Randlibjs

A lightweight JavaScript library for generating random numbers with various probability distributions.

Features

• Multiple Distributions: Generate random numbers from various probability distributions including normal, uniform, exponential, and more
• Reproducible Results: Set seeds to generate reproducible sequences of random numbers for testing and validation
• Lightweight: Minimal dependencies and optimized code for fast performance in both browser and Node.js environments
• Easy to Use: Simple and intuitive API with comprehensive documentation and examples

Installation

npm

npm install randlibjs

CDN

<script src="https://cdn.jsdelivr.net/npm/randlibjs@latest/dist/randlib.min.js"></script>

Quick Start

ES Modules

// Import specific functions
import { normal, uniform, seed } from 'randlibjs';

// Set a seed for reproducibility
seed(12345);

// Generate 10 numbers from normal distribution
const normalSample = normal(0, 1, 10);
console.log(normalSample);

// Generate 5 uniformly distributed numbers
const uniformSample = uniform(0, 10, 5);
console.log(uniformSample);

Browser

<script src="https://cdn.jsdelivr.net/npm/randlibjs@latest/dist/randlib.min.js"></script>
<script>
  // Set a seed
  randlib.seed(12345);
  
  // Generate random numbers
  const normalSample = randlib.normal(0, 1, 10);
  console.log(normalSample);
</script>

Available Distributions

Continuous Distributions

Function	Description	Parameters
uniform(low, high, size)	Uniform distribution	low: Lower bound (default: 0) high: Upper bound (default: 1) size: Sample size (default: 1)
normal(mean, std, size)	Normal (Gaussian) distribution	mean: Mean (default: 0) std: Standard deviation (default: 1) size: Sample size (default: 1)
exponential(lambda, size)	Exponential distribution	lambda: Rate parameter (default: 1) size: Sample size (default: 1)
cauchy(loc, scale, size)	Cauchy distribution	loc: Location parameter (default: 0) scale: Scale parameter (default: 1) size: Sample size (default: 1)
lognormal(mean, sigma, size)	Lognormal distribution	mean: Mean of logarithm (default: 0) sigma: Standard deviation of logarithm (default: 1) size: Sample size (default: 1)
pareto(alpha, size)	Pareto distribution	alpha: Shape parameter (default: 1) size: Sample size (default: 1)
triangular(low, high, mode, size)	Triangular distribution	low: Lower bound (default: 0) high: Upper bound (default: 1) mode: Mode (default: 0.5) size: Sample size (default: 1)

Discrete Distributions

Function	Description	Parameters
randint(low, high, size)	Discrete uniform distribution	low: Lower bound (included, default: 0) high: Upper bound (excluded) size: Sample size (default: 1)
poisson(lambda, size)	Poisson distribution	lambda: Rate parameter (default: 1) size: Sample size (default: 1)
binomial(n, p, size)	Binomial distribution	n: Number of trials p: Success probability size: Sample size (default: 1)
geometric(p, size)	Geometric distribution	p: Success probability size: Sample size (default: 1)
chisquare(df, size)	Chi-square distribution	df: Degrees of freedom size: Sample size (default: 1)

Multivariate Distributions

Function	Description	Parameters
multivariateNormal(mean, cov, size)	Multivariate normal distribution	mean: Mean vector cov: Covariance matrix size: Sample size (default: 1)

Utilities

Function	Description	Parameters
seed(value)	Set random seed	value: Seed value
shuffle(array)	Randomly shuffle array	array: Array to shuffle
permutation(n)	Random permutation of integers	n: Number of elements
mixture(distributions, weights, size)	Mixture of distributions	distributions: Array of distribution functions weights: Array of weights size: Sample size (default: 1)

Examples

Monte Carlo Simulation

import { uniform } from 'randlibjs';

function estimatePi(samples = 1000) {
  let insideCircle = 0;
  
  for (let i = 0; i < samples; i++) {
    const x = uniform(-1, 1);
    const y = uniform(-1, 1);
    
    if (x*x + y*y <= 1) {
      insideCircle++;
    }
  }
  
  return 4 * insideCircle / samples;
}

// Estimate π with 10,000 samples
console.log(estimatePi(10000));

Random Walk Simulation

import { randint } from 'randlibjs';

function randomWalk(steps = 100, p = 0.5) {
  let position = 0;
  const positions = [position];
  
  for (let i = 0; i < steps; i++) {
    // Movement: +1 with probability p, -1 with probability 1-p
    const step = Math.random() < p ? 1 : -1;
    position += step;
    positions.push(position);
  }
  
  return positions;
}

// Perform a random walk of 100 steps
const walk = randomWalk(100);
console.log(walk);

Documentation

For detailed documentation and more examples, visit:

• Official Website
• API Documentation
• GitHub Repository

License

BSD-3-Clause License

Author

Adrián Muñoz Perera