1//===----------------------------------------------------------------------===//
2//
3//                     The LLVM Compiler Infrastructure
4//
5// This file is dual licensed under the MIT and the University of Illinois Open
6// Source Licenses. See LICENSE.TXT for details.
7//
8//===----------------------------------------------------------------------===//
9
10// <random>
11
12// class bernoulli_distribution
13
14// template<class _URNG> result_type operator()(_URNG& g, const param_type& parm);
15
16#include <random>
17#include <numeric>
18#include <vector>
19#include <cassert>
20
21template <class T>
22inline
23T
24sqr(T x)
25{
26    return x * x;
27}
28
29int main()
30{
31    {
32        typedef std::bernoulli_distribution D;
33        typedef D::param_type P;
34        typedef std::minstd_rand G;
35        G g;
36        D d(.75);
37        P p(.25);
38        const int N = 100000;
39        std::vector<D::result_type> u;
40        for (int i = 0; i < N; ++i)
41            u.push_back(d(g, p));
42        double mean = std::accumulate(u.begin(), u.end(),
43                                              double(0)) / u.size();
44        double var = 0;
45        double skew = 0;
46        double kurtosis = 0;
47        for (int i = 0; i < u.size(); ++i)
48        {
49            double d = (u[i] - mean);
50            double d2 = sqr(d);
51            var += d2;
52            skew += d * d2;
53            kurtosis += d2 * d2;
54        }
55        var /= u.size();
56        double dev = std::sqrt(var);
57        skew /= u.size() * dev * var;
58        kurtosis /= u.size() * var * var;
59        kurtosis -= 3;
60        double x_mean = p.p();
61        double x_var = p.p()*(1-p.p());
62        double x_skew = (1 - 2 * p.p())/std::sqrt(x_var);
63        double x_kurtosis = (6 * sqr(p.p()) - 6 * p.p() + 1)/x_var;
64        assert(std::abs((mean - x_mean) / x_mean) < 0.01);
65        assert(std::abs((var - x_var) / x_var) < 0.01);
66        assert(std::abs((skew - x_skew) / x_skew) < 0.01);
67        assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.02);
68    }
69    {
70        typedef std::bernoulli_distribution D;
71        typedef D::param_type P;
72        typedef std::minstd_rand G;
73        G g;
74        D d(.25);
75        P p(.75);
76        const int N = 100000;
77        std::vector<D::result_type> u;
78        for (int i = 0; i < N; ++i)
79            u.push_back(d(g, p));
80        double mean = std::accumulate(u.begin(), u.end(),
81                                              double(0)) / u.size();
82        double var = 0;
83        double skew = 0;
84        double kurtosis = 0;
85        for (int i = 0; i < u.size(); ++i)
86        {
87            double d = (u[i] - mean);
88            double d2 = sqr(d);
89            var += d2;
90            skew += d * d2;
91            kurtosis += d2 * d2;
92        }
93        var /= u.size();
94        double dev = std::sqrt(var);
95        skew /= u.size() * dev * var;
96        kurtosis /= u.size() * var * var;
97        kurtosis -= 3;
98        double x_mean = p.p();
99        double x_var = p.p()*(1-p.p());
100        double x_skew = (1 - 2 * p.p())/std::sqrt(x_var);
101        double x_kurtosis = (6 * sqr(p.p()) - 6 * p.p() + 1)/x_var;
102        assert(std::abs((mean - x_mean) / x_mean) < 0.01);
103        assert(std::abs((var - x_var) / x_var) < 0.01);
104        assert(std::abs((skew - x_skew) / x_skew) < 0.01);
105        assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.02);
106    }
107}
108