/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.commons.math3.optim;
import org.apache.commons.math3.exception.MathIllegalStateException;
import org.apache.commons.math3.exception.NotStrictlyPositiveException;
import org.apache.commons.math3.exception.TooManyEvaluationsException;
import org.apache.commons.math3.random.RandomVectorGenerator;
/**
* Base class multi-start optimizer for a multivariate function.
* <br/>
* This class wraps an optimizer in order to use it several times in
* turn with different starting points (trying to avoid being trapped
* in a local extremum when looking for a global one).
* <em>It is not a "user" class.</em>
*
* @param <PAIR> Type of the point/value pair returned by the optimization
* algorithm.
*
* @since 3.0
*/
public abstract class BaseMultiStartMultivariateOptimizer<PAIR>
extends BaseMultivariateOptimizer<PAIR> {
/** Underlying classical optimizer. */
private final BaseMultivariateOptimizer<PAIR> optimizer;
/** Number of evaluations already performed for all starts. */
private int totalEvaluations;
/** Number of starts to go. */
private int starts;
/** Random generator for multi-start. */
private RandomVectorGenerator generator;
/** Optimization data. */
private OptimizationData[] optimData;
/**
* Location in {@link #optimData} where the updated maximum
* number of evaluations will be stored.
*/
private int maxEvalIndex = -1;
/**
* Location in {@link #optimData} where the updated start value
* will be stored.
*/
private int initialGuessIndex = -1;
/**
* Create a multi-start optimizer from a single-start optimizer.
* <p>
* Note that if there are bounds constraints (see {@link #getLowerBound()}
* and {@link #getUpperBound()}), then a simple rejection algorithm is used
* at each restart. This implies that the random vector generator should have
* a good probability to generate vectors in the bounded domain, otherwise the
* rejection algorithm will hit the {@link #getMaxEvaluations()} count without
* generating a proper restart point. Users must be take great care of the <a
* href="http://en.wikipedia.org/wiki/Curse_of_dimensionality">curse of dimensionality</a>.
* </p>
* @param optimizer Single-start optimizer to wrap.
* @param starts Number of starts to perform. If {@code starts == 1},
* the {@link #optimize(OptimizationData[]) optimize} will return the
* same solution as the given {@code optimizer} would return.
* @param generator Random vector generator to use for restarts.
* @throws NotStrictlyPositiveException if {@code starts < 1}.
*/
public BaseMultiStartMultivariateOptimizer(final BaseMultivariateOptimizer<PAIR> optimizer,
final int starts,
final RandomVectorGenerator generator) {
super(optimizer.getConvergenceChecker());
if (starts < 1) {
throw new NotStrictlyPositiveException(starts);
}
this.optimizer = optimizer;
this.starts = starts;
this.generator = generator;
}
/** {@inheritDoc} */
@Override
public int getEvaluations() {
return totalEvaluations;
}
/**
* Gets all the optima found during the last call to {@code optimize}.
* The optimizer stores all the optima found during a set of
* restarts. The {@code optimize} method returns the best point only.
* This method returns all the points found at the end of each starts,
* including the best one already returned by the {@code optimize} method.
* <br/>
* The returned array as one element for each start as specified
* in the constructor. It is ordered with the results from the
* runs that did converge first, sorted from best to worst
* objective value (i.e in ascending order if minimizing and in
* descending order if maximizing), followed by {@code null} elements
* corresponding to the runs that did not converge. This means all
* elements will be {@code null} if the {@code optimize} method did throw
* an exception.
* This also means that if the first element is not {@code null}, it is
* the best point found across all starts.
* <br/>
* The behaviour is undefined if this method is called before
* {@code optimize}; it will likely throw {@code NullPointerException}.
*
* @return an array containing the optima sorted from best to worst.
*/
public abstract PAIR[] getOptima();
/**
* {@inheritDoc}
*
* @throws MathIllegalStateException if {@code optData} does not contain an
* instance of {@link MaxEval} or {@link InitialGuess}.
*/
@Override
public PAIR optimize(OptimizationData... optData) {
// Store arguments in order to pass them to the internal optimizer.
optimData = optData;
// Set up base class and perform computations.
return super.optimize(optData);
}
/** {@inheritDoc} */
@Override
protected PAIR doOptimize() {
// Remove all instances of "MaxEval" and "InitialGuess" from the
// array that will be passed to the internal optimizer.
// The former is to enforce smaller numbers of allowed evaluations
// (according to how many have been used up already), and the latter
// to impose a different start value for each start.
for (int i = 0; i < optimData.length; i++) {
if (optimData[i] instanceof MaxEval) {
optimData[i] = null;
maxEvalIndex = i;
}
if (optimData[i] instanceof InitialGuess) {
optimData[i] = null;
initialGuessIndex = i;
continue;
}
}
if (maxEvalIndex == -1) {
throw new MathIllegalStateException();
}
if (initialGuessIndex == -1) {
throw new MathIllegalStateException();
}
RuntimeException lastException = null;
totalEvaluations = 0;
clear();
final int maxEval = getMaxEvaluations();
final double[] min = getLowerBound();
final double[] max = getUpperBound();
final double[] startPoint = getStartPoint();
// Multi-start loop.
for (int i = 0; i < starts; i++) {
// CHECKSTYLE: stop IllegalCatch
try {
// Decrease number of allowed evaluations.
optimData[maxEvalIndex] = new MaxEval(maxEval - totalEvaluations);
// New start value.
double[] s = null;
if (i == 0) {
s = startPoint;
} else {
int attempts = 0;
while (s == null) {
if (attempts++ >= getMaxEvaluations()) {
throw new TooManyEvaluationsException(getMaxEvaluations());
}
s = generator.nextVector();
for (int k = 0; s != null && k < s.length; ++k) {
if ((min != null && s[k] < min[k]) || (max != null && s[k] > max[k])) {
// reject the vector
s = null;
}
}
}
}
optimData[initialGuessIndex] = new InitialGuess(s);
// Optimize.
final PAIR result = optimizer.optimize(optimData);
store(result);
} catch (RuntimeException mue) {
lastException = mue;
}
// CHECKSTYLE: resume IllegalCatch
totalEvaluations += optimizer.getEvaluations();
}
final PAIR[] optima = getOptima();
if (optima.length == 0) {
// All runs failed.
throw lastException; // Cannot be null if starts >= 1.
}
// Return the best optimum.
return optima[0];
}
/**
* Method that will be called in order to store each found optimum.
*
* @param optimum Result of an optimization run.
*/
protected abstract void store(PAIR optimum);
/**
* Method that will called in order to clear all stored optima.
*/
protected abstract void clear();
}