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
18 package org.apache.commons.math3.optimization.general;
19
20 /**
21 * This interface represents a preconditioner for differentiable scalar
22 * objective function optimizers.
23 * @version $Id: Preconditioner.java 1422230 2012-12-15 12:11:13Z erans $
24 * @deprecated As of 3.1 (to be removed in 4.0).
25 * @since 2.0
26 */
27 @Deprecated
28 public interface Preconditioner {
29 /**
30 * Precondition a search direction.
31 * <p>
32 * The returned preconditioned search direction must be computed fast or
33 * the algorithm performances will drop drastically. A classical approach
34 * is to compute only the diagonal elements of the hessian and to divide
35 * the raw search direction by these elements if they are all positive.
36 * If at least one of them is negative, it is safer to return a clone of
37 * the raw search direction as if the hessian was the identity matrix. The
38 * rationale for this simplified choice is that a negative diagonal element
39 * means the current point is far from the optimum and preconditioning will
40 * not be efficient anyway in this case.
41 * </p>
42 * @param point current point at which the search direction was computed
43 * @param r raw search direction (i.e. opposite of the gradient)
44 * @return approximation of H<sup>-1</sup>r where H is the objective function hessian
45 */
46 double[] precondition(double[] point, double[] r);
47 }