1 /* 2 * Licensed to the Apache Software Foundation (ASF) under one or more 3 * contributor license agreements. See the NOTICE file distributed with 4 * this work for additional information regarding copyright ownership. 5 * The ASF licenses this file to You under the Apache License, Version 2.0 6 * (the "License"); you may not use this file except in compliance with 7 * the License. You may obtain a copy of the License at 8 * 9 * http://www.apache.org/licenses/LICENSE-2.0 10 * 11 * Unless required by applicable law or agreed to in writing, software 12 * distributed under the License is distributed on an "AS IS" BASIS, 13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 14 * See the License for the specific language governing permissions and 15 * limitations under the License. 16 */ 17 package org.apache.commons.math.analysis.integration; 18 19 import org.apache.commons.math.ConvergenceException; 20 import org.apache.commons.math.FunctionEvaluationException; 21 import org.apache.commons.math.MathRuntimeException; 22 import org.apache.commons.math.MaxIterationsExceededException; 23 import org.apache.commons.math.analysis.UnivariateRealFunction; 24 import org.apache.commons.math.exception.util.LocalizedFormats; 25 import org.apache.commons.math.util.FastMath; 26 27 /** 28 * Implements the <a href="http://mathworld.wolfram.com/Legendre-GaussQuadrature.html"> 29 * Legendre-Gauss</a> quadrature formula. 30 * <p> 31 * Legendre-Gauss integrators are efficient integrators that can 32 * accurately integrate functions with few functions evaluations. A 33 * Legendre-Gauss integrator using an n-points quadrature formula can 34 * integrate exactly 2n-1 degree polynomials. 35 * </p> 36 * <p> 37 * These integrators evaluate the function on n carefully chosen 38 * abscissas in each step interval (mapped to the canonical [-1 1] interval). 39 * The evaluation abscissas are not evenly spaced and none of them are 40 * at the interval endpoints. This implies the function integrated can be 41 * undefined at integration interval endpoints. 42 * </p> 43 * <p> 44 * The evaluation abscissas x<sub>i</sub> are the roots of the degree n 45 * Legendre polynomial. The weights a<sub>i</sub> of the quadrature formula 46 * integrals from -1 to +1 ∫ Li<sup>2</sup> where Li (x) = 47 * ∏ (x-x<sub>k</sub>)/(x<sub>i</sub>-x<sub>k</sub>) for k != i. 48 * </p> 49 * <p> 50 * @version $Revision: 1070725 $ $Date: 2011-02-15 02:31:12 +0100 (mar. 15 fvr. 2011) $ 51 * @since 1.2 52 */ 53 54 public class LegendreGaussIntegrator extends UnivariateRealIntegratorImpl { 55 56 /** Abscissas for the 2 points method. */ 57 private static final double[] ABSCISSAS_2 = { 58 -1.0 / FastMath.sqrt(3.0), 59 1.0 / FastMath.sqrt(3.0) 60 }; 61 62 /** Weights for the 2 points method. */ 63 private static final double[] WEIGHTS_2 = { 64 1.0, 65 1.0 66 }; 67 68 /** Abscissas for the 3 points method. */ 69 private static final double[] ABSCISSAS_3 = { 70 -FastMath.sqrt(0.6), 71 0.0, 72 FastMath.sqrt(0.6) 73 }; 74 75 /** Weights for the 3 points method. */ 76 private static final double[] WEIGHTS_3 = { 77 5.0 / 9.0, 78 8.0 / 9.0, 79 5.0 / 9.0 80 }; 81 82 /** Abscissas for the 4 points method. */ 83 private static final double[] ABSCISSAS_4 = { 84 -FastMath.sqrt((15.0 + 2.0 * FastMath.sqrt(30.0)) / 35.0), 85 -FastMath.sqrt((15.0 - 2.0 * FastMath.sqrt(30.0)) / 35.0), 86 FastMath.sqrt((15.0 - 2.0 * FastMath.sqrt(30.0)) / 35.0), 87 FastMath.sqrt((15.0 + 2.0 * FastMath.sqrt(30.0)) / 35.0) 88 }; 89 90 /** Weights for the 4 points method. */ 91 private static final double[] WEIGHTS_4 = { 92 (90.0 - 5.0 * FastMath.sqrt(30.0)) / 180.0, 93 (90.0 + 5.0 * FastMath.sqrt(30.0)) / 180.0, 94 (90.0 + 5.0 * FastMath.sqrt(30.0)) / 180.0, 95 (90.0 - 5.0 * FastMath.sqrt(30.0)) / 180.0 96 }; 97 98 /** Abscissas for the 5 points method. */ 99 private static final double[] ABSCISSAS_5 = { 100 -FastMath.sqrt((35.0 + 2.0 * FastMath.sqrt(70.0)) / 63.0), 101 -FastMath.sqrt((35.0 - 2.0 * FastMath.sqrt(70.0)) / 63.0), 102 0.0, 103 FastMath.sqrt((35.0 - 2.0 * FastMath.sqrt(70.0)) / 63.0), 104 FastMath.sqrt((35.0 + 2.0 * FastMath.sqrt(70.0)) / 63.0) 105 }; 106 107 /** Weights for the 5 points method. */ 108 private static final double[] WEIGHTS_5 = { 109 (322.0 - 13.0 * FastMath.sqrt(70.0)) / 900.0, 110 (322.0 + 13.0 * FastMath.sqrt(70.0)) / 900.0, 111 128.0 / 225.0, 112 (322.0 + 13.0 * FastMath.sqrt(70.0)) / 900.0, 113 (322.0 - 13.0 * FastMath.sqrt(70.0)) / 900.0 114 }; 115 116 /** Abscissas for the current method. */ 117 private final double[] abscissas; 118 119 /** Weights for the current method. */ 120 private final double[] weights; 121 122 /** 123 * Build a Legendre-Gauss integrator. 124 * @param n number of points desired (must be between 2 and 5 inclusive) 125 * @param defaultMaximalIterationCount maximum number of iterations 126 * @exception IllegalArgumentException if the number of points is not 127 * in the supported range 128 */ 129 public LegendreGaussIntegrator(final int n, final int defaultMaximalIterationCount) 130 throws IllegalArgumentException { 131 super(defaultMaximalIterationCount); 132 switch(n) { 133 case 2 : 134 abscissas = ABSCISSAS_2; 135 weights = WEIGHTS_2; 136 break; 137 case 3 : 138 abscissas = ABSCISSAS_3; 139 weights = WEIGHTS_3; 140 break; 141 case 4 : 142 abscissas = ABSCISSAS_4; 143 weights = WEIGHTS_4; 144 break; 145 case 5 : 146 abscissas = ABSCISSAS_5; 147 weights = WEIGHTS_5; 148 break; 149 default : 150 throw MathRuntimeException.createIllegalArgumentException( 151 LocalizedFormats.N_POINTS_GAUSS_LEGENDRE_INTEGRATOR_NOT_SUPPORTED, 152 n, 2, 5); 153 } 154 155 } 156 157 /** {@inheritDoc} */ 158 @Deprecated 159 public double integrate(final double min, final double max) 160 throws ConvergenceException, FunctionEvaluationException, IllegalArgumentException { 161 return integrate(f, min, max); 162 } 163 164 /** {@inheritDoc} */ 165 public double integrate(final UnivariateRealFunction f, final double min, final double max) 166 throws ConvergenceException, FunctionEvaluationException, IllegalArgumentException { 167 168 clearResult(); 169 verifyInterval(min, max); 170 verifyIterationCount(); 171 172 // compute first estimate with a single step 173 double oldt = stage(f, min, max, 1); 174 175 int n = 2; 176 for (int i = 0; i < maximalIterationCount; ++i) { 177 178 // improve integral with a larger number of steps 179 final double t = stage(f, min, max, n); 180 181 // estimate error 182 final double delta = FastMath.abs(t - oldt); 183 final double limit = 184 FastMath.max(absoluteAccuracy, 185 relativeAccuracy * (FastMath.abs(oldt) + FastMath.abs(t)) * 0.5); 186 187 // check convergence 188 if ((i + 1 >= minimalIterationCount) && (delta <= limit)) { 189 setResult(t, i); 190 return result; 191 } 192 193 // prepare next iteration 194 double ratio = FastMath.min(4, FastMath.pow(delta / limit, 0.5 / abscissas.length)); 195 n = FastMath.max((int) (ratio * n), n + 1); 196 oldt = t; 197 198 } 199 200 throw new MaxIterationsExceededException(maximalIterationCount); 201 202 } 203 204 /** 205 * Compute the n-th stage integral. 206 * @param f the integrand function 207 * @param min the lower bound for the interval 208 * @param max the upper bound for the interval 209 * @param n number of steps 210 * @return the value of n-th stage integral 211 * @throws FunctionEvaluationException if an error occurs evaluating the 212 * function 213 */ 214 private double stage(final UnivariateRealFunction f, 215 final double min, final double max, final int n) 216 throws FunctionEvaluationException { 217 218 // set up the step for the current stage 219 final double step = (max - min) / n; 220 final double halfStep = step / 2.0; 221 222 // integrate over all elementary steps 223 double midPoint = min + halfStep; 224 double sum = 0.0; 225 for (int i = 0; i < n; ++i) { 226 for (int j = 0; j < abscissas.length; ++j) { 227 sum += weights[j] * f.value(midPoint + halfStep * abscissas[j]); 228 } 229 midPoint += step; 230 } 231 232 return halfStep * sum; 233 234 } 235 236 } 237