• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
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.math.linear;
19 
20 
21 
22 /**
23  * An interface to classes that implement an algorithm to calculate the
24  * Singular Value Decomposition of a real matrix.
25  * <p>
26  * The Singular Value Decomposition of matrix A is a set of three matrices: U,
27  * &Sigma; and V such that A = U &times; &Sigma; &times; V<sup>T</sup>. Let A be
28  * a m &times; n matrix, then U is a m &times; p orthogonal matrix, &Sigma; is a
29  * p &times; p diagonal matrix with positive or null elements, V is a p &times;
30  * n orthogonal matrix (hence V<sup>T</sup> is also orthogonal) where
31  * p=min(m,n).
32  * </p>
33  * <p>This interface is similar to the class with similar name from the
34  * <a href="http://math.nist.gov/javanumerics/jama/">JAMA</a> library, with the
35  * following changes:</p>
36  * <ul>
37  *   <li>the <code>norm2</code> method which has been renamed as {@link #getNorm()
38  *   getNorm},</li>
39  *   <li>the <code>cond</code> method which has been renamed as {@link
40  *   #getConditionNumber() getConditionNumber},</li>
41  *   <li>the <code>rank</code> method which has been renamed as {@link #getRank()
42  *   getRank},</li>
43  *   <li>a {@link #getUT() getUT} method has been added,</li>
44  *   <li>a {@link #getVT() getVT} method has been added,</li>
45  *   <li>a {@link #getSolver() getSolver} method has been added,</li>
46  *   <li>a {@link #getCovariance(double) getCovariance} method has been added.</li>
47  * </ul>
48  * @see <a href="http://mathworld.wolfram.com/SingularValueDecomposition.html">MathWorld</a>
49  * @see <a href="http://en.wikipedia.org/wiki/Singular_value_decomposition">Wikipedia</a>
50  * @version $Revision: 928081 $ $Date: 2010-03-26 23:36:38 +0100 (ven. 26 mars 2010) $
51  * @since 2.0
52  */
53 public interface SingularValueDecomposition {
54 
55     /**
56      * Returns the matrix U of the decomposition.
57      * <p>U is an orthogonal matrix, i.e. its transpose is also its inverse.</p>
58      * @return the U matrix
59      * @see #getUT()
60      */
getU()61     RealMatrix getU();
62 
63     /**
64      * Returns the transpose of the matrix U of the decomposition.
65      * <p>U is an orthogonal matrix, i.e. its transpose is also its inverse.</p>
66      * @return the U matrix (or null if decomposed matrix is singular)
67      * @see #getU()
68      */
getUT()69     RealMatrix getUT();
70 
71     /**
72      * Returns the diagonal matrix &Sigma; of the decomposition.
73      * <p>&Sigma; is a diagonal matrix. The singular values are provided in
74      * non-increasing order, for compatibility with Jama.</p>
75      * @return the &Sigma; matrix
76      */
getS()77     RealMatrix getS();
78 
79     /**
80      * Returns the diagonal elements of the matrix &Sigma; of the decomposition.
81      * <p>The singular values are provided in non-increasing order, for
82      * compatibility with Jama.</p>
83      * @return the diagonal elements of the &Sigma; matrix
84      */
getSingularValues()85     double[] getSingularValues();
86 
87     /**
88      * Returns the matrix V of the decomposition.
89      * <p>V is an orthogonal matrix, i.e. its transpose is also its inverse.</p>
90      * @return the V matrix (or null if decomposed matrix is singular)
91      * @see #getVT()
92      */
getV()93     RealMatrix getV();
94 
95     /**
96      * Returns the transpose of the matrix V of the decomposition.
97      * <p>V is an orthogonal matrix, i.e. its transpose is also its inverse.</p>
98      * @return the V matrix (or null if decomposed matrix is singular)
99      * @see #getV()
100      */
getVT()101     RealMatrix getVT();
102 
103     /**
104      * Returns the n &times; n covariance matrix.
105      * <p>The covariance matrix is V &times; J &times; V<sup>T</sup>
106      * where J is the diagonal matrix of the inverse of the squares of
107      * the singular values.</p>
108      * @param minSingularValue value below which singular values are ignored
109      * (a 0 or negative value implies all singular value will be used)
110      * @return covariance matrix
111      * @exception IllegalArgumentException if minSingularValue is larger than
112      * the largest singular value, meaning all singular values are ignored
113      */
getCovariance(double minSingularValue)114     RealMatrix getCovariance(double minSingularValue) throws IllegalArgumentException;
115 
116     /**
117      * Returns the L<sub>2</sub> norm of the matrix.
118      * <p>The L<sub>2</sub> norm is max(|A &times; u|<sub>2</sub> /
119      * |u|<sub>2</sub>), where |.|<sub>2</sub> denotes the vectorial 2-norm
120      * (i.e. the traditional euclidian norm).</p>
121      * @return norm
122      */
getNorm()123     double getNorm();
124 
125     /**
126      * Return the condition number of the matrix.
127      * @return condition number of the matrix
128      */
getConditionNumber()129     double getConditionNumber();
130 
131     /**
132      * Return the effective numerical matrix rank.
133      * <p>The effective numerical rank is the number of non-negligible
134      * singular values. The threshold used to identify non-negligible
135      * terms is max(m,n) &times; ulp(s<sub>1</sub>) where ulp(s<sub>1</sub>)
136      * is the least significant bit of the largest singular value.</p>
137      * @return effective numerical matrix rank
138      */
getRank()139     int getRank();
140 
141     /**
142      * Get a solver for finding the A &times; X = B solution in least square sense.
143      * @return a solver
144      */
getSolver()145     DecompositionSolver getSolver();
146 
147 }
148