<agrum/FMDP/learning/datastructure/incrementalGraphLearner> More...

#include <incrementalGraphLearner.h>

Inheritance diagram for gum::IncrementalGraphLearner< AttributeSelection, isScalar >:

Collaboration diagram for gum::IncrementalGraphLearner< AttributeSelection, isScalar >:

Public Member Functions
Size	size ()
Function Graph Updating methods
virtual void	updateFunctionGraph ()=0
	Updates target to currently learned graph structure.
Visit Methods
NodeId	root () const
bool	isTerminal (NodeId ni) const
const DiscreteVariable *	nodeVar (NodeId ni) const
NodeId	nodeSon (NodeId ni, Idx modality) const
Idx	nodeNbObservation (NodeId ni) const
virtual void	insertSetOfVars (MultiDimFunctionGraph< double > *ret) const

Protected Attributes
MultiDimFunctionGraph< double > *	target_
	The final diagram we're building.
gum::VariableSet	setOfVars_
const DiscreteVariable *	value_
Sequence< ValueType >	valueAssumed_
bool	needUpdate_
Model handling datastructures
NodeGraphPart	model_
	The source of nodeId.
NodeId	root_
	The root of the ordered tree.
HashTable< NodeId, const DiscreteVariable * >	nodeVarMap_
	Gives for any node its associated variable.
HashTable< NodeId, NodeId * >	nodeSonsMap_
	A table giving for any node a table mapping to its son idx is the modality of associated variable.
HashTable< const DiscreteVariable , LinkedList< NodeId > >	var2Node_
	Associates to any variable the list of all nodes associated to this variable.
HashTable< NodeId, NodeDatabase< AttributeSelection, isScalar > * >	nodeId2Database_
	This hashtable binds every node to an associated NodeDatabase which handles every observation that concerns that node.
HashTable< NodeId, Set< const Observation * > * >	leafDatabase_
	This hashtable binds to every leaf an associated set of all hte observations compatible with it.

Private Types
using	ValueType = typename ValueSelect< isScalar, double, Idx >::type

Constructor & destructor.
	IncrementalGraphLearner (MultiDimFunctionGraph< double > target, gum::VariableSet attributesSet, const DiscreteVariable learnVariable)
	Default constructor.
virtual	~IncrementalGraphLearner ()
	Default destructor.
void	_clearValue_ ()
	Template function dispatcher.
void	_clearValue_ (Int2Type< true >)
	In the case where we're learning a function of real values this has to be wiped out upon destruction (to be deprecated).
void	_clearValue_ (Int2Type< false >)
	In case where we're learning function of variable behaviour, this should do nothing.

New Observation insertion methods
virtual void	addObservation (const Observation *obs)
	Inserts a new observation.
void	_assumeValue_ (const Observation *obs)
	Get value assumed by studied variable for current observation.
void	_assumeValue_ (const Observation *obs, Int2Type< true >)
	Inserts a new observation.
void	_assumeValue_ (const Observation *obs, Int2Type< false >)
	Inserts a new observation.
Idx	_branchObs_ (const Observation obs, const DiscreteVariable var)
	Seek modality assumed in obs for given var.
Idx	_branchObs_ (const Observation obs, const DiscreteVariable var, Int2Type< true >)
	Inserts a new observation.
Idx	_branchObs_ (const Observation obs, const DiscreteVariable var, Int2Type< false >)
	Inserts a new observation.
virtual void	updateNodeWithObservation_ (const Observation *newObs, NodeId currentNodeId)
	Will update internal graph's NodeDatabase of given node with the new observation.

Graph Structure update methods
virtual void	updateVar (const DiscreteVariable *)
	If a new modality appears to exists for given variable, call this method to turn every associated node to this variable into leaf. Graph has then indeed to be revised.
virtual void	updateGraph ()=0
	Updates the tree after a new observation has been added.
void	updateNode_ (NodeId nody, gum::VariableSet &bestVars)
	From the given sets of node, selects randomly one and installs it on given node.
virtual void	convertNode2Leaf_ (NodeId)
	Turns the given node into a leaf if not already so.
virtual void	transpose_ (NodeId, const DiscreteVariable *)
	Installs given variable to the given node, ensuring that the variable is not present in its subtree.
virtual NodeId	insertNode_ (NodeDatabase< AttributeSelection, isScalar > nDB, const DiscreteVariable boundVar)
	inserts a new node in internal graph
virtual NodeId	insertInternalNode_ (NodeDatabase< AttributeSelection, isScalar > nDB, const DiscreteVariable boundVar, NodeId *sonsMap)
	inserts a new internal node in internal graph
virtual NodeId	insertLeafNode_ (NodeDatabase< AttributeSelection, isScalar > nDB, const DiscreteVariable boundVar, Set< const Observation * > *obsSet)
	inserts a new leaf node in internal graohs
virtual void	chgNodeBoundVar_ (NodeId chgedNodeId, const DiscreteVariable *desiredVar)
	Changes the associated variable of a node.
virtual void	removeNode_ (NodeId removedNodeId)
	Removes a node from the internal graph.

Detailed Description

template<TESTNAME AttributeSelection, bool isScalar = false>
class gum::IncrementalGraphLearner< AttributeSelection, isScalar >

Abstract class for incrementally learn a graphical representation of a function. Can handle both function of real values, and function explaining the behaviour of a variable given set of other variables (as typically in conditional probabilities)

Maintains two graph in memory, one which is incrementally updated and the other one which is updated on demand and is usable by the outside.

Definition at line 83 of file incrementalGraphLearner.h.

Member Typedef Documentation

◆ ValueType

template<TESTNAME AttributeSelection, bool isScalar = false>

using gum::IncrementalGraphLearner< AttributeSelection, isScalar >::ValueType = typename ValueSelect< isScalar, double, Idx >::type

private

Definition at line 84 of file incrementalGraphLearner.h.

Constructor & Destructor Documentation

◆ IncrementalGraphLearner()

template<TESTNAME AttributeSelection, bool isScalar>

gum::IncrementalGraphLearner< AttributeSelection, isScalar >::IncrementalGraphLearner	(	MultiDimFunctionGraph< double > *	target,
		gum::VariableSet	varList,
		const DiscreteVariable *	value )

Default constructor.

Parameters

target	: the output diagram usable by the outside
attributesSet	: set of variables from which we try to describe the learned function
learnVariable	: if we tried to learn a the behaviour of a variable given variable given another set of variables, this is the one. If we are learning a function of real value, this is just a computationnal trick (and is to be deprecated)

Definition at line 85 of file incrementalGraphLearner_tpl.h.

                                     : target_(target), setOfVars_(varList), value_(value) {
    GUM_CONSTRUCTOR(IncrementalGraphLearner);
 
    for (auto varIter = setOfVars_.cbeginSafe(); varIter != setOfVars_.cendSafe(); ++varIter)
      var2Node_.insert(*varIter, new LinkedList< NodeId >());
    var2Node_.insert(value_, new LinkedList< NodeId >());
 
    model_.addNode();
    this->root_ = IncrementalGraphLearner< AttributeSelection, isScalar >::insertLeafNode_(
        new NodeDatabase< AttributeSelection, isScalar >(&setOfVars_, value_),
        value_,
        new Set< const Observation* >());
  }

References IncrementalGraphLearner(), insertLeafNode_(), model_, root_, setOfVars_, target_, value_, and var2Node_.

Referenced by IncrementalGraphLearner().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ ~IncrementalGraphLearner()

template<TESTNAME AttributeSelection, bool isScalar>

gum::IncrementalGraphLearner< AttributeSelection, isScalar >::~IncrementalGraphLearner ( )

virtual

Default destructor.

Definition at line 106 of file incrementalGraphLearner_tpl.h.

                                                                                    {
    for (auto nodeIter = nodeId2Database_.beginSafe(); nodeIter != nodeId2Database_.endSafe();
         ++nodeIter)
      delete nodeIter.val();
 
    for (auto nodeIter = nodeSonsMap_.beginSafe(); nodeIter != nodeSonsMap_.endSafe(); ++nodeIter)
      SOA_DEALLOCATE(nodeIter.val(), sizeof(NodeId) * nodeVarMap_[nodeIter.key()]->domainSize());
 
    for (auto varIter = var2Node_.beginSafe(); varIter != var2Node_.endSafe(); ++varIter)
      delete varIter.val();
 
    for (auto nodeIter = leafDatabase_.beginSafe(); nodeIter != leafDatabase_.endSafe(); ++nodeIter)
      delete nodeIter.val();
 
    _clearValue_();
 
    GUM_DESTRUCTOR(IncrementalGraphLearner);
  }

References nodeId2Database_.

Member Function Documentation

◆ _assumeValue_() [1/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_assumeValue_ ( const Observation * obs )

inlineprivate

Get value assumed by studied variable for current observation.

Definition at line 153 of file incrementalGraphLearner.h.

153{ _assumeValue_(obs, Int2Type< isScalar >()); }

gum::IncrementalGraphLearner::_assumeValue_

void _assumeValue_(const Observation *obs)

Get value assumed by studied variable for current observation.

Definition incrementalGraphLearner.h:153

References _assumeValue_().

Referenced by _assumeValue_(), and addObservation().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ _assumeValue_() [2/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_assumeValue_	(	const Observation *	obs,
		Int2Type< false >	)

inlineprivate

Inserts a new observation.

Parameters

the	new observation to learn

Definition at line 159 of file incrementalGraphLearner.h.

                                                                  {
      if (!valueAssumed_.exists(obs->modality(value_))) valueAssumed_ << obs->modality(value_);
    }

References gum::Observation::modality(), value_, and valueAssumed_.

Here is the call graph for this function:

◆ _assumeValue_() [3/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_assumeValue_	(	const Observation *	obs,
		Int2Type< true >	)

inlineprivate

Inserts a new observation.

Parameters

the	new observation to learn

Definition at line 155 of file incrementalGraphLearner.h.

                                                                 {
      if (!valueAssumed_.exists(obs->reward())) valueAssumed_ << obs->reward();
    }

References gum::Observation::reward(), and valueAssumed_.

Here is the call graph for this function:

◆ _branchObs_() [1/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

Idx gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_branchObs_	(	const Observation *	obs,
		const DiscreteVariable *	var )

inlineprivate

Seek modality assumed in obs for given var.

Definition at line 168 of file incrementalGraphLearner.h.

                                                                         {
      return _branchObs_(obs, var, Int2Type< isScalar >());
    }

References _branchObs_().

Referenced by _branchObs_().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ _branchObs_() [2/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

Idx gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_branchObs_	(	const Observation *	obs,
		const DiscreteVariable *	var,
		Int2Type< false >	)

inlineprivate

Inserts a new observation.

Parameters

the	new observation to learn

Definition at line 176 of file incrementalGraphLearner.h.

                                                                                            {
      return obs->modality(var);
    }

References gum::Observation::modality().

Here is the call graph for this function:

◆ _branchObs_() [3/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

Idx gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_branchObs_	(	const Observation *	obs,
		const DiscreteVariable *	var,
		Int2Type< true >	)

inlineprivate

Inserts a new observation.

Parameters

the	new observation to learn

Definition at line 172 of file incrementalGraphLearner.h.

                                                                                           {
      return obs->rModality(var);
    }

References gum::Observation::rModality().

Here is the call graph for this function:

◆ _clearValue_() [1/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_clearValue_ ( )

inlineprivate

Template function dispatcher.

Definition at line 117 of file incrementalGraphLearner.h.

117{ _clearValue_(Int2Type< isScalar >()); }

References _clearValue_().

Referenced by _clearValue_().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ _clearValue_() [2/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_clearValue_ ( Int2Type< false > )

inlineprivate

In case where we're learning function of variable behaviour, this should do nothing.

Definition at line 129 of file incrementalGraphLearner.h.

129{}

◆ _clearValue_() [3/3]

template<TESTNAME AttributeSelection, bool isScalar = false>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::_clearValue_ ( Int2Type< true > )

inlineprivate

In the case where we're learning a function of real values this has to be wiped out upon destruction (to be deprecated).

Definition at line 123 of file incrementalGraphLearner.h.

123{ delete value_; }

References value_.

◆ addObservation()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::addObservation ( const Observation * newObs )

virtual

Inserts a new observation.

Parameters

the	new observation to learn

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

Definition at line 136 of file incrementalGraphLearner_tpl.h.

                                 {
    _assumeValue_(newObs);
 
    // The we go across the tree
    NodeId currentNodeId = root_;
 
    while (nodeSonsMap_.exists(currentNodeId)) {
      // On each encountered node, we update the database
      updateNodeWithObservation_(newObs, currentNodeId);
 
      // The we select the next to go throught
      currentNodeId = nodeSonsMap_[currentNodeId][_branchObs_(newObs, nodeVarMap_[currentNodeId])];
    }
 
    // On final insertion into the leave we reach
    updateNodeWithObservation_(newObs, currentNodeId);
    leafDatabase_[currentNodeId]->insert(newObs);
  }

References _assumeValue_(), nodeSonsMap_, and root_.

Referenced by gum::IMDDI< AttributeSelection, isScalar >::addObservation(), and gum::ITI< AttributeSelection, isScalar >::addObservation().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ chgNodeBoundVar_()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::chgNodeBoundVar_	(	NodeId	currentNodeId,
		const DiscreteVariable *	desiredVar )

protectedvirtual

Changes the associated variable of a node.

Parameters

chgedNodeId	: the node to change
desiredVar	: its new associated variable

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

Definition at line 412 of file incrementalGraphLearner_tpl.h.

                                          {
    if (nodeVarMap_[currentNodeId] == desiredVar) return;
 
    var2Node_[nodeVarMap_[currentNodeId]]->searchAndRemoveLink(currentNodeId);
    var2Node_[desiredVar]->addLink(currentNodeId);
    nodeVarMap_[currentNodeId] = desiredVar;
 
    if (nodeVarMap_[currentNodeId] != value_ && leafDatabase_.exists(currentNodeId)) {
      delete leafDatabase_[currentNodeId];
      leafDatabase_.erase(currentNodeId);
    }
 
    if (nodeVarMap_[currentNodeId] == value_ && !leafDatabase_.exists(currentNodeId)) {
      leafDatabase_.insert(currentNodeId, new Set< const Observation* >());
    }
 
    needUpdate_ = true;
  }

References leafDatabase_, needUpdate_, nodeVarMap_, value_, and var2Node_.

Referenced by gum::IMDDI< AttributeSelection, isScalar >::chgNodeBoundVar_(), and gum::ITI< AttributeSelection, isScalar >::chgNodeBoundVar_().

Here is the caller graph for this function:

◆ convertNode2Leaf_()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::convertNode2Leaf_ ( NodeId currentNodeId )

protectedvirtual

Turns the given node into a leaf if not already so.

Definition at line 217 of file incrementalGraphLearner_tpl.h.

                            {
    if (nodeVarMap_[currentNodeId] != value_) {
      leafDatabase_.insert(currentNodeId, new Set< const Observation* >());
 
      // Resolving tensor sons issue
      for (Idx modality = 0; modality < nodeVarMap_[currentNodeId]->domainSize(); ++modality) {
        NodeId sonId = nodeSonsMap_[currentNodeId][modality];
        convertNode2Leaf_(sonId);
        (*leafDatabase_[currentNodeId]) = (*leafDatabase_[currentNodeId]) + *(leafDatabase_[sonId]);
        removeNode_(sonId);
      }
 
      SOA_DEALLOCATE(nodeSonsMap_[currentNodeId],
                     sizeof(NodeId) * nodeVarMap_[currentNodeId]->domainSize());
      nodeSonsMap_.erase(currentNodeId);
 
      chgNodeBoundVar_(currentNodeId, value_);
    }
  }

References leafDatabase_, nodeSonsMap_, nodeVarMap_, and value_.

Referenced by updateNode_(), and updateVar().

Here is the caller graph for this function:

◆ insertInternalNode_()

template<TESTNAME AttributeSelection, bool isScalar>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::insertInternalNode_	(	NodeDatabase< AttributeSelection, isScalar > *	nDB,
		const DiscreteVariable *	boundVar,
		NodeId *	sonsMap )

protectedvirtual

inserts a new internal node in internal graph

Parameters

nDB	: the associated database
boundVar	: the associated variable
sonsMap	: a table giving node's sons node

Returns: the newly created node's id

Definition at line 376 of file incrementalGraphLearner_tpl.h.

                                                             {
    NodeId newNodeId = this->insertNode_(nDB, boundVar);
    nodeSonsMap_.insert(newNodeId, sonsMap);
    return newNodeId;
  }

References insertNode_(), and nodeSonsMap_.

Here is the call graph for this function:

◆ insertLeafNode_()

template<TESTNAME AttributeSelection, bool isScalar>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::insertLeafNode_	(	NodeDatabase< AttributeSelection, isScalar > *	nDB,
		const DiscreteVariable *	boundVar,
		Set< const Observation * > *	obsSet )

protectedvirtual

inserts a new leaf node in internal graohs

Parameters

nDB	: the associated database
boundVar	: the associated variable
obsSet	: the set of observation this leaf retains

Returns: the newly created node's id

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >.

Definition at line 395 of file incrementalGraphLearner_tpl.h.

                                                            {
    NodeId newNodeId = this->insertNode_(nDB, boundVar);
    leafDatabase_.insert(newNodeId, obsSet);
    return newNodeId;
  }

References insertNode_(), and leafDatabase_.

Referenced by IncrementalGraphLearner(), and gum::IMDDI< AttributeSelection, isScalar >::insertLeafNode_().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ insertNode_()

template<TESTNAME AttributeSelection, bool isScalar>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::insertNode_	(	NodeDatabase< AttributeSelection, isScalar > *	nDB,
		const DiscreteVariable *	boundVar )

protectedvirtual

inserts a new node in internal graph

Parameters

nDB	: the associated database
boundVar	: the associated variable

Returns: the newly created node's id

Reimplemented in gum::ITI< AttributeSelection, isScalar >.

Definition at line 353 of file incrementalGraphLearner_tpl.h.

                                                              {
    NodeId newNodeId = model_.addNode();
    nodeVarMap_.insert(newNodeId, boundVar);
    nodeId2Database_.insert(newNodeId, nDB);
    var2Node_[boundVar]->addLink(newNodeId);
 
    needUpdate_ = true;
 
    return newNodeId;
  }

References model_, needUpdate_, nodeId2Database_, nodeVarMap_, and var2Node_.

Referenced by insertInternalNode_(), insertLeafNode_(), and gum::ITI< AttributeSelection, isScalar >::insertNode_().

Here is the caller graph for this function:

◆ insertSetOfVars()

template<TESTNAME AttributeSelection, bool isScalar = false>

virtual void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::insertSetOfVars ( MultiDimFunctionGraph< double > * ret ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >.

Definition at line 349 of file incrementalGraphLearner.h.

                                                                             {
      for (SetIteratorSafe< const DiscreteVariable* > varIter = setOfVars_.beginSafe();
           varIter != setOfVars_.endSafe();
           ++varIter)
        ret->add(**varIter);
    }

References gum::MultiDimFunctionGraph< GUM_SCALAR, TerminalNodePolicy >::add(), and setOfVars_.

Here is the call graph for this function:

◆ isTerminal()

template<TESTNAME AttributeSelection, bool isScalar = false>

bool gum::IncrementalGraphLearner< AttributeSelection, isScalar >::isTerminal ( NodeId ni ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Definition at line 329 of file incrementalGraphLearner.h.

329{ return !this->nodeSonsMap_.exists(ni); }

References nodeSonsMap_.

◆ nodeNbObservation()

template<TESTNAME AttributeSelection, bool isScalar = false>

Idx gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeNbObservation ( NodeId ni ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Definition at line 344 of file incrementalGraphLearner.h.

344{ return this->nodeId2Database_[ni]->nbObservation(); }

References nodeId2Database_.

◆ nodeSon()

template<TESTNAME AttributeSelection, bool isScalar = false>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeSon	(	NodeId	ni,
		Idx	modality ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Definition at line 339 of file incrementalGraphLearner.h.

339{ return this->nodeSonsMap_[ni][modality]; }

References nodeSonsMap_.

◆ nodeVar()

template<TESTNAME AttributeSelection, bool isScalar = false>

const DiscreteVariable * gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeVar ( NodeId ni ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Definition at line 334 of file incrementalGraphLearner.h.

334{ return this->nodeVarMap_[ni]; }

References nodeVarMap_.

◆ removeNode_()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::removeNode_ ( NodeId currentNodeId )

protectedvirtual

Removes a node from the internal graph.

Parameters

removedNodeId : the node to remove

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

Definition at line 440 of file incrementalGraphLearner_tpl.h.

                                                                                                {
    // Retriat de l'id
    model_.eraseNode(currentNodeId);
 
    // Retrait du vecteur fils
    if (nodeSonsMap_.exists(currentNodeId)) {
      SOA_DEALLOCATE(nodeSonsMap_[currentNodeId],
                     sizeof(NodeId) * nodeVarMap_[currentNodeId]->domainSize());
      nodeSonsMap_.erase(currentNodeId);
    }
 
    if (leafDatabase_.exists(currentNodeId)) {
      delete leafDatabase_[currentNodeId];
      leafDatabase_.erase(currentNodeId);
    }
 
    // Retrait de la variable
    var2Node_[nodeVarMap_[currentNodeId]]->searchAndRemoveLink(currentNodeId);
    nodeVarMap_.erase(currentNodeId);
 
    // Retrait du NodeDatabase
    delete nodeId2Database_[currentNodeId];
    nodeId2Database_.erase(currentNodeId);
 
    needUpdate_ = true;
  }

References leafDatabase_, model_, needUpdate_, nodeId2Database_, nodeSonsMap_, nodeVarMap_, SOA_DEALLOCATE, and var2Node_.

Referenced by gum::IMDDI< AttributeSelection, isScalar >::removeNode_(), and gum::ITI< AttributeSelection, isScalar >::removeNode_().

Here is the caller graph for this function:

◆ root()

template<TESTNAME AttributeSelection, bool isScalar = false>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::root ( ) const

inlinevirtual

Implements gum::IVisitableGraphLearner.

Definition at line 324 of file incrementalGraphLearner.h.

324{ return this->root_; }

References root_.

◆ size()

template<TESTNAME AttributeSelection, bool isScalar = false>

Size gum::IncrementalGraphLearner< AttributeSelection, isScalar >::size ( )

inline

Definition at line 313 of file incrementalGraphLearner.h.

313{ return nodeVarMap_.size(); }

References nodeVarMap_.

◆ transpose_()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::transpose_	(	NodeId	currentNodeId,
		const DiscreteVariable *	desiredVar )

protectedvirtual

Installs given variable to the given node, ensuring that the variable is not present in its subtree.

Definition at line 243 of file incrementalGraphLearner_tpl.h.

                                          {
    // **************************************************************************************
    // Si le noeud courant contient déjà la variable qu'on souhaite lui amener
    // Il n'y a rien à faire
    if (nodeVarMap_[currentNodeId] == desiredVar) { return; }
 
    // **************************************************************************************
    // Si le noeud courant est terminal,
    // Il faut artificiellement insérer un noeud liant à la variable
    if (nodeVarMap_[currentNodeId] == value_) {
      // We turned this leaf into an internal node.
      // This mean that we'll need to install children leaves for each value of
      // desiredVar
 
      // First We must prepare these new leaves NodeDatabases and Sets<const
      // Observation*>
      NodeDatabase< AttributeSelection, isScalar >** dbMap
          = static_cast< NodeDatabase< AttributeSelection, isScalar >** >(SOA_ALLOCATE(
              sizeof(NodeDatabase< AttributeSelection, isScalar >*) * desiredVar->domainSize()));
      Set< const Observation* >** obsetMap = static_cast< Set< const Observation* >** >(
          SOA_ALLOCATE(sizeof(Set< const Observation* >*) * desiredVar->domainSize()));
      for (Idx modality = 0; modality < desiredVar->domainSize(); ++modality) {
        dbMap[modality]    = new NodeDatabase< AttributeSelection, isScalar >(&setOfVars_, value_);
        obsetMap[modality] = new Set< const Observation* >();
      }
      for (SetIteratorSafe< const Observation* > obsIter
           = leafDatabase_[currentNodeId]->beginSafe();
           leafDatabase_[currentNodeId]->endSafe() != obsIter;
           ++obsIter) {
        dbMap[_branchObs_(*obsIter, desiredVar)]->addObservation(*obsIter);
        obsetMap[_branchObs_(*obsIter, desiredVar)]->insert(*obsIter);
      }
 
      // Then we can install each new leaves (and put in place the sonsMap)
      NodeId* sonsMap
          = static_cast< NodeId* >(SOA_ALLOCATE(sizeof(NodeId) * desiredVar->domainSize()));
      for (Idx modality = 0; modality < desiredVar->domainSize(); ++modality)
        sonsMap[modality] = insertLeafNode_(dbMap[modality], value_, obsetMap[modality]);
 
      // Some necessary clean up
      SOA_DEALLOCATE(dbMap,
                     sizeof(NodeDatabase< AttributeSelection, isScalar >*)
                         * desiredVar->domainSize());
      SOA_DEALLOCATE(obsetMap, sizeof(Set< const Observation* >*) * desiredVar->domainSize());
 
      // And finally we can turn the node into an internal node associated to
      // desiredVar
      chgNodeBoundVar_(currentNodeId, desiredVar);
      nodeSonsMap_.insert(currentNodeId, sonsMap);
 
      return;
    }
 
    // *************************************************************************************
    // Remains the general case where currentNodeId is an internal node.
 
    // First we ensure that children node use desiredVar as variable
    for (Idx modality = 0; modality < nodeVarMap_[currentNodeId]->domainSize(); ++modality)
      transpose_(nodeSonsMap_[currentNodeId][modality], desiredVar);
 
    //        Sequence<NodeDatabase<AttributeSelection, isScalar>*>
    //        sonsNodeDatabase =
    //        nodeId2Database_[currentNodeId]->splitOnVar(desiredVar);
    NodeId* sonsMap
        = static_cast< NodeId* >(SOA_ALLOCATE(sizeof(NodeId) * desiredVar->domainSize()));
 
    // Then we create the new mapping
    for (Idx desiredVarModality = 0; desiredVarModality < desiredVar->domainSize();
         ++desiredVarModality) {
      NodeId* grandSonsMap = static_cast< NodeId* >(
          SOA_ALLOCATE(sizeof(NodeId) * nodeVarMap_[currentNodeId]->domainSize()));
      NodeDatabase< AttributeSelection, isScalar >* sonDB
          = new NodeDatabase< AttributeSelection, isScalar >(&setOfVars_, value_);
      for (Idx currentVarModality = 0;
           currentVarModality < nodeVarMap_[currentNodeId]->domainSize();
           ++currentVarModality) {
        grandSonsMap[currentVarModality]
            = nodeSonsMap_[nodeSonsMap_[currentNodeId][currentVarModality]][desiredVarModality];
        sonDB->operator+=((*nodeId2Database_[grandSonsMap[currentVarModality]]));
      }
 
      sonsMap[desiredVarModality]
          = insertInternalNode_(sonDB, nodeVarMap_[currentNodeId], grandSonsMap);
    }
 
    // Finally we clean the old remaining nodes
    for (Idx currentVarModality = 0; currentVarModality < nodeVarMap_[currentNodeId]->domainSize();
         ++currentVarModality) {
      removeNode_(nodeSonsMap_[currentNodeId][currentVarModality]);
    }
 
    // We suppress the old sons map and remap to the new one
    SOA_DEALLOCATE(nodeSonsMap_[currentNodeId],
                   sizeof(NodeId) * nodeVarMap_[currentNodeId]->domainSize());
    nodeSonsMap_[currentNodeId] = sonsMap;
 
    chgNodeBoundVar_(currentNodeId, desiredVar);
  }

◆ updateFunctionGraph()

template<TESTNAME AttributeSelection, bool isScalar = false>

virtual void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::updateFunctionGraph ( )

pure virtual

Updates target to currently learned graph structure.

Implemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

◆ updateGraph()

template<TESTNAME AttributeSelection, bool isScalar = false>

virtual void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::updateGraph ( )

pure virtual

Updates the tree after a new observation has been added.

Implemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

◆ updateNode_()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::updateNode_	(	NodeId	updatedNode,
		gum::VariableSet &	varsOfInterest )

protected

From the given sets of node, selects randomly one and installs it on given node.

Chechks of course if node's current variable is not in that set first.

Parameters

nody	: the node we update
bestVars	: the set of interessting vars to be installed here

Chechks of course if node's current variable is not in that set first.

Parameters

nody	: the node we update
bestVar	: the set of interessting vars to be installed here

Definition at line 187 of file incrementalGraphLearner_tpl.h.

                                      {
    // If this node has no interesting variable, we turn it into a leaf
    if (varsOfInterest.empty()) {
      convertNode2Leaf_(updatedNode);
      return;
    }
 
    // If this node has already one of the best variable intalled as test, we
    // move on
    if (nodeVarMap_.exists(updatedNode) && varsOfInterest.exists(nodeVarMap_[updatedNode])) {
      return;
    }
 
    // In any other case we have to install variable as best test
    Idx randy = randomValue(varsOfInterest.size()), basc = 0;
    SetConstIteratorSafe< const DiscreteVariable* > varIter;
    for (varIter = varsOfInterest.cbeginSafe(), basc = 0;
         varIter != varsOfInterest.cendSafe() && basc < randy;
         ++varIter, basc++)
      ;
 
    transpose_(updatedNode, *varIter);
  }

References gum::Set< Key >::cbeginSafe(), gum::Set< Key >::cendSafe(), convertNode2Leaf_(), gum::Set< Key >::empty(), gum::Set< Key >::exists(), nodeVarMap_, gum::randomValue(), and gum::Set< Key >::size().

Here is the call graph for this function:

◆ updateNodeWithObservation_()

template<TESTNAME AttributeSelection, bool isScalar = false>

virtual void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::updateNodeWithObservation_	(	const Observation *	newObs,
		NodeId	currentNodeId )

inlineprotectedvirtual

Will update internal graph's NodeDatabase of given node with the new observation.

Parameters

newObs
currentNodeId

Reimplemented in gum::IMDDI< AttributeSelection, isScalar >, and gum::ITI< AttributeSelection, isScalar >.

Definition at line 189 of file incrementalGraphLearner.h.

                                                                                             {
      nodeId2Database_[currentNodeId]->addObservation(newObs);
    }

References nodeId2Database_.

Referenced by gum::IMDDI< AttributeSelection, isScalar >::updateNodeWithObservation_(), and gum::ITI< AttributeSelection, isScalar >::updateNodeWithObservation_().

Here is the caller graph for this function:

◆ updateVar()

template<TESTNAME AttributeSelection, bool isScalar>

void gum::IncrementalGraphLearner< AttributeSelection, isScalar >::updateVar ( const DiscreteVariable * var )

virtual

If a new modality appears to exists for given variable, call this method to turn every associated node to this variable into leaf. Graph has then indeed to be revised.

Definition at line 166 of file incrementalGraphLearner_tpl.h.

                                   {
    Link< NodeId >* nodIter = var2Node_[var]->list();
    Link< NodeId >* nni     = nullptr;
    while (nodIter) {
      nni = nodIter->nextLink();
      convertNode2Leaf_(nodIter->element());
      nodIter = nni;
    }
  }

References convertNode2Leaf_(), gum::Link< T >::element(), gum::Link< T >::nextLink(), and var2Node_.

Here is the call graph for this function:

Member Data Documentation

◆ leafDatabase_

template<TESTNAME AttributeSelection, bool isScalar = false>

HashTable< NodeId, Set< const Observation* >* > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::leafDatabase_

protected

This hashtable binds to every leaf an associated set of all hte observations compatible with it.

Definition at line 403 of file incrementalGraphLearner.h.

Referenced by chgNodeBoundVar_(), convertNode2Leaf_(), insertLeafNode_(), and removeNode_().

◆ model_

template<TESTNAME AttributeSelection, bool isScalar = false>

NodeGraphPart gum::IncrementalGraphLearner< AttributeSelection, isScalar >::model_

protected

The source of nodeId.

Definition at line 369 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner(), insertNode_(), and removeNode_().

◆ needUpdate_

template<TESTNAME AttributeSelection, bool isScalar = false>

bool gum::IncrementalGraphLearner< AttributeSelection, isScalar >::needUpdate_

protected

Definition at line 416 of file incrementalGraphLearner.h.

Referenced by chgNodeBoundVar_(), insertNode_(), and removeNode_().

◆ nodeId2Database_

template<TESTNAME AttributeSelection, bool isScalar = false>

HashTable< NodeId, NodeDatabase< AttributeSelection, isScalar >* > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeId2Database_

protected

This hashtable binds every node to an associated NodeDatabase which handles every observation that concerns that node.

Definition at line 397 of file incrementalGraphLearner.h.

Referenced by ~IncrementalGraphLearner(), insertNode_(), nodeNbObservation(), removeNode_(), and updateNodeWithObservation_().

◆ nodeSonsMap_

template<TESTNAME AttributeSelection, bool isScalar = false>

HashTable< NodeId, NodeId* > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeSonsMap_

protected

A table giving for any node a table mapping to its son idx is the modality of associated variable.

Definition at line 385 of file incrementalGraphLearner.h.

Referenced by addObservation(), convertNode2Leaf_(), insertInternalNode_(), isTerminal(), nodeSon(), and removeNode_().

◆ nodeVarMap_

template<TESTNAME AttributeSelection, bool isScalar = false>

HashTable< NodeId, const DiscreteVariable* > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::nodeVarMap_

protected

Gives for any node its associated variable.

Definition at line 379 of file incrementalGraphLearner.h.

Referenced by chgNodeBoundVar_(), convertNode2Leaf_(), insertNode_(), nodeVar(), removeNode_(), size(), and updateNode_().

◆ root_

template<TESTNAME AttributeSelection, bool isScalar = false>

NodeId gum::IncrementalGraphLearner< AttributeSelection, isScalar >::root_

protected

The root of the ordered tree.

Definition at line 374 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner(), addObservation(), and root().

◆ setOfVars_

template<TESTNAME AttributeSelection, bool isScalar = false>

gum::VariableSet gum::IncrementalGraphLearner< AttributeSelection, isScalar >::setOfVars_

protected

Definition at line 411 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner(), and insertSetOfVars().

◆ target_

template<TESTNAME AttributeSelection, bool isScalar = false>

MultiDimFunctionGraph< double >* gum::IncrementalGraphLearner< AttributeSelection, isScalar >::target_

protected

The final diagram we're building.

Definition at line 409 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner().

◆ value_

template<TESTNAME AttributeSelection, bool isScalar = false>

const DiscreteVariable* gum::IncrementalGraphLearner< AttributeSelection, isScalar >::value_

protected

Definition at line 413 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner(), _assumeValue_(), _clearValue_(), chgNodeBoundVar_(), and convertNode2Leaf_().

◆ valueAssumed_

template<TESTNAME AttributeSelection, bool isScalar = false>

Sequence< ValueType > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::valueAssumed_

protected

Definition at line 414 of file incrementalGraphLearner.h.

Referenced by _assumeValue_(), and _assumeValue_().

◆ var2Node_

template<TESTNAME AttributeSelection, bool isScalar = false>

HashTable< const DiscreteVariable*, LinkedList< NodeId >* > gum::IncrementalGraphLearner< AttributeSelection, isScalar >::var2Node_

protected

Associates to any variable the list of all nodes associated to this variable.

Definition at line 391 of file incrementalGraphLearner.h.

Referenced by IncrementalGraphLearner(), chgNodeBoundVar_(), insertNode_(), removeNode_(), and updateVar().

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

agrum/FMDP/learning/datastructure/incrementalGraphLearner.h
agrum/FMDP/learning/datastructure/incrementalGraphLearner_tpl.h

Public Member Functions

Protected Attributes

Private Types

Constructor & destructor.

New Observation insertion methods

Graph Structure update methods

Detailed Description

Member Typedef Documentation

◆ ValueType

Constructor & Destructor Documentation

◆ IncrementalGraphLearner()

◆ ~IncrementalGraphLearner()

Member Function Documentation

◆ _assumeValue_() [1/3]

◆ _assumeValue_() [2/3]

◆ _assumeValue_() [3/3]

◆ _branchObs_() [1/3]

◆ _branchObs_() [2/3]

◆ _branchObs_() [3/3]

◆ _clearValue_() [1/3]

◆ _clearValue_() [2/3]

◆ _clearValue_() [3/3]

◆ addObservation()

◆ chgNodeBoundVar_()

◆ convertNode2Leaf_()

◆ insertInternalNode_()

◆ insertLeafNode_()

◆ insertNode_()

◆ insertSetOfVars()

◆ isTerminal()

◆ nodeNbObservation()

◆ nodeSon()

◆ nodeVar()

◆ removeNode_()

◆ root()

◆ size()

◆ transpose_()

◆ updateFunctionGraph()

◆ updateGraph()

◆ updateNode_()

◆ updateNodeWithObservation_()

◆ updateVar()

Member Data Documentation

◆ leafDatabase_

◆ model_

◆ needUpdate_

◆ nodeId2Database_

◆ nodeSonsMap_

◆ nodeVarMap_

◆ root_

◆ setOfVars_

◆ target_

◆ value_

◆ valueAssumed_

◆ var2Node_