1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 package org.apache.commons.math4.legacy.stat.inference;
18
19 import org.apache.commons.math4.legacy.exception.NoDataException;
20 import org.apache.commons.math4.legacy.exception.NullArgumentException;
21 import org.junit.Assert;
22 import org.junit.Test;
23
24
25
26
27
28
29
30 public class MannWhitneyUTestTest {
31
32 protected MannWhitneyUTest testStatistic = new MannWhitneyUTest();
33
34 @Test
35 public void testMannWhitneyUSimple() {
36
37
38
39
40
41
42 final double x[] = {19, 22, 16, 29, 24};
43 final double y[] = {20, 11, 17, 12};
44
45 Assert.assertEquals(3, testStatistic.mannWhitneyU(x, y), 1e-10);
46 Assert.assertEquals(0.08641, testStatistic.mannWhitneyUTest(x, y), 1e-5);
47 }
48
49
50 @Test
51 public void testMannWhitneyUInputValidation() {
52
53
54 try {
55 testStatistic.mannWhitneyUTest(new double[] { }, new double[] { 1.0 });
56 Assert.fail("x does not contain samples (exact), NoDataException expected");
57 } catch (NoDataException ex) {
58
59 }
60
61 try {
62 testStatistic.mannWhitneyUTest(new double[] { 1.0 }, new double[] { });
63 Assert.fail("y does not contain samples (exact), NoDataException expected");
64 } catch (NoDataException ex) {
65
66 }
67
68
69
70
71 try {
72 testStatistic.mannWhitneyUTest(null, null);
73 Assert.fail("x and y is null (exact), NullArgumentException expected");
74 } catch (NullArgumentException ex) {
75
76 }
77
78 try {
79 testStatistic.mannWhitneyUTest(null, null);
80 Assert.fail("x and y is null (asymptotic), NullArgumentException expected");
81 } catch (NullArgumentException ex) {
82
83 }
84
85
86
87
88 try {
89 testStatistic.mannWhitneyUTest(null, new double[] { 1.0 });
90 Assert.fail("x is null (exact), NullArgumentException expected");
91 } catch (NullArgumentException ex) {
92
93 }
94
95 try {
96 testStatistic.mannWhitneyUTest(new double[] { 1.0 }, null);
97 Assert.fail("y is null (exact), NullArgumentException expected");
98 } catch (NullArgumentException ex) {
99
100 }
101 }
102
103 @Test
104 public void testBigDataSet() {
105 double[] d1 = new double[1500];
106 double[] d2 = new double[1500];
107 for (int i = 0; i < 1500; i++) {
108 d1[i] = 2 * i;
109 d2[i] = 2 * i + 1;
110 }
111 double result = testStatistic.mannWhitneyUTest(d1, d2);
112 Assert.assertTrue(result > 0.1);
113 }
114
115 @Test
116 public void testBigDataSetOverflow() {
117
118 double[] d1 = new double[110000];
119 double[] d2 = new double[110000];
120 for (int i = 0; i < 110000; i++) {
121 d1[i] = i;
122 d2[i] = i;
123 }
124 double result = testStatistic.mannWhitneyUTest(d1, d2);
125 Assert.assertEquals(1.0, result, 0.0);
126 }
127 }