KNN.h File Reference

#include <dignea/core/Solution.h>
#include <dignea/distances/Distance.h>
#include <dignea/searches/NoveltySearch.h>
#include <functional>
#include <iostream>
#include <string>
#include <utility>
#include <vector>

Go to the source code of this file.

Classes
class	KNN
	K-Nearest Neighbours Algorithm. More...

Typedefs
using	neighborT = pair< float, int >
	Data type which represents a close neighbours. The first value is the distance to the individual The second value is the index of the neighbour in the array (Distance_to_other, other_index) More...
using	neighVector = vector< neighborT >
	Vector of neighbours Each element is of type neighborT. [(Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1)]. More...
using	neighMatrix = vector< neighVector >
	Data type which represents a matrix of close neighbours. The first value is the index of the neighbour in the array The second value is the distance to the individual [ [ (Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1) ], [ (Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1) ] ]. More...
using	Descriptor = vector< float >

Functions
vector< float >	sparseness (const vector< Descriptor > &population, Distance< float > *metric, const int k)
	Computes the sparseness of all individuals in the population vector Based on the Novelty Search from Exploiting Open-Endedness to Solve Problems Through the Search for Novelty. More...
float	sparseness (const Descriptor &solution, const vector< Descriptor > &population, Distance< float > *metric, const int k)
	Computes the sparseness score of the individual against the rest of individuals in the population list Based on the Novelty Search from Exploiting Open-Endedness to Solve Problems Through the Search for Novelty. More...

Detailed Description

Author

Alejandro Marrero (amarr.nosp@m.erd@.nosp@m.ull.e.nosp@m.du.e.nosp@m.s)

Version

0.1

Date

2022-10-11

Copyright

Copyright (c) 2022

Typedef Documentation

neighborT

using neighborT = pair<float, int>

Data type which represents a close neighbours. The first value is the distance to the individual The second value is the index of the neighbour in the array (Distance_to_other, other_index)

Template Parameters

T

neighMatrix

using neighMatrix = vector<neighVector>

Data type which represents a matrix of close neighbours. The first value is the index of the neighbour in the array The second value is the distance to the individual [ [ (Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1) ], [ (Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1) ] ].

Template Parameters

T

neighVector

using neighVector = vector<neighborT>

Vector of neighbours Each element is of type neighborT. [(Distance_to_other_0, other_index_0), (Distance_to_other_1, other_index_1), (...), (Distance_to_other_n-1, other_index_n-1)].

Template Parameters

T

Function Documentation

sparseness() [1/2]

float sparseness	(	const Descriptor &	solution,
		const vector< Descriptor > &	population,
		Distance< float > *	metric,
		const int	k
	)

Computes the sparseness score of the individual against the rest of individuals in the population list Based on the Novelty Search from Exploiting Open-Endedness to Solve Problems Through the Search for Novelty.

Template Parameters

T

Parameters

individual
neighbours
metric
k

Returns

T

sparseness() [2/2]

vector<float> sparseness	(	const vector< Descriptor > &	population,
		Distance< float > *	metric,
		const int	k
	)

Computes the sparseness of all individuals in the population vector Based on the Novelty Search from Exploiting Open-Endedness to Solve Problems Through the Search for Novelty.

Template Parameters

T

Parameters

individual
neighbours
metric
k

Returns

T