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 */
017
018 package org.apache.commons.math.optimization.fitting;
019
020 import org.apache.commons.math.FunctionEvaluationException;
021 import org.apache.commons.math.MathRuntimeException;
022 import org.apache.commons.math.analysis.polynomials.PolynomialFunction;
023 import org.apache.commons.math.optimization.DifferentiableMultivariateVectorialOptimizer;
024 import org.apache.commons.math.optimization.OptimizationException;
025
026 /** This class implements a curve fitting specialized for polynomials.
027 * <p>Polynomial fitting is a very simple case of curve fitting. The
028 * estimated coefficients are the polynomial coefficients. They are
029 * searched by a least square estimator.</p>
030 * @version $Revision: 811685 $ $Date: 2009-09-05 13:36:48 -0400 (Sat, 05 Sep 2009) $
031 * @since 2.0
032 */
033
034 public class PolynomialFitter {
035
036 /** Fitter for the coefficients. */
037 private final CurveFitter fitter;
038
039 /** Polynomial degree. */
040 private final int degree;
041
042 /** Simple constructor.
043 * <p>The polynomial fitter built this way are complete polynomials,
044 * ie. a n-degree polynomial has n+1 coefficients.</p>
045 * @param degree maximal degree of the polynomial
046 * @param optimizer optimizer to use for the fitting
047 */
048 public PolynomialFitter(int degree, final DifferentiableMultivariateVectorialOptimizer optimizer) {
049 this.fitter = new CurveFitter(optimizer);
050 this.degree = degree;
051 }
052
053 /** Add an observed weighted (x,y) point to the sample.
054 * @param weight weight of the observed point in the fit
055 * @param x abscissa of the point
056 * @param y observed value of the point at x, after fitting we should
057 * have P(x) as close as possible to this value
058 */
059 public void addObservedPoint(double weight, double x, double y) {
060 fitter.addObservedPoint(weight, x, y);
061 }
062
063 /** Get the polynomial fitting the weighted (x, y) points.
064 * @return polynomial function best fitting the observed points
065 * @exception OptimizationException if the algorithm failed to converge
066 */
067 public PolynomialFunction fit()
068 throws OptimizationException {
069 try {
070 return new PolynomialFunction(fitter.fit(new ParametricPolynomial(), new double[degree + 1]));
071 } catch (FunctionEvaluationException fee) {
072 // this should never happen
073 throw MathRuntimeException.createInternalError(fee);
074 }
075 }
076
077 /** Dedicated parametric polynomial class. */
078 private static class ParametricPolynomial implements ParametricRealFunction {
079
080 /** {@inheritDoc} */
081 public double[] gradient(double x, double[] parameters)
082 throws FunctionEvaluationException {
083 final double[] gradient = new double[parameters.length];
084 double xn = 1.0;
085 for (int i = 0; i < parameters.length; ++i) {
086 gradient[i] = xn;
087 xn *= x;
088 }
089 return gradient;
090 }
091
092 /** {@inheritDoc} */
093 public double value(final double x, final double[] parameters) {
094 double y = 0;
095 for (int i = parameters.length - 1; i >= 0; --i) {
096 y = y * x + parameters[i];
097 }
098 return y;
099 }
100
101 }
102
103 }