Random Number Generation

Numpy’s random number generation

Stochastic relies on numpy for random number generation. Since numpy 1.17, the newer Generator objects provide improved performance:

Note

From numpy docs: The Generator’s normal, exponential and gamma functions use 256-step Ziggurat methods which are 2-10 times faster than NumPy’s Box-Muller or inverse CDF implementations.

By default, the stochastic package uses numpy’s faster Generator random number generation. With a function call, we can change the default back to the legacy random number generation, which uses RandomState objects.

If no rng arg is passed when instantiating process instances, each instance will reference the stochastic.random module’s generator attribute for random number generation.

Examples

Changing the default random number generation on instances without specified rng:

from stochastic.processes import GaussianNoise
from stochastic import random

gn = GaussianNoise()
print(gn.rng)
# Generator(PCG64)

# use the legacy random number generator
random.use_randomstate()

print(gn.rng)
# <module 'numpy.random' from '/path/to/site-packages/numpy/random/__init__.py'>

# use the newer Generator
random.use_generator()

print(gn.rng)
# Generator(PCG64)

Setting the seed value:

from stochastic.processes import GaussianNoise
from stochastic import random

gn = GaussianNoise()
print(gn.rng)
# Generator(PCG64)

random.seed(42)
print(gn.rng.bit_generator.state)
# {'bit_generator': 'PCG64', 'state': {'state': 274674114334540486603088602300644985544, 'inc': 332724090758049132448979897138935081983}, 'has_uint32': 0, 'uinteger': 0}
print(gn.sample(4))
# [ 0.15235854 -0.51999205  0.3752256   0.47028236]

random.seed(42)
print(gn.rng.bit_generator.state)
# {'bit_generator': 'PCG64', 'state': {'state': 274674114334540486603088602300644985544, 'inc': 332724090758049132448979897138935081983}, 'has_uint32': 0, 'uinteger': 0}
print(gn.sample(4))
# [ 0.15235854 -0.51999205  0.3752256   0.47028236]

Passing custom generators to process instances at instantiation:

from numpy.random import Generator
from numpy.random import PCG64
from stochastic.processes import GaussianNoise
from stochastic import random

generator = Generator(PCG64(seed=42))

gn = GaussianNoise(rng=generator)
# generator specific to this gaussian noise instance
print(gn.rng.bit_generator.state)
# {'bit_generator': 'PCG64', 'state': {'state': 274674114334540486603088602300644985544, 'inc': 332724090758049132448979897138935081983}, 'has_uint32': 0, 'uinteger': 0}

# stochastic's global generator, different from the one attached to `gn`
print(random.generator.bit_generator.state)
# {'bit_generator': 'PCG64', 'state': {'state': 228239801863081385502825691348763076514, 'inc': 61631449755775032062670113901777656135}, 'has_uint32': 0, 'uinteger': 0}

Documentation

stochastic.random.generator = Generator(PCG64) at 0x7FE9A17D09E0

The default random number generator for the stochastic package

stochastic.random.seed(value)

Sets the seed for numpy legacy or default_rng generators.

If using the legacy generator, this will call numpy.random.seed(value). Otherwise a new random number generator is created using numpy.random.default_rng(value).

stochastic.random.use_generator(rng=None)

Use the new numpy Generator as default for stochastic.

Sets the default random number generator for stochastic processes to the newer np.random.default_rng().

Note

This is the default generator and there is no need to call this function unless returning to the default after switching away from it.

Parameters

rng (numpy.random.Generator) – a Generator instance to use as the default random number generator for stochastic.

stochastic.random.use_randomstate(rng=None)

Use the legacy numpy RandomState generator as default for stochastic.

Sets the default random number generator for stochastic processes to the legacy np.random.

Parameters

rng (numpy.random.RandomState) – a RandomState instance to use as the default random number generator for stochastic.