Overview

Miso.Random provides a pseudo-random number generator for miso components and their test infrastructure. It is built on the SplitMix32 algorithm, implemented as a stateful JavaScript function stored in a Function.

Two usage styles are available:

• Explicit generator — pass a StdGen through your code using next (analogous to System.Random).
• Global generator — use replicateRM or access globalStdGen directly for fire-and-forget random values.

Quick start

import Miso.Random

-- Explicit generator
example :: IO ()
example = do
  gen         <- newStdGen
  let (v, g') = next gen     -- v :: Double in [0, 1)
  print v

-- Global generator (convenience)
tenValues :: IO [Double]
tenValues = replicateRM 10

-- Reproducible seed for tests
deterministicGen :: StdGen
deterministicGen = mkStdGen 42

Seeding

• newStdGen seeds from crypto.getRandomValues() — cryptographically random, non-reproducible.
• mkStdGen takes an explicit Seed (Int) — reproducible, useful for property tests or simulations.
• globalStdGen is seeded once at module load time from Math.random().

See also

• splitmix32 — the raw JS PRNG primitive
• getRandomValue — crypto.getRandomValues() used for seeding