1 | //===----------------------------------------------------------------------===// |
2 | // |
3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
4 | // See https://llvm.org/LICENSE.txt for license information. |
5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
6 | // |
7 | //===----------------------------------------------------------------------===// |
8 | // |
9 | // REQUIRES: long_tests |
10 | |
11 | // <random> |
12 | |
13 | // template<class RealType = double> |
14 | // class gamma_distribution |
15 | |
16 | // template<class _URNG> result_type operator()(_URNG& g, const param_type& parm); |
17 | |
18 | #include <random> |
19 | #include <cassert> |
20 | #include <vector> |
21 | #include <numeric> |
22 | |
23 | #include "test_macros.h" |
24 | |
25 | template <class T> |
26 | inline |
27 | T |
28 | sqr(T x) |
29 | { |
30 | return x * x; |
31 | } |
32 | |
33 | int main(int, char**) |
34 | { |
35 | { |
36 | typedef std::gamma_distribution<> D; |
37 | typedef D::param_type P; |
38 | typedef std::mt19937 G; |
39 | G g; |
40 | D d(0.5, 2); |
41 | P p(1, .5); |
42 | const int N = 1000000; |
43 | std::vector<D::result_type> u; |
44 | for (int i = 0; i < N; ++i) |
45 | { |
46 | D::result_type v = d(g, p); |
47 | assert(d.min() < v); |
48 | u.push_back(v); |
49 | } |
50 | double mean = std::accumulate(u.begin(), u.end(), 0.0) / u.size(); |
51 | double var = 0; |
52 | double skew = 0; |
53 | double kurtosis = 0; |
54 | for (unsigned i = 0; i < u.size(); ++i) |
55 | { |
56 | double dbl = (u[i] - mean); |
57 | double d2 = sqr(dbl); |
58 | var += d2; |
59 | skew += dbl * d2; |
60 | kurtosis += d2 * d2; |
61 | } |
62 | var /= u.size(); |
63 | double dev = std::sqrt(x: var); |
64 | skew /= u.size() * dev * var; |
65 | kurtosis /= u.size() * var * var; |
66 | kurtosis -= 3; |
67 | double x_mean = p.alpha() * p.beta(); |
68 | double x_var = p.alpha() * sqr(p.beta()); |
69 | double x_skew = 2 / std::sqrt(x: p.alpha()); |
70 | double x_kurtosis = 6 / p.alpha(); |
71 | assert(std::abs((mean - x_mean) / x_mean) < 0.01); |
72 | assert(std::abs((var - x_var) / x_var) < 0.01); |
73 | assert(std::abs((skew - x_skew) / x_skew) < 0.01); |
74 | assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.01); |
75 | } |
76 | { |
77 | typedef std::gamma_distribution<> D; |
78 | typedef D::param_type P; |
79 | typedef std::mt19937 G; |
80 | G g; |
81 | D d(1, .5); |
82 | P p(2, 3); |
83 | const int N = 1000000; |
84 | std::vector<D::result_type> u; |
85 | for (int i = 0; i < N; ++i) |
86 | { |
87 | D::result_type v = d(g, p); |
88 | assert(d.min() < v); |
89 | u.push_back(v); |
90 | } |
91 | double mean = std::accumulate(u.begin(), u.end(), 0.0) / u.size(); |
92 | double var = 0; |
93 | double skew = 0; |
94 | double kurtosis = 0; |
95 | for (unsigned i = 0; i < u.size(); ++i) |
96 | { |
97 | double dbl = (u[i] - mean); |
98 | double d2 = sqr(dbl); |
99 | var += d2; |
100 | skew += dbl * d2; |
101 | kurtosis += d2 * d2; |
102 | } |
103 | var /= u.size(); |
104 | double dev = std::sqrt(x: var); |
105 | skew /= u.size() * dev * var; |
106 | kurtosis /= u.size() * var * var; |
107 | kurtosis -= 3; |
108 | double x_mean = p.alpha() * p.beta(); |
109 | double x_var = p.alpha() * sqr(p.beta()); |
110 | double x_skew = 2 / std::sqrt(p.alpha()); |
111 | double x_kurtosis = 6 / p.alpha(); |
112 | assert(std::abs((mean - x_mean) / x_mean) < 0.01); |
113 | assert(std::abs((var - x_var) / x_var) < 0.01); |
114 | assert(std::abs((skew - x_skew) / x_skew) < 0.01); |
115 | assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.01); |
116 | } |
117 | { |
118 | typedef std::gamma_distribution<> D; |
119 | typedef D::param_type P; |
120 | typedef std::mt19937 G; |
121 | G g; |
122 | D d(2, 3); |
123 | P p(.5, 2); |
124 | const int N = 1000000; |
125 | std::vector<D::result_type> u; |
126 | for (int i = 0; i < N; ++i) |
127 | { |
128 | D::result_type v = d(g, p); |
129 | assert(d.min() < v); |
130 | u.push_back(v); |
131 | } |
132 | double mean = std::accumulate(u.begin(), u.end(), 0.0) / u.size(); |
133 | double var = 0; |
134 | double skew = 0; |
135 | double kurtosis = 0; |
136 | for (unsigned i = 0; i < u.size(); ++i) |
137 | { |
138 | double dbl = (u[i] - mean); |
139 | double d2 = sqr(dbl); |
140 | var += d2; |
141 | skew += dbl * d2; |
142 | kurtosis += d2 * d2; |
143 | } |
144 | var /= u.size(); |
145 | double dev = std::sqrt(x: var); |
146 | skew /= u.size() * dev * var; |
147 | kurtosis /= u.size() * var * var; |
148 | kurtosis -= 3; |
149 | double x_mean = p.alpha() * p.beta(); |
150 | double x_var = p.alpha() * sqr(p.beta()); |
151 | double x_skew = 2 / std::sqrt(p.alpha()); |
152 | double x_kurtosis = 6 / p.alpha(); |
153 | assert(std::abs((mean - x_mean) / x_mean) < 0.01); |
154 | assert(std::abs((var - x_var) / x_var) < 0.01); |
155 | assert(std::abs((skew - x_skew) / x_skew) < 0.01); |
156 | assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.01); |
157 | } |
158 | |
159 | return 0; |
160 | } |
161 | |