CVT Map Elites Archive class. More...

#include <cvtMapElitesArchive.h>

Inheritance diagram for Selector::MapElites::CvtMapElitesArchive:

Public Member Functions
	CvtMapElitesArchive (Mutator::RNG &rng, size_t nbCentroids, size_t nbDescriptors, double minValue, double maxValue, size_t nbIterationInit, size_t nbDotsInit, double a1, double b1, double a2, double b2)
	Constructor of the CVT Map Elites Archive.

double	dist_squared (const std::vector< double > &a, const std::vector< double > &b)
	Calculate squared Euclidean distance between two points.

std::vector< double >	add (const std::vector< double > &a, const std::vector< double > &b)
	Vector addition.

std::vector< double >	scalar_mult (const std::vector< double > &a, double s)
	Scalar multiplication.

std::vector< double >	average (const std::vector< std::vector< double > > &points)
	Compute the average of a set of points.

std::vector< double >	random_point (Mutator::RNG &rng)
	Generate a random point within the descriptor bounds.

size_t	nearest (const std::vector< double > &point, const std::vector< std::vector< double > > &centroids)
	Find the nearest centroid to a given point.

void	initialize_cvt (Mutator::RNG &rng)
	Initialize the CVT centroids using Lloyd's algorithm.

size_t	getIndexForDescriptor (const std::vector< double > &descriptors) const
	Get the index for given descriptors.

const std::pair< std::shared_ptr< Learn::EvaluationResult >, const TPG::TPGVertex * > &	getArchiveFromDescriptors (const std::vector< double > &descriptors) const override
	Get the archive content at given descriptors.

void	setArchiveFromDescriptors (const TPG::TPGVertex *vertex, std::shared_ptr< Learn::EvaluationResult > eval, const std::vector< double > &descriptors) override
	Set the archive content at given descriptors.

virtual const std::vector< std::vector< double > > &	getCentroids () const
	Return the centroids used by the CVT.

Public Member Functions inherited from Selector::MapElites::MapElitesArchive
	MapElitesArchive (size_t nbBinPerDescriptor, size_t nbDescriptors, double minValue, double maxValue)
	Constructor of the MapElitesArchive.

uint64_t	size () const
	Get the size of the archive.

std::pair< uint64_t, uint64_t >	getDimensions () const
	Get the dimensions of the archive.

std::vector< double >	getArchiveLimits () const
	get the archive limits

virtual const std::vector< std::pair< std::shared_ptr< Learn::EvaluationResult >, const TPG::TPGVertex * > > &	getAllArchive () const
	Get all the archive content.

virtual const std::pair< std::shared_ptr< Learn::EvaluationResult >, const TPG::TPGVertex * > &	getArchiveAt (const std::vector< uint64_t > &indices) const
	Get the archive content at given indices.

virtual void	setArchiveAt (const TPG::TPGVertex *vertex, std::shared_ptr< Learn::EvaluationResult > eval, const std::vector< uint64_t > &indices)
	Set the archive content at given indices.

virtual uint64_t	getIndexArchive (double value) const
	Get the index in one dimension of the archive from a value.

virtual uint64_t	computeLinearIndex (const std::vector< uint64_t > &indices) const
	Compute the linear index from multi-dimensional indices.

virtual std::vector< uint64_t >	computeIndices (uint64_t index) const
	Compute the multi-dimensional indices from linear index.

virtual bool	containsRoot (const TPG::TPGVertex *root) const
	Check if the archive contains a root TPGVertex.

virtual void	removeRootFromArchiveIfNotComplete (const TPG::TPGVertex *root, size_t maxNbEvaluation)
	Remove a root TPGVertex from the archive if its number of evaluation is below maxNbEvaluation.

virtual void	removeRootFromArchive (const TPG::TPGVertex *root, size_t maxNbEvaluation)
	Remove a root TPGVertex from the archive.

virtual std::set< const TPG::TPGVertex * >	getVerticesInArchive () const
	Return a set with the current vectors in the archive.

Protected Attributes
size_t	nbCentroids
	Number of centroids.

size_t	nbIterationInit
	Number of iterations for initialization.

size_t	nbDotsInit
	Number of dots for initialization.

double	a1
	CVT parameters.

double	b1
	CVT parameters.

double	a2
	CVT parameters.

double	b2
	CVT parameters.

std::vector< std::vector< double > >	centroids
	Centroids of the CVT.

Protected Attributes inherited from Selector::MapElites::MapElitesArchive
uint64_t	nbBinPerDescriptor
	Number of bins per descriptor.

uint64_t	nbDescriptors
	Number of descriptors for this archive.

double	minValue
	Minimum value for the descriptors.

double	maxValue
	Maximum value for the descriptors.

std::vector< double >	archiveLimits
	Limits of the bins for the descriptors.

std::vector< std::pair< std::shared_ptr< Learn::EvaluationResult >, const TPG::TPGVertex * > >	archive

Detailed Description

CVT Map Elites Archive class.

This class implements a CVT (Centroidal Voronoi Tessellation) based Map Elites Archive It inherits from the MapElitesArchive class.

Constructor & Destructor Documentation

◆ CvtMapElitesArchive()

Selector::MapElites::CvtMapElitesArchive::CvtMapElitesArchive	(	Mutator::RNG &	rng,
		size_t	nbCentroids,
		size_t	nbDescriptors,
		double	minValue,
		double	maxValue,
		size_t	nbIterationInit,
		size_t	nbDotsInit,
		double	a1,
		double	b1,
		double	a2,
		double	b2 )

inline

Constructor of the CVT Map Elites Archive.

Parameters

[in]	rng	Random number generator
[in]	nbCentroids	Number of centroids
[in]	nbDescriptors	Number of descriptors
[in]	minValue	Minimum value for the descriptors
[in]	maxValue	Maximum value for the descriptors
[in]	nbIterationInit	Number of iterations for initialization
[in]	nbDotsInit	Number of dots for initialization
[in]	a1	CVT parameter
[in]	b1	CVT parameter
[in]	a2	CVT parameter
[in]	b2	CVT parameter

Member Function Documentation

◆ add()

std::vector< double > Selector::MapElites::CvtMapElitesArchive::add	(	const std::vector< double > &	a,
		const std::vector< double > &	b )

Vector addition.

Parameters

[in]	a	First vector
[in]	b	Second vector

◆ average()

std::vector< double > Selector::MapElites::CvtMapElitesArchive::average ( const std::vector< std::vector< double > > & points )

Compute the average of a set of points.

Parameters

[in] points Set of points

◆ dist_squared()

double Selector::MapElites::CvtMapElitesArchive::dist_squared	(	const std::vector< double > &	a,
		const std::vector< double > &	b )

Calculate squared Euclidean distance between two points.

Parameters

[in]	a	First point
[in]	b	Second point

Copyright or © or Copr. IETR/INSA - Rennes (2025) :

Quentin Vacher qvach.nosp@m.er@i.nosp@m.nsa-r.nosp@m.enne.nosp@m.s.fr (2025)

GEGELATI is an open-source reinforcement learning framework for training artificial intelligence based on Tangled Program Graphs (TPGs).

This software is governed by the CeCILL-C license under French law and abiding by the rules of distribution of free software. You can use, modify and/ or redistribute the software under the terms of the CeCILL-C license as circulated by CEA, CNRS and INRIA at the following URL "http://www.cecill.info".

As a counterpart to the access to the source code and rights to copy, modify and redistribute granted by the license, users are provided only with a limited warranty and the software's author, the holder of the economic rights, and the successive licensors have only limited liability.

In this respect, the user's attention is drawn to the risks associated with loading, using, modifying and/or developing or reproducing the software by the user in light of its specific status of free software, that may mean that it is complicated to manipulate, and that also therefore means that it is reserved for developers and experienced professionals having in-depth computer knowledge. Users are therefore encouraged to load and test the software's suitability as regards their requirements in conditions enabling the security of their systems and/or data to be ensured and, more generally, to use and operate it in the same conditions as regards security.

The fact that you are presently reading this means that you have had knowledge of the CeCILL-C license and that you accept its terms.

◆ getArchiveFromDescriptors()

const std::pair< std::shared_ptr< Learn::EvaluationResult >, const TPG::TPGVertex * > & Selector::MapElites::CvtMapElitesArchive::getArchiveFromDescriptors ( const std::vector< double > & descriptors ) const

overridevirtual

Get the archive content at given descriptors.

Reimplemented from Selector::MapElites::MapElitesArchive.

◆ getIndexForDescriptor()

size_t Selector::MapElites::CvtMapElitesArchive::getIndexForDescriptor ( const std::vector< double > & descriptors ) const

Get the index for given descriptors.

Parameters

[in] descriptors The descriptors to get the index for

◆ initialize_cvt()

void Selector::MapElites::CvtMapElitesArchive::initialize_cvt ( Mutator::RNG & rng )

Initialize the CVT centroids using Lloyd's algorithm.

Parameters

[in] rng Random number generator

◆ nearest()

size_t Selector::MapElites::CvtMapElitesArchive::nearest	(	const std::vector< double > &	point,
		const std::vector< std::vector< double > > &	centroids )

Find the nearest centroid to a given point.

Parameters

[in]	point	The point to find the nearest centroid for
[in]	centroids	The list of centroids

◆ random_point()

std::vector< double > Selector::MapElites::CvtMapElitesArchive::random_point ( Mutator::RNG & rng )

Generate a random point within the descriptor bounds.

Parameters

[in] rng Random number generator

◆ scalar_mult()

std::vector< double > Selector::MapElites::CvtMapElitesArchive::scalar_mult	(	const std::vector< double > &	a,
		double	s )

Scalar multiplication.

Parameters

[in]	a	Vector
[in]	s	Scalar

◆ setArchiveFromDescriptors()

void Selector::MapElites::CvtMapElitesArchive::setArchiveFromDescriptors	(	const TPG::TPGVertex *	vertex,
		std::shared_ptr< Learn::EvaluationResult >	eval,
		const std::vector< double > &	descriptors )

overridevirtual

Set the archive content at given descriptors.

Reimplemented from Selector::MapElites::MapElitesArchive.

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

D:/a/gegelati/gegelati/gegelatilib/include/selector/mapElites/cvtMapElitesArchive.h
D:/a/gegelati/gegelati/gegelatilib/src/selector/mapElites/cvtMapElitesArchive.cpp

Public Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ CvtMapElitesArchive()

Member Function Documentation

◆ add()

◆ average()

◆ dist_squared()

◆ getArchiveFromDescriptors()

◆ getIndexForDescriptor()

◆ initialize_cvt()

◆ nearest()

◆ random_point()

◆ scalar_mult()

◆ setArchiveFromDescriptors()