[1593] | 1 | // $Id: kolmogorov_smirnov_test.cc 1688 2008-12-30 23:14:12Z peter $ |
---|
| 2 | |
---|
| 3 | /* |
---|
| 4 | Copyright (C) 2008 Peter Johansson |
---|
| 5 | |
---|
| 6 | This file is part of the yat library, http://dev.thep.lu.se/yat |
---|
| 7 | |
---|
| 8 | The yat library is free software; you can redistribute it and/or |
---|
| 9 | modify it under the terms of the GNU General Public License as |
---|
| 10 | published by the Free Software Foundation; either version 3 of the |
---|
| 11 | License, or (at your option) any later version. |
---|
| 12 | |
---|
| 13 | The yat library is distributed in the hope that it will be useful, |
---|
| 14 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
| 15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
---|
| 16 | General Public License for more details. |
---|
| 17 | |
---|
| 18 | You should have received a copy of the GNU General Public License |
---|
| 19 | along with yat. If not, see <http://www.gnu.org/licenses/>. |
---|
| 20 | */ |
---|
| 21 | |
---|
| 22 | #include "Suite.h" |
---|
| 23 | |
---|
[1600] | 24 | #include "yat/statistics/Averager.h" |
---|
[1593] | 25 | #include "yat/statistics/KolmogorovSmirnov.h" |
---|
[1617] | 26 | #include "yat/random/random.h" |
---|
[1593] | 27 | |
---|
[1600] | 28 | #include <cmath> |
---|
| 29 | #include <iostream> |
---|
[1593] | 30 | |
---|
[1600] | 31 | using namespace theplu::yat; |
---|
| 32 | |
---|
| 33 | void test_one_sample(test::Suite&); |
---|
[1617] | 34 | void test_two_sample(test::Suite&); |
---|
[1600] | 35 | void test_p_value(test::Suite&); |
---|
[1612] | 36 | void test_reset(test::Suite&); |
---|
[1687] | 37 | void test_ties(test::Suite&); |
---|
[1600] | 38 | |
---|
[1593] | 39 | int main(int argc, char* argv[]) |
---|
| 40 | { |
---|
| 41 | test::Suite suite(argc, argv); |
---|
| 42 | |
---|
[1600] | 43 | test_one_sample(suite); |
---|
[1617] | 44 | test_two_sample(suite); |
---|
[1626] | 45 | test_p_value(suite); |
---|
[1612] | 46 | test_reset(suite); |
---|
[1687] | 47 | test_ties(suite); |
---|
[1593] | 48 | |
---|
[1600] | 49 | return suite.return_value(); |
---|
| 50 | } |
---|
[1593] | 51 | |
---|
[1600] | 52 | void test_one_sample(test::Suite& suite) |
---|
| 53 | { |
---|
| 54 | std::vector<double> correct(11); |
---|
| 55 | for (size_t i=0; i<correct.size(); ++i) { |
---|
| 56 | double s1 = 1.0 - i/10.0; |
---|
| 57 | double s2 = 0.0-i/10.0; |
---|
| 58 | if (std::abs(s1)>std::abs(s2)) |
---|
| 59 | correct[i] = s1; |
---|
| 60 | else |
---|
| 61 | correct[i] = s2; |
---|
| 62 | } |
---|
[1593] | 63 | |
---|
[1600] | 64 | for (size_t i=0; i<11; ++i) { |
---|
| 65 | statistics::KolmogorovSmirnov ks; |
---|
| 66 | for (size_t j=0; j<11; ++j) { |
---|
| 67 | ks.add(j, i==j); |
---|
| 68 | } |
---|
| 69 | double score = ks.signed_score(); |
---|
| 70 | if (!suite.add(suite.equal(score, correct[i]))) { |
---|
| 71 | suite.err() << "signed_score(void) failed\n"; |
---|
| 72 | } |
---|
| 73 | } |
---|
| 74 | |
---|
| 75 | statistics::KolmogorovSmirnov ks; |
---|
| 76 | for (size_t i=0; i<11; ++i) { |
---|
| 77 | ks.add(i, i==0); |
---|
| 78 | } |
---|
| 79 | size_t n=110000; |
---|
| 80 | double p = ks.p_value(n); |
---|
| 81 | double p_correct = 2.0/11.0; |
---|
[1618] | 82 | double margin = 10*std::sqrt(p_correct*(1-p_correct)/n); |
---|
[1600] | 83 | if (p>p_correct+margin || p<p_correct-margin) { |
---|
| 84 | suite.err() << "Error: p-value: " << p << "\n" |
---|
| 85 | << "expected approximately: " << p_correct << "\n" |
---|
| 86 | << "and at most " << margin << "deviation\n"; |
---|
| 87 | suite.add(false); |
---|
| 88 | } |
---|
[1593] | 89 | } |
---|
| 90 | |
---|
[1617] | 91 | |
---|
| 92 | void test_two_sample(test::Suite& suite) |
---|
| 93 | { |
---|
| 94 | suite.err() << "testing two sample\n"; |
---|
| 95 | statistics::KolmogorovSmirnov ks; |
---|
| 96 | for (size_t i=0; i<5; ++i) |
---|
| 97 | ks.add(i+0.5, i<2); |
---|
| 98 | suite.add(suite.equal(ks.score(), 1.0)); |
---|
[1618] | 99 | size_t n=100000; |
---|
[1617] | 100 | double p = ks.p_value(n); |
---|
| 101 | double p_correct = 0.2; |
---|
[1618] | 102 | double margin=10*std::sqrt(p_correct*(1-p_correct)/n); |
---|
[1617] | 103 | if (std::abs(p-p_correct)>margin) { |
---|
| 104 | suite.add(false); |
---|
| 105 | suite.err() << "Error: p = " << p << "\n" |
---|
| 106 | << "correct p would be: " << p_correct << "\n" |
---|
| 107 | << "expected a difference less than " << margin << "margin\n"; |
---|
| 108 | suite.err() << p << std::endl; |
---|
| 109 | } |
---|
| 110 | } |
---|
| 111 | |
---|
| 112 | |
---|
[1600] | 113 | void test_p_value(test::Suite& suite) |
---|
| 114 | { |
---|
| 115 | statistics::KolmogorovSmirnov ks; |
---|
| 116 | for (size_t i=0; i<100; ++i) { |
---|
| 117 | ks.add(i, true); |
---|
| 118 | ks.add(i+14.5, false); |
---|
| 119 | } |
---|
[1626] | 120 | suite.add(suite.equal(ks.score(), 0.15, 10)); |
---|
| 121 | |
---|
[1600] | 122 | statistics::Averager a; |
---|
| 123 | for (size_t n=0; n<100; ++n) { |
---|
| 124 | a.add(ks.p_value(100)); |
---|
| 125 | } |
---|
| 126 | double margin = 5 * a.standard_error(); |
---|
| 127 | double p_approx = ks.p_value(); |
---|
| 128 | if (std::abs(a.mean()-p_approx)>margin) { |
---|
| 129 | suite.add(false); |
---|
| 130 | suite.err() << "Error: unexpected large deviation between p_values\n" |
---|
| 131 | << "permutation p-value: " << a.mean() << "\n" |
---|
| 132 | << "analytical approximation: " << p_approx << "\n" |
---|
| 133 | << "expected deviation to be smaller than " << margin << "\n"; |
---|
| 134 | } |
---|
| 135 | |
---|
| 136 | } |
---|
[1612] | 137 | |
---|
| 138 | void test_reset(test::Suite& suite) |
---|
| 139 | { |
---|
| 140 | suite.err() << "testing reset\n"; |
---|
| 141 | statistics::KolmogorovSmirnov ks; |
---|
| 142 | ks.add(1.0, true); |
---|
| 143 | ks.add(2.0, false); |
---|
| 144 | ks.add(3.0, true); |
---|
| 145 | double score = ks.score(); |
---|
| 146 | double p = ks.p_value(); |
---|
| 147 | ks.reset(); |
---|
| 148 | ks.add(1.0, true); |
---|
| 149 | ks.add(2.0, false); |
---|
| 150 | ks.add(3.0, true); |
---|
| 151 | suite.add(suite.equal(ks.score(), score)); |
---|
| 152 | suite.add(suite.equal(ks.p_value(), p)); |
---|
| 153 | } |
---|
[1687] | 154 | |
---|
| 155 | |
---|
| 156 | void test_ties(test::Suite& suite) |
---|
| 157 | { |
---|
| 158 | suite.err() << "test ties" << std::endl; |
---|
| 159 | statistics::KolmogorovSmirnov ks; |
---|
| 160 | for (size_t i=0; i<5; ++i) |
---|
| 161 | ks.add(i, true); |
---|
| 162 | ks.add(0, false); |
---|
[1688] | 163 | suite.add(suite.equal(ks.score(), 1.0-0.2)); |
---|
[1687] | 164 | } |
---|