This is basically a one-to-one port of Processing's noise() functionality from Java to Javascript for Node.js. Since Javascript's Math.random() has no seeding capability, I've used the wonderful Alea pseudo-random number generator instead.
The typical way:
npm install proc-noise
And that's it.
Using it is pretty simple:
var PerlinGenerator = require("proc-noise");
var Perlin = new PerlinGenerator(); // seeds itself if no seed is given as an argument
console.log( Perlin.noise( 817.2 ) ); // one dimensional
console.log( Perlin.noise( 9192, 818.53 ) ); // two dimensional
console.log( Perlin.noise( 5, 7, 9.22 ) ); // three dimensional
Initializing the module returns a randomly-seeded Perlin Noise object. If you want, you can pass it a seed either when initializing it with new, or later using the noiseSeed(seed) method.
The generated Perlin object has four methods: noise(x, y, z), noiseSeed(seed), noiseReseed(), and noiseDetail(lod, falloff).
The noise(x, y, z) function can take one, two, or three arguments, depending on how many dimensions you want to use. This function will always return a number between 0.0 and 1.0.
By default, the Perlin object seeds itself using a random Alea seed, but you can set your own seed using noiseSeed(seed). This obviously resets the generator with new values.
To reseed the Perlin object with a new random seed, just use noiseReseed(). This obviously resets the generator with new values.
To change the level of detail and falloff of the Perlin object, use noiseDetail(lod, falloff). The falloff argument is optional. By default, the level of detail is 4 octaves, and the amplitude falloff is 0.5. To be totally honest, I don't fully understand the math of this, but it's here if you want to use it.
Two examples are provided in the test/ directory. To run them, use node, of course. test.js just prints out some noise values. visual.js is more complicated, providing a webpage for you to see the noise function in action in two dimensions via the often-seen randomly-generated noise cloud.
The original code from Processing is deep in here. It has plenty of comments if you're interested in learning more about it.
I tried using the simplex-noise module and it gave me very different results than what I was used to in Processing. I'm not sure why, so I just ported the Processing noise() implementation over as the quickest solution. That's probably a silly way to go about it, but whatever, lol.