Rotate-Multiply (Romu) Generators

class randomgen.romu.Romu(seed=None, variant='quad')

Mark A. Overton’s quad and trio rotate-multiply-based generators


A seed to initialize the BitGenerator. If None, then fresh, unpredictable entropy will be pulled from the OS. If an int or array_like[ints] is passed, then it will be passed to SeedSequence to derive the initial BitGenerator state. One may also pass in a SeedSequence instance.


The variant to use. The Quad variant is somewhat slower but has a larger state.


Romu is a rotate-multiply based generator either either a 256-bit state (default, “quad”) or a 192-bit state (“trio”) ([1], [2]). Romu has a large capacity and has a tiny chance of overlap, especially when using the default “quad” variant. It is extremely fast.

Romu provides a capsule containing function pointers that produce doubles, and unsigned 32 and 64- bit integers. These are not directly consumable in Python and must be consumed by a Generator or similar object that supports low-level access.

State and Seeding

The Romu state vector consists of 3 (“trio”) or 4 unsigned 64-bit values. The input seed is processed by SeedSequence to generate all values, then the Romu algorithm is iterated a small number of times to mix.

Compatibility Guarantee

Romu makes a guarantee that a fixed seed will always produce the same random integer stream.


Romu supports parallel application using distinct seed values.

>>> from randomgen import SeedSequence, Romu
>>> NUM_STREAMS = 8192
>>> seed_seq = SeedSequence(489048146361948)
>>> bit_gens = [Romu(child) for child in seed_seq.spawn(NUM_STREAMS)]
>>> [bg.random_raw() for bg in bit_gens[:3]]
[9130649916277841551, 2156737186088199787, 12803009197309261862]


Seeding and State


Seed the generator


Get or set the PRNG state



CFFI interface


ctypes interface


random_raw([size, output])

Return randoms as generated by the underlying BitGenerator