• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
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 #include <cstddef>
21 
22 template <class T>
23 inline
24 T
sqr(T x)25 sqr(T x)
26 {
27     return x * x;
28 }
29 
main()30 int main()
31 {
32     {
33         typedef std::bernoulli_distribution D;
34         typedef D::param_type P;
35         typedef std::minstd_rand G;
36         G g;
37         D d(.75);
38         P p(.25);
39         const int N = 100000;
40         std::vector<D::result_type> u;
41         for (int i = 0; i < N; ++i)
42             u.push_back(d(g, p));
43         double mean = std::accumulate(u.begin(), u.end(),
44                                               double(0)) / u.size();
45         double var = 0;
46         double skew = 0;
47         double kurtosis = 0;
48         for (std::size_t i = 0; i < u.size(); ++i)
49         {
50             double dbl = (u[i] - mean);
51             double d2 = sqr(dbl);
52             var += d2;
53             skew += dbl * d2;
54             kurtosis += d2 * d2;
55         }
56         var /= u.size();
57         double dev = std::sqrt(var);
58         skew /= u.size() * dev * var;
59         kurtosis /= u.size() * var * var;
60         kurtosis -= 3;
61         double x_mean = p.p();
62         double x_var = p.p()*(1-p.p());
63         double x_skew = (1 - 2 * p.p())/std::sqrt(x_var);
64         double x_kurtosis = (6 * sqr(p.p()) - 6 * p.p() + 1)/x_var;
65         assert(std::abs((mean - x_mean) / x_mean) < 0.01);
66         assert(std::abs((var - x_var) / x_var) < 0.01);
67         assert(std::abs((skew - x_skew) / x_skew) < 0.01);
68         assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.02);
69     }
70     {
71         typedef std::bernoulli_distribution D;
72         typedef D::param_type P;
73         typedef std::minstd_rand G;
74         G g;
75         D d(.25);
76         P p(.75);
77         const int N = 100000;
78         std::vector<D::result_type> u;
79         for (int i = 0; i < N; ++i)
80             u.push_back(d(g, p));
81         double mean = std::accumulate(u.begin(), u.end(),
82                                               double(0)) / u.size();
83         double var = 0;
84         double skew = 0;
85         double kurtosis = 0;
86         for (std::size_t i = 0; i < u.size(); ++i)
87         {
88             double dbl = (u[i] - mean);
89             double d2 = sqr(dbl);
90             var += d2;
91             skew += dbl * d2;
92             kurtosis += d2 * d2;
93         }
94         var /= u.size();
95         double dev = std::sqrt(var);
96         skew /= u.size() * dev * var;
97         kurtosis /= u.size() * var * var;
98         kurtosis -= 3;
99         double x_mean = p.p();
100         double x_var = p.p()*(1-p.p());
101         double x_skew = (1 - 2 * p.p())/std::sqrt(x_var);
102         double x_kurtosis = (6 * sqr(p.p()) - 6 * p.p() + 1)/x_var;
103         assert(std::abs((mean - x_mean) / x_mean) < 0.01);
104         assert(std::abs((var - x_var) / x_var) < 0.01);
105         assert(std::abs((skew - x_skew) / x_skew) < 0.01);
106         assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.02);
107     }
108 }
109