Normal

Random normal initializer which simulate a normal distribution with specified parameters.

Parameters

• mu

  Default → 0.0

  The mean of the normal distribution

• sigma

  Default → 1.0

  The standard deviation of the normal distribution

• seed

  Type → int | None

  Default → None

  Random number generation seed that can be set for reproducibility.

Examples

from river import optim

init = optim.initializers.Normal(mu=0, sigma=1, seed=42)

init(shape=1)

np.float64(0.4967141...)

init(shape=2)

array([-0.1382643 ,  0.64768854])

Methods

call

Returns a fresh set of weights.

Parameters

• shape — defaults to 1