com.tecacet.math.extensions.RandomExtensions.kt Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of math-extensions Show documentation

Koltin extensions for matrix and vector algebra, random number generation, and arithmetic on boolean arrays.

The newest version!

/**
 * Extend java.util.Random to generate random numbers based on various distributions
 */
package com.tecacet.math.extensions

import java.util.Random
import kotlin.math.ceil
import kotlin.math.ln

/**
 * Gaussian with specific mean and standard deviation
 * @param mean the mean
 * @param std the standard deviation
 */
fun Random.nextGaussian(mean: Double = 0.0, std: Double = 1.0) = std * this.nextGaussian() + mean

/**
 * Pick a number uniformly inside the interval [from,to]
 * @param from start of the interval
 * @param to end of the interval
 */
fun Random.nextUniform(from: Double, to: Double) = (this.nextDouble() * (to - from)) + from

/**
 * Perform n binomial trials with probability p. Return the number of successes
 * @param n the number of trials
 * @param p the probability of success for each trial
 */
fun Random.nextBinomial(n: Int, p: Double): Int {
    if (p < 0 || p > 1) throw IllegalArgumentException("Invalid probability value $p.")
    return (0 until n).map { nextDouble() }.filter { it < p }.count()
}

/**
 * Bernoulli distribution: Return true with probability p
 */
fun Random.nextBoolean(p: Double) = nextDouble() < p


/**
 * Generates an exponentially distributed random number with the given mean
 *
 * @param mean expected value of the distribution.
 */
fun Random.nextExponential(mean: Double) = -1.0 * mean * ln(nextDouble())

/**
 * X ~ G(p) : X = [X'~E(-ln(1-p))]
 * @param p the probability
 */
fun Random.nextGeometrical(p: Double): Long {
    if (p < 0 || p > 1) throw IllegalArgumentException("Invalid probability value $p.")
    return ceil(nextExponential(-ln(1.0 - p))).toLong()
}