ExRandom  2.1
Simple examples
Forward to Getting started. Up to Contents.

Here is how to sample exactly from the normal distribution. The code is simple_normal.cpp

#include <iostream>
#include <random>
int main() {
unsigned s = std::random_device()(); // Set seed from random_device
std::mt19937 g(s); // Initialize URNG
std::cout << "Seed set to " << s << "\n\n";
std::cout
<< "Sampling doubles exactly from the unit normal distribution:\n";
for (int i = 0; i < 48; ++i)
std::cout << N(g) << ((i + 1) % 8 ? ' ' : '\n');
std::cout << "\n";
}

Compile this code with a C++11 compliant compiler setting the include path to the parent of the exrandom directory, e.g.,

g++ -o simple_normal -O3 -std=c++11 -I../include simple_normal.cpp

Running the resulting program gives, for example,

Seed set to 1930225588

Sampling doubles exactly from the unit normal distribution:
-0.86918 -2.04981 1.62507 0.37086 -0.0259856 0.179536 -1.37951 2.17251
0.608488 0.700322 -1.26223 0.0148746 -1.10179 0.155343 -0.0498893 -0.298053
-1.12402 1.1431 0.355696 -0.203227 -1.35181 -0.0228599 -1.4729 -0.333065
-1.25213 1.47739 -0.159445 1.27897 -0.649038 0.915326 0.0316318 1.33156
0.135598 -1.1963 1.02445 0.195189 0.0944316 1.89514 0.928068 -0.933905
0.0120612 0.293793 1.84805 1.12541 -1.44845 0.183437 0.676166 1.68539

In this example you can substitute unit_exponential_distribution or unit_uniform_distribution for unit_normal_distribution; you can also change double, the template parameter for the constructor, to float or to long double (or even to a multi-precision floating point type, see Multi-precision floating-point types). See simple_exponential.cpp and simple_uniform.cpp.

And here is how to sample exactly from the discrete normal distribution. The code is simple_discrete_normal.cpp

#include <iostream>
#include <random>
int main() {
unsigned s = std::random_device()(); // Set seed from random_device
std::mt19937 g(s); // Initialize URNG
std::cout << "Seed set to " << s << "\n\n";
std::cout
<< "Sampling exactly from the discrete normal distribution,\n"
<< " mu = " << N.mu_num() << "/" << N.mu_den()
<< ", sigma = " << N.sigma_num() << "/" << N.sigma_den() << ":\n";
for (int i = 0; i < 100; ++i)
std::cout << N(g) << ((i + 1) % 20 ? ' ' : '\n');
std::cout << "\n";
}

Running this program gives, for example,

Seed set to 3530269933

Sampling exactly from the discrete normal distribution,
  mu = 1/3, sigma = 129/2:
66 37 -17 107 -108 34 -51 -8 29 -55 31 111 7 -122 36 -43 39 67 31 -65
-39 106 -11 91 -45 -49 -32 16 -90 53 61 56 -122 46 -37 -53 -39 10 -34 -69
39 -83 -46 -58 -5 -24 77 -65 51 -38 77 -55 -20 -24 27 -110 119 40 -2 10
42 53 69 244 -5 2 21 -9 -119 -4 -38 -86 67 -33 -21 115 -121 61 56 19
66 -30 63 37 19 -38 50 -23 122 -59 3 88 75 66 -15 -107 15 105 -38 85
Forward to Getting started. Up to Contents.