aGrUM  0.13.2
gum::learning::ParamEstimator< IdSetAlloc, CountAlloc > Class Template Referenceabstract

The base class for estimating parameters of CPTsThe class should be used as follows: first, to speed-up computations, you should consider computing all the parameters you need in one pass. More...

#include <paramEstimator.h>

+ Inheritance diagram for gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >:
+ Collaboration diagram for gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >:

Public Member Functions

Constructors / Destructors
template<typename RowFilter >
 ParamEstimator (const RowFilter &filter, const std::vector< Size > &var_modalities, Apriori< IdSetAlloc, CountAlloc > &apriori, const ScoreInternalApriori< IdSetAlloc, CountAlloc > &score_internal_apriori)
 default constructor More...
 
virtual ParamEstimator< IdSetAlloc, CountAlloc > * copyFactory () const =0
 virtual copy factory More...
 
virtual ~ParamEstimator ()
 destructor More...
 
Accessors / Modifiers
Idx addNodeSet (Idx var)
 add a new CPT with a single variable to be estimated More...
 
Idx addNodeSet (Idx var, const std::vector< Idx > &conditioning_ids)
 add a new target variable plus some conditioning vars More...
 
void clear ()
 clears all the data structures from memory More...
 
virtual const std::vector< double, CountAlloc > & parameters (Idx nodeset_index)=0
 returns the CPT's parameters corresponding to a given nodeset More...
 
void setParameters (Idx nodeset_index, Potential< double > &pot)
 sets the CPT's parameters corresponding to a given nodeset More...
 
template<typename GUM_SCALAR >
void setParameters (Idx nodeset_index, Potential< GUM_SCALAR > &pot)
 sets the CPT's parameters corresponding to a given nodeset More...
 
void setRange (Size min_range, Size max_range)
 sets the range of records taken into account by the counter More...
 

Protected Attributes

Apriori< IdSetAlloc, CountAlloc > * _apriori
 the a priori used by the score More...
 
ScoreInternalApriori< IdSetAlloc, CountAlloc > * _score_internal_apriori
 the score that was use for structure learning (used for its apriori) More...
 
std::vector< bool_is_normalized
 indicate whether we have already normalized the parameters More...
 

Protected Member Functions

const std::vector< double, CountAlloc > & _getAllApriori (Idx index)
 returns the apriori vector for a given (conditioned) target set More...
 
const std::vector< double, CountAlloc > & _getConditioningApriori (Idx index)
 returns the apriori vector for a conditioning set More...
 
void _insertScoreApriori ()
 if needed insert the score apriori into the countings More...
 
 ParamEstimator (const ParamEstimator< IdSetAlloc, CountAlloc > &)
 copy constructor More...
 
 ParamEstimator (ParamEstimator< IdSetAlloc, CountAlloc > &&)
 move constructor More...
 

Detailed Description

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
class gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >

The base class for estimating parameters of CPTs

The class should be used as follows: first, to speed-up computations, you should consider computing all the parameters you need in one pass.

To do so, use the appropriate addNodeSet methods. These will compute everything you need. The addNodeSet methods where you do not specify a set of conditioning nodes assume that this set is empty. Once the computations have been performed, use methods _getAllCounts and _getConditioningCounts to retrieve the parameters of interest.

Definition at line 69 of file paramEstimator.h.

Constructor & Destructor Documentation

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
template<typename RowFilter >
gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::ParamEstimator ( const RowFilter &  filter,
const std::vector< Size > &  var_modalities,
Apriori< IdSetAlloc, CountAlloc > &  apriori,
const ScoreInternalApriori< IdSetAlloc, CountAlloc > &  score_internal_apriori 
)

default constructor

Parameters
filterthe row filter that will be used to read the database
var_modalitiesthe domain sizes of the variables in the database
apriorithe a priori that is taken into account in the score/countings
score_internal_aprioriThe score internal apriori.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
virtual gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::~ParamEstimator ( )
virtual

destructor

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::ParamEstimator ( const ParamEstimator< IdSetAlloc, CountAlloc > &  )
protected

copy constructor

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::ParamEstimator ( ParamEstimator< IdSetAlloc, CountAlloc > &&  )
protected

move constructor

Member Function Documentation

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::Counter< IdSetAlloc, CountAlloc >::_count ( )
protectedinherited

perform the computation of the countings

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< double, CountAlloc >& gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_getAllApriori ( Idx  index)
protected

returns the apriori vector for a given (conditioned) target set

This method returns the observation countings for the set of variables whose index was returned by method addNodeSet. If the set was conditioned, the countings correspond to the target variables and the conditioning variables. If you wish to get only the countings for the conditioning variables, prefer using method _getConditioningApriori.

Warning
the dimensions of the vector are as follows: first come the nodes of the conditioning set (in the order in which they were specified when callind addNodeset, and then the target nodes.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< double, CountAlloc >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getAllCounts ( Idx  index)
protectedinherited

returns the counting vector for a given (conditioned) target set

This method returns the observtion countings for the set of variables whose index was returned by method addNodeSet or addNodeSet. If the set was conditioned, the countings correspond to the target variables and the conditioning variables. If you wish to get only the countings for the conditioning variables, prefer using method countConditioning.

Warning
the dimensions of the vector are as follows: first come the nodes of the conditioning set (in the order in which they were specified when callind addNodeset, and then the target nodes).
whenever you call this function, if the counts have not been computed yet, they are computed before the function returns.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< Idx, IdSetAlloc >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getAllNodes ( Idx  index) const
protectednoexceptinherited

returns the set of target + conditioning nodes

conditioning nodes are always the first ones in the vector and targets are the last ones

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< std::pair< std::vector< Idx, IdSetAlloc >, Idx >* >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getAllNodes ( ) const
protectednoexceptinherited

returns all the sets of target + cond nodes, and their counting indices

conditioning nodes are always the first ones in the vector and targets are the last ones

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< double, CountAlloc >& gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_getConditioningApriori ( Idx  index)
protected

returns the apriori vector for a conditioning set

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< double, CountAlloc >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getConditioningCounts ( Idx  index)
protectedinherited

returns the counting vector for a conditioning set

Warning
whenever you call this function, if the counts have not been computed yet, they are computed before the function returns.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< Idx, IdSetAlloc >* gum::learning::Counter< IdSetAlloc, CountAlloc >::_getConditioningNodes ( Idx  index) const
protectednoexceptinherited

returns the conditioning nodes (nullptr if there are no such nodes)

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< std::pair< std::vector< Idx, IdSetAlloc >, Idx >* >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getConditioningNodes ( ) const
protectednoexceptinherited

returns all the sets of conditioning nodes

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
std::vector< std::vector< double, CountAlloc > >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_getCounts ( )
protectednoexceptinherited

returns all the countings performed (both targets and conditioned)

this method returns the countings of the record counter. It should be used in conjunction with methods _getConditioningNodes () and _getTargetNodes () that indicate, for each nodeset, the index of the corresponding counting in the vector returned by _getCounts ().

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_insertScoreApriori ( )
protected

if needed insert the score apriori into the countings

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addEmptyNodeSet ( )
inherited

adds an empty set of variables to count

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var)

add a new CPT with a single variable to be estimated

Parameters
varrepresents the index of the variable in the filtered rows produced by the database cell filters whose observations shall be counted
Returns
the index of the produced counting vector: the user should use class ParamEstimator to compute in one pass several CPT's parameters. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the observed countings of "var" in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the corresponding counting vectors.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var,
const std::vector< Idx > &  conditioning_ids 
)

add a new target variable plus some conditioning vars

Parameters
varrepresents the index of the target variable in the filtered rows produced by the database cell filters
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows
Returns
the index of the produced counting vector: the user should use class ParamEstimator to compute in one pass several CPT's parameters. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the observed countings of "var" in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the corresponding counting vectors.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var1,
Idx  var2 
)
inherited

add a new target node conditioned by another node to be counted

Parameters
var1represents the index of the target variable in the filtered rows produced by the database cell filters
var2represents the index of the conditioning variable in the filtered rows produced by the database cell filters
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores or independence tests. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the observed countings of (var2,var1) [in this order] and var2 respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( const std::pair< Idx, Idx > &  vars)
inherited

add a new target node conditioned by another node to be counted

Parameters
varscontains the index of the target variable (first) in the filtered rows produced by the database cell filters, and the index of the conditioning variable (second).
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the observed countings of (vars.second, vars.first) [in this order] and vars.second respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var,
std::vector< Idx > &&  conditioning_ids 
)
inherited

add a new target variable plus some conditioning vars

Parameters
varrepresents the index of the target variable in the filtered rows produced by the database cell filters
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores or independence tests. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the countings of (var | conditioning_ids) in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the counting vectors of (conditioning_ids,vars) [in this order] and conditioning_ids respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var1,
Idx  var2,
const std::vector< Idx > &  conditioning_ids 
)
inherited

add a target conditioned by other variables to be counted

Parameters
var1represents the index of the target variable in the filtered rows produced by the database cell filters
var2represents the index of the last conditioning variable in the filtered rows produced by the database cell filters
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows (minus var2, which is subsequently apended to it).
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the countings of (conditioning_ids, var2, var1) [in this order] and (conditioning_ids, var2) [in this order] respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( Idx  var1,
Idx  var2,
std::vector< Idx > &&  conditioning_ids 
)
inherited

add a target conditioned by other variables to be counted

Parameters
var1represents the index of the target variable in the filtered rows produced by the database cell filters
var2represents the index of the last conditioning variable in the filtered rows produced by the database cell filters
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows (minus var2, which is subsequently apended to it).
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the countings of (conditioning_ids, var2, var1) [in this order] and (conditioning_ids, var2) [in this order] respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( const std::pair< Idx, Idx > &  vars,
const std::vector< Idx > &  conditioning_ids 
)
inherited

add a target conditioned by other variables to be counted

Parameters
varsrepresents the index of the target variable (first) in the filtered rows produced by the database cell filters, and the index of the last conditioning variable (second)
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows (minus vars.second which is appended to it)
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the observed countings of (conditioning_ids, vars.second, vars.first) [in this order] and (conditioning_ids, vars.second) [in this order] respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Idx gum::learning::Counter< IdSetAlloc, CountAlloc >::addNodeSet ( const std::pair< Idx, Idx > &  vars,
std::vector< Idx > &&  conditioning_ids 
)
inherited

add a target conditioned by other variables to be counted

Parameters
varsrepresents the index of the target variable (first) in the filtered rows produced by the database cell filters, and the index of the last conditioning variable (second)
conditioning_idsthe indices of the variables of the conditioning set in the filtered rows (minus vars.second which is appended to it)
Returns
the index of the produced counting vector: the user should use class Counter to compute in one pass several scores. These and their corresponding countings in the database are stored into a vector and the value returned by method addNodeSet is the index of the counts in this vector. The user shall pass this index as argument to methods _getAllCounts and _getConditioningCounts to get the observed countings of (conditioning_ids, vars.second, vars.first) [in this order] and (conditioning_ids, vars.second) [in this order] respectively.
template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::clear ( )

clears all the data structures from memory

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
virtual ParamEstimator< IdSetAlloc, CountAlloc >* gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::copyFactory ( ) const
pure virtual

virtual copy factory

Implemented in gum::learning::ParamEstimatorML< IdSetAlloc, CountAlloc >.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< Size >& gum::learning::Counter< IdSetAlloc, CountAlloc >::modalities ( ) const
noexceptinherited

returns the modalities of the variables

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
ParamEstimator< IdSetAlloc, CountAlloc >& gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::operator= ( const ParamEstimator< IdSetAlloc, CountAlloc > &  )
privatedelete

prevent copy operator

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
virtual const std::vector< double, CountAlloc >& gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::parameters ( Idx  nodeset_index)
pure virtual

returns the CPT's parameters corresponding to a given nodeset

The vector contains the parameters of an n-dimensional CPT. The distribution of the dimensions of the CPT within the vector is as follows: first, there are the conditioning nodes (in the order in which they were specified) and, then, the target node.

Implemented in gum::learning::ParamEstimatorML< IdSetAlloc, CountAlloc >.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::Counter< IdSetAlloc, CountAlloc >::setMaxNbThreads ( Size  nb)
noexceptinherited

sets the maximum number of threads used to perform countings

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::setParameters ( Idx  nodeset_index,
Potential< double > &  pot 
)

sets the CPT's parameters corresponding to a given nodeset

The order of the variables in the potential and in the nodeset are assumed to be identical

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
template<typename GUM_SCALAR >
void gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::setParameters ( Idx  nodeset_index,
Potential< GUM_SCALAR > &  pot 
)

sets the CPT's parameters corresponding to a given nodeset

The order of the variables in the potential and in the nodeset are assumed to be identical

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
void gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::setRange ( Size  min_range,
Size  max_range 
)

sets the range of records taken into account by the counter

Parameters
min_rangehe number of the first record to be taken into account during learning
max_rangethe number of the record after the last one taken into account

Member Data Documentation

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const double gum::learning::Counter< IdSetAlloc, CountAlloc >::_1log2 {M_LOG2E}
protectedinherited

1 / log(2)

Definition at line 342 of file counter.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
bool gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::__apriori_computed {false}
private

has the a priori been computed

Definition at line 241 of file paramEstimator.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
bool gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::__score_apriori_inserted {false}
private

has the score's internal apriori been inserted into the countings ?

Definition at line 244 of file paramEstimator.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
Apriori< IdSetAlloc, CountAlloc >* gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_apriori
protected

the a priori used by the score

Definition at line 177 of file paramEstimator.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
std::vector< std::pair< std::vector< Idx, IdSetAlloc >, Idx >* > gum::learning::Counter< IdSetAlloc, CountAlloc >::_conditioning_nodesets
protectedinherited

the conditioning id sets to count and their indices in the record counter

Definition at line 361 of file counter.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
bool gum::learning::Counter< IdSetAlloc, CountAlloc >::_counts_computed {false}
protectedinherited

indicates whether we have already computed the countings of the nodesets

Definition at line 349 of file counter.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
std::vector< bool > gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_is_normalized
protected

indicate whether we have already normalized the parameters

Definition at line 183 of file paramEstimator.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
const std::vector< Size >& gum::learning::Counter< IdSetAlloc, CountAlloc >::_modalities
protectedinherited

the modalities of the variables

Definition at line 345 of file counter.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
RecordCounter< IdSetAlloc, CountAlloc > gum::learning::Counter< IdSetAlloc, CountAlloc >::_record_counter
protectedinherited

the recordCounter that will parse the database

Definition at line 352 of file counter.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
ScoreInternalApriori< IdSetAlloc, CountAlloc >* gum::learning::ParamEstimator< IdSetAlloc, CountAlloc >::_score_internal_apriori
protected

the score that was use for structure learning (used for its apriori)

Definition at line 180 of file paramEstimator.h.

template<typename IdSetAlloc = std::allocator< Idx >, typename CountAlloc = std::allocator< double >>
std::vector< std::pair< std::vector< Idx, IdSetAlloc >, Idx >* > gum::learning::Counter< IdSetAlloc, CountAlloc >::_target_nodesets
protectedinherited

the target id sets to count and their indices in the record counter

Definition at line 356 of file counter.h.


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