are given in a vector of plVlaues and an array of plProbValues of this range with the actual learn probabilities is returned.

The result is normalized.

void plLearnUnKLaplace::select_plProbValue (  vector< plProbValue > &  prob_ref,
const vector< plValues > &  selected_values,
double  alpha
)  const
 

The selected values are given in a vector of plVlaues and a vector of plProbValues of this range id filled with the actual learn probabilities.

The result is normalized.

plProbValue* plLearnUnKLaplace::select_plProbValue (  const vector< plValues > &  selected_values  )  const [virtual]
 

The selected values are given in a vector of plVlaues and an array of plProbValues of this range with the actual learn probabilities is returned.

The result is normalized.

Reimplemented from plLearnUnKHistogram.

void plLearnUnKLaplace::select_plProbValue (  vector< plProbValue > &  prob_ref,
const vector< plValues > &  selected_values
)  const [virtual]
 

The selected values are given in a vector of plVlaues and a vector of plProbValues of this range id filled with the actual learn probabilities.

The result is normalized.

Reimplemented from plLearnUnKHistogram.

Referenced by get_distribution().

The documentation for this class was generated from the following file:
Generated on Fri Apr 1 10:59:20 2005 for ProBT by  doxygen 1.4.1