!NPM Version
!downloads
!David
!David
 

seeded-random-utilities

Common random functions that are seedable written in TypeScript with TypeScript support.

The rand-seed npm package provides a random number generator similar to Math.random except with seeding. This package uses rand-seed to provide random numbers, but also implements some common easy-to-use random utilities.

Installation

This package is available through _npm_:

``
npm install --save seeded-random-utilities
`

Usage

Either import directly

`html

`

or import in your own scripts using

`javascript
import SeededRandomUtilities from 'seeded-random-utilities';
`

Then simply create a new instance with an (optional) seed:

`javascript
const rand = new SeededRandomUtilities('1234');

rand.getRandomBool(); // Generate a new random number
`

If no seed is specified the call to rand.random() will simply be forwarded to Math.random(). So it won't operate in a repeatable seeded fashion if no seed is supplied.

`javascript
// Create a new random number generator using the xoshiro128** algorithm
const rand = new SeededRandomUtilities('1234', PRNG.xoshiro128ss);
`

An interface is provided for the main random class, RandomUtilities.

Example

A simple example is included. This may be run with _node_: node sample/index.js

Another example that was used to verify that the package can be installed and used properly can be found in sample/test-random-unique-integers-lists. A test package is in there that one can use to examine large sets of the output from generateRandomArrayOfUniqueIntegers`.

API

| Method | Description |
|:------------------------------|:-------------|
| random(): number, getRandom(): number | Generate a random integer. |
| getRandomIntegar(max: number, min = zero): number | Generate a random integer. |
| getRandomArbitrary(max: number, min = zero): number | Generate a random arbitary. |
| getRandomIntInclusive(max: number, min = zero): number | Generate a random max inclusive integer. |
| getRandomBool(): boolean | Generate a random boolean (true/false). |
| getRandomChar(): string | Generate a random character. |
|selectRandomElement(source: T[]): T||Selects a random element out of the provided array|
|selectUniqueRandomElements(source: T[], picks: number): T[]|Select a number of random unique elments in a provided array|
|shuffle(array: T[], copy?: boolean): T[]|string| Randomly shuffle a provided array|
|chooseBooleanRandomlyWithProbability(itemCount: number, picks?: number): boolean| Choose a number of boolean randomly with the provide percentage|
|generateRandomArrayOfUniqueIntegers(amount: number, maxValue: number): number[]| Choose a list of unique integers out of a list of consecutive integers|

Contributing

Pull requests and stars are highly welcome.

For bugs and feature requests, please create an issue.

Motivations

In a recent project I was working on I needed the use of a seeded random number generator, preferably written in TypeScript. I found rand-seed, but while it provided the basic engine to run randomization it didn't have any implementation for picking unique numbers out of a set or really anything other than the root output similar to Math.random(). I had been waiting for an excuse to make a TypeScript style npm package, in order to provide tangible proof of my understanding of TypeScript. So this seemed like the perfect excuse to attempt to fill a narrow niche with a package that had a set of random utilities as well as one that was seeded and in TypeScript.
I started reading different TypeScript npm package tutorials and finally settled on using How to Create and Publish an NPM module in TypeScript
I first attemptted implementing simple utilities like generating a random integer. For and more, I kept seeing people point to the [MDN article on Math.Random], and noticed that it some nice basic implementations of many common random utilities. So I adapted most of those utilities to the new seeded TypeScript package I was buliding. Then I noticed the random-utility - npm, while I was getting close to have creating a unique element picking algorithm. I adapted some of the easy to adapte to seeded TypeScript versions.
Finally I noticed more and more I liked the TypeScript project setup in rand-seed and ended up copying more and more of it i into my project. It gave me chance to work on rollup which I hadn't gotten into yet. Thanks to all mentioned and those unmentioned that I don't know of.

Links

* "rand-seed - npm" - Provides the seeded random implementation powering the utilities. Also directly used a lot of the TypeScript setup and project design from here, blending it partly with the following tutorial link and other random utilities.
* "How to Create and Publish an NPM module in TypeScript" - One of the best tutorials on how to publish a TypeScript package that is usefulable in both JavaScript and TypeScript.
* "Math.random() - JavaScript | MDN" - This MDN web doc provided the inspriation and logic behind many of the utilities. They were adpated to both seeding and into TypeScript.
* "random-utility - npm" - I created seeded versions of a few of these utitlities. It wasn't a straight rip, more just inspiration.