1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
|
/*
File : NSLStatisticalTestTest.cpp
Project : LabPlot
Description : NSL Tests for statistical functions
--------------------------------------------------------------------
SPDX-FileCopyrightText: 2025 Kuntal Bar <barkuntal6@gmail.com>
SPDX-License-Identifier: GPL-2.0-or-later
*/
#include "NSLStatisticalTestTest.h"
extern "C" {
#include "backend/nsl/nsl_statistical_test.h"
}
void NSLStatisticalTestTest::testMannWhitney() {
const double sample1[] = {1.0, 2.0, 3.0};
const double sample2[] = {4.0, 5.0, 6.0};
const size_t n1 = sizeof(sample1) / sizeof(sample1[0]);
const size_t n2 = sizeof(sample2) / sizeof(sample2[0]);
double U = nsl_stats_mannwhitney_u(sample1, n1, sample2, n2);
printf("Mann-Whitney U: %f\n", U);
double p = nsl_stats_mannwhitney_p(U, n1, n2);
printf("Mann-Whitney p-value: %f\n", p);
double expected_U = 0.0;
double expected_p = 0.0808556;
const double epsilon = 1e-4;
QVERIFY(std::abs(U - expected_U) < epsilon);
QVERIFY(std::abs(p - expected_p) < epsilon);
}
void NSLStatisticalTestTest::testAnovaOneWay() {
const double group1[] = {1.0, 2.0, 3.0};
const double group2[] = {2.0, 3.0, 4.0};
const double group3[] = {5.0, 6.0, 7.0};
size_t sizes[] = {3, 3, 3};
const size_t n_groups = 3;
double* groups[] = {const_cast<double*>(group1), const_cast<double*>(group2), const_cast<double*>(group3)};
double F = nsl_stats_anova_oneway_f(groups, sizes, n_groups);
printf("Computed F-statistic: %f", F);
double p = nsl_stats_anova_oneway_p(groups, sizes, n_groups);
printf("Computed p-value: %f\n", p);
double expected_F = 13.0;
double expected_p = 0.0065;
const double epsilon = 1e-4;
QVERIFY(std::abs(F - expected_F) < epsilon);
QVERIFY(std::abs(p - expected_p) < epsilon);
}
void NSLStatisticalTestTest::testKruskalWallis() {
const double group1[] = {1.0, 2.0, 3.0};
const double group2[] = {2.0, 3.0, 4.0};
const double group3[] = {5.0, 6.0, 7.0};
size_t sizes[] = {3, 3, 3};
const size_t n_groups = 3;
double* groups[] = {const_cast<double*>(group1), const_cast<double*>(group2), const_cast<double*>(group3)};
double H = nsl_stats_kruskal_wallis_h(groups, sizes, n_groups);
double p = nsl_stats_kruskal_wallis_p(groups, sizes, n_groups);
printf("Computed h-statistic: %f\n", H);
printf("Computed p-value: %f\n", p);
double expected_H = 6.0564;
double expected_p = 0.0484;
const double epsilon = 1e-4;
QVERIFY(std::abs(H - expected_H) < epsilon);
QVERIFY(std::abs(p - expected_p) < epsilon);
}
void NSLStatisticalTestTest::testLogRankTest() {
// ----------------------------------------------------------------------------------
// Test Case 1: No Censored Data
// ----------------------------------------------------------------------------------
const double time1_group1[] = {5, 6, 6, 2, 4, 4};
const int status1_group1[] = {1, 1, 1, 1, 1, 1}; // 1 = event occurred
const double time1_group2[] = {1, 3, 3, 8, 9, 9};
const int status1_group2[] = {1, 1, 1, 1, 1, 1}; // 1 = event occurred
size_t group1_indices1[] = {0, 1, 2, 3, 4, 5};
size_t size1_1 = sizeof(group1_indices1) / sizeof(group1_indices1[0]);
size_t group2_indices1[] = {6, 7, 8, 9, 10, 11};
size_t size2_1 = sizeof(group2_indices1) / sizeof(group2_indices1[0]);
double time1_combined[12];
int status1_combined[12];
for (size_t i = 0; i < size1_1; i++) {
time1_combined[i] = time1_group1[i];
status1_combined[i] = status1_group1[i];
}
for (size_t i = 0; i < size2_1; i++) {
time1_combined[size1_1 + i] = time1_group2[i];
status1_combined[size1_1 + i] = status1_group2[i];
}
double H1 = nsl_stats_log_rank_test_statistic(time1_combined, status1_combined, group1_indices1, size1_1, group2_indices1, size2_1);
double p1 = nsl_stats_log_rank_test_p(time1_combined, status1_combined, group1_indices1, size1_1, group2_indices1, size2_1);
printf("Computed H-value: %f\n", H1);
printf("Computed p-value: %f\n", p1);
double expected_H1 = 1.009510;
double expected_p1 = 0.31502028;
double epsilon = 1e-4;
QVERIFY(std::abs(H1 - expected_H1) < epsilon);
QVERIFY(std::abs(p1 - expected_p1) < epsilon);
// ----------------------------------------------------------------------------------
// Test Case 2: Censored Data
// ----------------------------------------------------------------------------------
const double time2_group1[] = {5, 6, 6, 2, 4, 4};
const int status2_group1[] = {1, 1, 0, 1, 1, 0}; // 1 = event occurred, 0 = censored
const double time2_group2[] = {1, 3, 3, 8, 9, 9};
const int status2_group2[] = {1, 0, 1, 1, 0, 1}; // 1 = event occurred, 0 = censored
size_t group1_indices2[] = {0, 1, 2, 3, 4, 5};
size_t size1_2 = sizeof(group1_indices2) / sizeof(group1_indices2[0]);
size_t group2_indices2[] = {6, 7, 8, 9, 10, 11};
size_t size2_2 = sizeof(group2_indices2) / sizeof(group2_indices2[0]);
double time2_combined[12];
int status2_combined[12];
for (size_t i = 0; i < size1_2; i++) {
time2_combined[i] = time2_group1[i];
status2_combined[i] = status2_group1[i];
}
for (size_t i = 0; i < size2_2; i++) {
time2_combined[size1_2 + i] = time2_group2[i];
status2_combined[size1_2 + i] = status2_group2[i];
}
double H2 = nsl_stats_log_rank_test_statistic(time2_combined, status2_combined, group1_indices2, size1_2, group2_indices2, size2_2);
double p2 = nsl_stats_log_rank_test_p(time2_combined, status2_combined, group1_indices2, size1_2, group2_indices2, size2_2);
double expected_H2 = 0.586871;
double expected_p2 = 0.44362;
printf("Computed H-value: %f\n", H2);
printf("Computed p-value: %f\n", p2);
QVERIFY(std::abs(H2 - expected_H2) < epsilon);
QVERIFY(std::abs(p2 - expected_p2) < epsilon);
}
void NSLStatisticalTestTest::testIndependentT() {
const double sample1[] = {12.5, 13.3, 14.2, 12.7, 13.9};
const double sample2[] = {15.1, 14.8, 15.5, 15.2, 14.9};
size_t n1 = 5;
size_t n2 = 5;
double t_stat = nsl_stats_independent_t(sample1, n1, sample2, n2);
double p_value = nsl_stats_independent_t_p(sample1, n1, sample2, n2);
double expected_t = -5.0671;
double expected_p = 0.0010;
const double epsilon = 1e-4;
printf("Computed t-statistic: %f\n", t_stat);
printf("Computed p-value: %f\n", p_value);
QVERIFY(std::abs(t_stat - expected_t) < epsilon);
QVERIFY(std::abs(p_value - expected_p) < epsilon);
}
void NSLStatisticalTestTest::testOneSampleT() {
const double sample[] = {15.2, 14.8, 15.5, 14.9, 15.1};
size_t n = 5;
double hypothesized_mean = 15.0;
double t_stat = nsl_stats_one_sample_t(sample, n, hypothesized_mean);
double p_value = nsl_stats_one_sample_t_p(sample, n, hypothesized_mean, 0);
double expected_t = 0.8165;
double expected_p = 0.4601;
const double epsilon = 1e-4;
printf("Computed t-statistic: %f\n", t_stat);
printf("Computed p-value: %f\n", p_value);
QVERIFY(std::abs(t_stat - expected_t) < epsilon);
QVERIFY(std::abs(p_value - expected_p) < epsilon);
}
QTEST_MAIN(NSLStatisticalTestTest)
|
