1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 package org.apache.commons.math.stat.descriptive.moment;
18
19 import java.io.Serializable;
20 import java.util.Arrays;
21
22 import org.apache.commons.math.DimensionMismatchException;
23 import org.apache.commons.math.linear.RealMatrix;
24 import org.apache.commons.math.linear.RealMatrixImpl;
25
26
27
28
29
30
31 public class VectorialCovariance implements Serializable {
32
33
34 private static final long serialVersionUID = 4118372414238930270L;
35
36
37 private double[] sums;
38
39
40 private double[] productsSums;
41
42
43 private boolean isBiasCorrected;
44
45
46 private long n;
47
48
49
50
51
52
53 public VectorialCovariance(int dimension, boolean isBiasCorrected) {
54 sums = new double[dimension];
55 productsSums = new double[dimension * (dimension + 1) / 2];
56 n = 0;
57 this.isBiasCorrected = isBiasCorrected;
58 }
59
60
61
62
63
64
65 public void increment(double[] v) throws DimensionMismatchException {
66 if (v.length != sums.length) {
67 throw new DimensionMismatchException(v.length, sums.length);
68 }
69 int k = 0;
70 for (int i = 0; i < v.length; ++i) {
71 sums[i] += v[i];
72 for (int j = 0; j <= i; ++j) {
73 productsSums[k++] += v[i] * v[j];
74 }
75 }
76 n++;
77 }
78
79
80
81
82
83 public RealMatrix getResult() {
84
85 int dimension = sums.length;
86 RealMatrixImpl result = new RealMatrixImpl(dimension, dimension);
87
88 if (n > 1) {
89 double[][] resultData = result.getDataRef();
90 double c = 1.0 / (n * (isBiasCorrected ? (n - 1) : n));
91 int k = 0;
92 for (int i = 0; i < dimension; ++i) {
93 for (int j = 0; j <= i; ++j) {
94 double e = c * (n * productsSums[k++] - sums[i] * sums[j]);
95 resultData[i][j] = e;
96 resultData[j][i] = e;
97 }
98 }
99 }
100
101 return result;
102
103 }
104
105
106
107
108
109 public long getN() {
110 return n;
111 }
112
113
114
115
116 public void clear() {
117 n = 0;
118 Arrays.fill(sums, 0.0);
119 Arrays.fill(productsSums, 0.0);
120 }
121
122 }