RNGenie.Distributions 0.1.0-alpha.2

🎩 RNGenie.Distributions 📈

Lightweight Probability Distributions for Deterministic Sampling

RNGenie.Distributions provides ready-to-use sampling primitives (e.g. Uniform, Triangular, Gaussian-approx)
that integrate with IRandomSource for reproducible simulations.

✨ Features

• Common distributions (e.g. Uniform, Triangular, Gaussian approximation (using Box-Muller algorithm).
• All samplers implement IDistribution<T> → compose and test easily.
• Deterministic when sampled with a seeded IRandomSource.

📄 Documentation

See the Distribution Docs for usage and API details.

🚀 Quick Start

Install Core + Distributions:

dotnet add package RNGenie.Core
dotnet add package RNGenie.Distributions

Basic usage:

using RNGenie.Core.Sources;
using RNGenie.Distributions.Continuous;

// Seedable RNG for reproducibility
var rng = new Pcg32Source(seed: 123);

// Sample from a Gaussian-like distribution
var normal = new Gaussian(mean: 0.0, stdDev: 1.0);
double x = normal.Sample(rng);

// Sample from a triangular distribution
var tri = Triangular(min: 0, mode: 10, max: 20);
double y = tri.Sample(rng);

Console.WriteLine($"Normal: {x:F3}, Triangular: {y:F3}");

Output:

Normal: -0.318, Triangular: 12.474

Implement your own distribution:

using RNGenie.Core.Abstractions;

public sealed class Bernoulli : IDistribution<bool>
{
	private readonly double _p;
	
	public Bernoulli(double p) => _p = Math.Clamp(p, 0.0, 1.0);
	
	public bool Sample(IRandomSource rng) => rng.NextDouble() < _p;
}

🧩 Extensibility

• Build compound models by sampling multiple distributions.
• Swap RNG sources (Pcg32Source, SystemRandomSource, CryptoRandomSource) without changing your model code.
• Use the same seed to replay experiments exactly.

📦 Roadmap

• Swappable Ziggurat algorithm for Gaussian distributions (currently uses Box-Muller).
• Exponential, Poisson, Gamma, Beta, Binomial samplers.
• Alias method for large discrete distributions.

👩‍💻 Contributing

Pull requests are welcome!

Good first issues:

• Add new distributions + unit tests
• Numerical validation harness

See CONTRIBUTING.md for guidance. Also, check here for existing distribution issue writeups.

📜 License

RNGenie is licensed under the MIT License. This means you're free to use it in open source, commercial, or personal projects.