001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  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 */
017package org.apache.commons.math3.analysis.integration;
018
019import org.apache.commons.math3.exception.MaxCountExceededException;
020import org.apache.commons.math3.exception.NotStrictlyPositiveException;
021import org.apache.commons.math3.exception.NumberIsTooLargeException;
022import org.apache.commons.math3.exception.NumberIsTooSmallException;
023import org.apache.commons.math3.exception.TooManyEvaluationsException;
024import org.apache.commons.math3.util.FastMath;
025
026/**
027 * Implements the <a href="http://mathworld.wolfram.com/RombergIntegration.html">
028 * Romberg Algorithm</a> for integration of real univariate functions. For
029 * reference, see <b>Introduction to Numerical Analysis</b>, ISBN 038795452X,
030 * chapter 3.
031 * <p>
032 * Romberg integration employs k successive refinements of the trapezoid
033 * rule to remove error terms less than order O(N^(-2k)). Simpson's rule
034 * is a special case of k = 2.</p>
035 *
036 * @since 1.2
037 */
038public class RombergIntegrator extends BaseAbstractUnivariateIntegrator {
039
040    /** Maximal number of iterations for Romberg. */
041    public static final int ROMBERG_MAX_ITERATIONS_COUNT = 32;
042
043    /**
044     * Build a Romberg integrator with given accuracies and iterations counts.
045     * @param relativeAccuracy relative accuracy of the result
046     * @param absoluteAccuracy absolute accuracy of the result
047     * @param minimalIterationCount minimum number of iterations
048     * @param maximalIterationCount maximum number of iterations
049     * (must be less than or equal to {@link #ROMBERG_MAX_ITERATIONS_COUNT})
050     * @exception NotStrictlyPositiveException if minimal number of iterations
051     * is not strictly positive
052     * @exception NumberIsTooSmallException if maximal number of iterations
053     * is lesser than or equal to the minimal number of iterations
054     * @exception NumberIsTooLargeException if maximal number of iterations
055     * is greater than {@link #ROMBERG_MAX_ITERATIONS_COUNT}
056     */
057    public RombergIntegrator(final double relativeAccuracy,
058                             final double absoluteAccuracy,
059                             final int minimalIterationCount,
060                             final int maximalIterationCount)
061        throws NotStrictlyPositiveException, NumberIsTooSmallException, NumberIsTooLargeException {
062        super(relativeAccuracy, absoluteAccuracy, minimalIterationCount, maximalIterationCount);
063        if (maximalIterationCount > ROMBERG_MAX_ITERATIONS_COUNT) {
064            throw new NumberIsTooLargeException(maximalIterationCount,
065                                                ROMBERG_MAX_ITERATIONS_COUNT, false);
066        }
067    }
068
069    /**
070     * Build a Romberg integrator with given iteration counts.
071     * @param minimalIterationCount minimum number of iterations
072     * @param maximalIterationCount maximum number of iterations
073     * (must be less than or equal to {@link #ROMBERG_MAX_ITERATIONS_COUNT})
074     * @exception NotStrictlyPositiveException if minimal number of iterations
075     * is not strictly positive
076     * @exception NumberIsTooSmallException if maximal number of iterations
077     * is lesser than or equal to the minimal number of iterations
078     * @exception NumberIsTooLargeException if maximal number of iterations
079     * is greater than {@link #ROMBERG_MAX_ITERATIONS_COUNT}
080     */
081    public RombergIntegrator(final int minimalIterationCount,
082                             final int maximalIterationCount)
083        throws NotStrictlyPositiveException, NumberIsTooSmallException, NumberIsTooLargeException {
084        super(minimalIterationCount, maximalIterationCount);
085        if (maximalIterationCount > ROMBERG_MAX_ITERATIONS_COUNT) {
086            throw new NumberIsTooLargeException(maximalIterationCount,
087                                                ROMBERG_MAX_ITERATIONS_COUNT, false);
088        }
089    }
090
091    /**
092     * Construct a Romberg integrator with default settings
093     * (max iteration count set to {@link #ROMBERG_MAX_ITERATIONS_COUNT})
094     */
095    public RombergIntegrator() {
096        super(DEFAULT_MIN_ITERATIONS_COUNT, ROMBERG_MAX_ITERATIONS_COUNT);
097    }
098
099    /** {@inheritDoc} */
100    @Override
101    protected double doIntegrate()
102        throws TooManyEvaluationsException, MaxCountExceededException {
103
104        final int m = getMaximalIterationCount() + 1;
105        double previousRow[] = new double[m];
106        double currentRow[]  = new double[m];
107
108        TrapezoidIntegrator qtrap = new TrapezoidIntegrator();
109        currentRow[0] = qtrap.stage(this, 0);
110        incrementCount();
111        double olds = currentRow[0];
112        while (true) {
113
114            final int i = getIterations();
115
116            // switch rows
117            final double[] tmpRow = previousRow;
118            previousRow = currentRow;
119            currentRow = tmpRow;
120
121            currentRow[0] = qtrap.stage(this, i);
122            incrementCount();
123            for (int j = 1; j <= i; j++) {
124                // Richardson extrapolation coefficient
125                final double r = (1L << (2 * j)) - 1;
126                final double tIJm1 = currentRow[j - 1];
127                currentRow[j] = tIJm1 + (tIJm1 - previousRow[j - 1]) / r;
128            }
129            final double s = currentRow[i];
130            if (i >= getMinimalIterationCount()) {
131                final double delta  = FastMath.abs(s - olds);
132                final double rLimit = getRelativeAccuracy() * (FastMath.abs(olds) + FastMath.abs(s)) * 0.5;
133                if ((delta <= rLimit) || (delta <= getAbsoluteAccuracy())) {
134                    return s;
135                }
136            }
137            olds = s;
138        }
139
140    }
141
142}