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.general;
019
020 import org.apache.commons.math.FunctionEvaluationException;
021
022 /**
023 * This interface represents a preconditioner for differentiable scalar
024 * objective function optimizers.
025 * @version $Revision: 1073158 $ $Date: 2011-02-21 22:46:52 +0100 (lun. 21 f??vr. 2011) $
026 * @since 2.0
027 */
028 public interface Preconditioner {
029
030 /**
031 * Precondition a search direction.
032 * <p>
033 * The returned preconditioned search direction must be computed fast or
034 * the algorithm performances will drop drastically. A classical approach
035 * is to compute only the diagonal elements of the hessian and to divide
036 * the raw search direction by these elements if they are all positive.
037 * If at least one of them is negative, it is safer to return a clone of
038 * the raw search direction as if the hessian was the identity matrix. The
039 * rationale for this simplified choice is that a negative diagonal element
040 * means the current point is far from the optimum and preconditioning will
041 * not be efficient anyway in this case.
042 * </p>
043 * @param point current point at which the search direction was computed
044 * @param r raw search direction (i.e. opposite of the gradient)
045 * @return approximation of H<sup>-1</sup>r where H is the objective function hessian
046 * @exception FunctionEvaluationException if no cost can be computed for the parameters
047 * @exception IllegalArgumentException if point dimension is wrong
048 */
049 double[] precondition(double[] point, double[] r)
050 throws FunctionEvaluationException, IllegalArgumentException;
051
052 }