Skip to content

giacomocavalieri/prng

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

60 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

prng

Package Version Hex Docs

🎲 A Pure Random Number Generator (PRNG) for Gleam

⚙️ This package works for both the Erlang and JavaScript target

Installation

To add this package to your Gleam project:

gleam add prng

Generating random values

This package can help you generate any kind of random values you can think of. It can be useful in many different scenarios: when you need to simulate non-deterministic actions, like rolling a dice or flipping a coin; when you need to write property based tests; or to make fun interactive games where you can spawn randomly generated enemies!

A mindset shift

The way random values are generated may be a bit confusing at first, especially coming from other languages like JavaScript. In such languages, random number generation can be as simple as this:

const random_value: number = Math.random()

However, this should raise a lot of questions: what if I need those numbers to be in a certain range? How can I generate more complex values, like lists of numbers? Can I set the random seed to get deterministic and reproducible results in a test environment?

The random package tries to address all these questions by providing a nice interface to define random value generators. Let's have a look at an example to get a taste of what generating random values will look like:

let generator: Generator(Float) = random.float(0.0, 1.0)
let random_value: Float = random.sample(generator)

Notice a subtle but fundamental difference: you're no longer simply generating a value, you're describing the values you want to generate and you can take those out of a generator with a variety of functions, like sample.

This neat trick can give two great features:

  • Composability: it's easy to describe simple generators and compose them together to generate complex data structures in an expressive way. The library has a rich API to create and compose generators, you can have a look at it here.
  • Reproducibility: you can decide the random seed used to generate the random values from a generator. The algorithm for the pseudo number generation will always yield the same results given the same starting seed!
    The random package also goes out of its way to make sure that the random number generation works exactly the same on all Gleam targets, so you won't get any discrepancies just by compiling to different targets

References

This package and its documentation are based on the awesome Elm implementation of Permuted Congruential Generators. They have great documentation full of clear and useful examples; if you are curious, give it a look and show Elm some love!

Contributing

If you think there's any way to improve this package, or if you spot a bug don't be afraid to open PRs, issues or requests of any kind! Any contribution is welcome 💜