simba

Note: this project is still in active development and not all implementations have corresponding testcase and as such have not been proven to be correct, please hold off on using this until these tests are added and this notice is removed.

This project provides implementations of pseudorandom number generators in pure Nim. It can be used to generate the same sequences as other PRNGs provided you know the parameters they use.

All PRNGs implemented in simba are not cryptographically secure. If you don't know if you need a CSPRNG or a PRNG you should opt for the first one and skip this library.

Usage

simba provides a high level API that works across all provided PRNGs. Take care that this high level API might implement these functions differently from other implementations even if the underlying PRNG yields the same values.

import simba
import simba/lcg
import simba/lfsr

let
  prng0 = newRANDU()
  prng1 = newXorShift64s()

echo prng0.randBits(8)
echo prng1.randBits(8)

If you need to do other conversions you can use the PRNGs directly, they will yield next values in the correct width.

import simba/mersenne

let prng = newMT19937()

echo next(prng)
echo next(prng)

PRNGs

simba currently implements the following types and specific PRNGs.

Linear Congruential Generators

Linear Feedback Shift Registers

Mersenne

MT19337
MT19337-64

Bindings

Work has started on providing Python bindings for simba, you can find them in the sarafi project.