001/*
002 * Java Genetic Algorithm Library (jenetics-7.1.0).
003 * Copyright (c) 2007-2022 Franz Wilhelmstötter
004 *
005 * Licensed under the Apache License, Version 2.0 (the "License");
006 * you may not use this file except in compliance with the License.
007 * You may obtain a copy of the License at
008 *
009 *      http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 *
017 * Author:
018 *    Franz Wilhelmstötter (franz.wilhelmstoetter@gmail.com)
019 */
020package io.jenetics;
021
022import java.util.Arrays;
023
024import io.jenetics.internal.math.DoubleAdder;
025import io.jenetics.stat.DoubleSummary;
026import io.jenetics.util.BaseSeq;
027import io.jenetics.util.Seq;
028
029/**
030 * The roulette-wheel selector is also known as fitness proportional selector,
031 * but in the <em>Jenetics</em> library it is implemented as probability selector.
032 * The fitness value <i>f<sub>i</sub></i>  is used to calculate the selection
033 * probability of individual <i>i</i>.
034 *
035 * @see <a href="http://en.wikipedia.org/wiki/Roulette_wheel_selection">
036 *          Wikipedia: Roulette wheel selection
037 *      </a>
038 *
039 * @author <a href="mailto:franz.wilhelmstoetter@gmail.com">Franz Wilhelmstötter</a>
040 * @since 1.0
041 * @version 5.0
042 */
043public class RouletteWheelSelector<
044        G extends Gene<?, G>,
045        N extends Number & Comparable<? super N>
046>
047        extends ProbabilitySelector<G, N>
048{
049
050        public RouletteWheelSelector() {
051                this(false);
052        }
053
054        protected RouletteWheelSelector(final boolean sorted) {
055                super(sorted);
056        }
057
058        @Override
059        protected double[] probabilities(
060                final Seq<Phenotype<G, N>> population,
061                final int count
062        ) {
063                assert population != null : "Population must not be null. ";
064                assert population.nonEmpty() : "Population is empty.";
065                assert count > 0 : "Population to select must be greater than zero. ";
066
067                final double[] fitness = fitnessOf(population);
068                sub(fitness, Math.min(DoubleSummary.min(fitness), 0.0));
069                final double sum = DoubleAdder.sum(fitness);
070
071                if (eq(sum, 0.0)) {
072                        Arrays.fill(fitness, 1.0/population.size());
073                } else {
074                        for (int i = fitness.length; --i >= 0;) {
075                                fitness[i] = fitness[i]/sum;
076                        }
077                }
078
079                return fitness;
080        }
081
082        private double[] fitnessOf(final BaseSeq<Phenotype<G, N>> population) {
083                final double[] fitness = new double[population.length()];
084                for (int i = fitness.length; --i >= 0;) {
085                        final double fit = population.get(i).fitness().doubleValue();
086                        fitness[i] = Double.isFinite(fit) ? fit : 0.0;
087                }
088                return fitness;
089        }
090
091        private static void sub(final double[] values, final double subtrahend) {
092                if (Double.compare(subtrahend, 0.0) != 0) {
093                        for (int i = values.length; --i >= 0;) {
094                                values[i] -= subtrahend;
095                        }
096                }
097        }
098
099        @Override
100        public String toString() {
101                return getClass().getSimpleName();
102        }
103
104}