eval_param.pass.cpp (5910B)
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 // REQUIRES: long_tests 11 12 // <random> 13 14 // template<class RealType = double> 15 // class weibull_distribution 16 17 // template<class _URNG> result_type operator()(_URNG& g, const param_type& parm); 18 19 #include <random> 20 #include <cassert> 21 #include <vector> 22 #include <numeric> 23 #include <cstddef> 24 25 template <class T> 26 inline 27 T 28 sqr(T x) 29 { 30 return x * x; 31 } 32 33 int main() 34 { 35 { 36 typedef std::weibull_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 (std::size_t 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(var); 64 skew /= u.size() * dev * var; 65 kurtosis /= u.size() * var * var; 66 kurtosis -= 3; 67 double x_mean = p.b() * std::tgamma(1 + 1/p.a()); 68 double x_var = sqr(p.b()) * std::tgamma(1 + 2/p.a()) - sqr(x_mean); 69 double x_skew = (sqr(p.b())*p.b() * std::tgamma(1 + 3/p.a()) - 70 3*x_mean*x_var - sqr(x_mean)*x_mean) / 71 (std::sqrt(x_var)*x_var); 72 double x_kurtosis = (sqr(sqr(p.b())) * std::tgamma(1 + 4/p.a()) - 73 4*x_skew*x_var*sqrt(x_var)*x_mean - 74 6*sqr(x_mean)*x_var - sqr(sqr(x_mean))) / sqr(x_var) - 3; 75 assert(std::abs((mean - x_mean) / x_mean) < 0.01); 76 assert(std::abs((var - x_var) / x_var) < 0.01); 77 assert(std::abs((skew - x_skew) / x_skew) < 0.01); 78 assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.01); 79 } 80 { 81 typedef std::weibull_distribution<> D; 82 typedef D::param_type P; 83 typedef std::mt19937 G; 84 G g; 85 D d(1, .5); 86 P p(2, 3); 87 const int N = 1000000; 88 std::vector<D::result_type> u; 89 for (int i = 0; i < N; ++i) 90 { 91 D::result_type v = d(g, p); 92 assert(d.min() <= v); 93 u.push_back(v); 94 } 95 double mean = std::accumulate(u.begin(), u.end(), 0.0) / u.size(); 96 double var = 0; 97 double skew = 0; 98 double kurtosis = 0; 99 for (std::size_t i = 0; i < u.size(); ++i) 100 { 101 double dbl = (u[i] - mean); 102 double d2 = sqr(dbl); 103 var += d2; 104 skew += dbl * d2; 105 kurtosis += d2 * d2; 106 } 107 var /= u.size(); 108 double dev = std::sqrt(var); 109 skew /= u.size() * dev * var; 110 kurtosis /= u.size() * var * var; 111 kurtosis -= 3; 112 double x_mean = p.b() * std::tgamma(1 + 1/p.a()); 113 double x_var = sqr(p.b()) * std::tgamma(1 + 2/p.a()) - sqr(x_mean); 114 double x_skew = (sqr(p.b())*p.b() * std::tgamma(1 + 3/p.a()) - 115 3*x_mean*x_var - sqr(x_mean)*x_mean) / 116 (std::sqrt(x_var)*x_var); 117 double x_kurtosis = (sqr(sqr(p.b())) * std::tgamma(1 + 4/p.a()) - 118 4*x_skew*x_var*sqrt(x_var)*x_mean - 119 6*sqr(x_mean)*x_var - sqr(sqr(x_mean))) / sqr(x_var) - 3; 120 assert(std::abs((mean - x_mean) / x_mean) < 0.01); 121 assert(std::abs((var - x_var) / x_var) < 0.01); 122 assert(std::abs((skew - x_skew) / x_skew) < 0.01); 123 assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.03); 124 } 125 { 126 typedef std::weibull_distribution<> D; 127 typedef D::param_type P; 128 typedef std::mt19937 G; 129 G g; 130 D d(2, 3); 131 P p(.5, 2); 132 const int N = 1000000; 133 std::vector<D::result_type> u; 134 for (int i = 0; i < N; ++i) 135 { 136 D::result_type v = d(g, p); 137 assert(d.min() <= v); 138 u.push_back(v); 139 } 140 double mean = std::accumulate(u.begin(), u.end(), 0.0) / u.size(); 141 double var = 0; 142 double skew = 0; 143 double kurtosis = 0; 144 for (std::size_t i = 0; i < u.size(); ++i) 145 { 146 double dbl = (u[i] - mean); 147 double d2 = sqr(dbl); 148 var += d2; 149 skew += dbl * d2; 150 kurtosis += d2 * d2; 151 } 152 var /= u.size(); 153 double dev = std::sqrt(var); 154 skew /= u.size() * dev * var; 155 kurtosis /= u.size() * var * var; 156 kurtosis -= 3; 157 double x_mean = p.b() * std::tgamma(1 + 1/p.a()); 158 double x_var = sqr(p.b()) * std::tgamma(1 + 2/p.a()) - sqr(x_mean); 159 double x_skew = (sqr(p.b())*p.b() * std::tgamma(1 + 3/p.a()) - 160 3*x_mean*x_var - sqr(x_mean)*x_mean) / 161 (std::sqrt(x_var)*x_var); 162 double x_kurtosis = (sqr(sqr(p.b())) * std::tgamma(1 + 4/p.a()) - 163 4*x_skew*x_var*sqrt(x_var)*x_mean - 164 6*sqr(x_mean)*x_var - sqr(sqr(x_mean))) / sqr(x_var) - 3; 165 assert(std::abs((mean - x_mean) / x_mean) < 0.01); 166 assert(std::abs((var - x_var) / x_var) < 0.01); 167 assert(std::abs((skew - x_skew) / x_skew) < 0.01); 168 assert(std::abs((kurtosis - x_kurtosis) / x_kurtosis) < 0.03); 169 } 170 }