arch.univariate.Normal¶

class arch.univariate.Normal(*, seed: int | RandomState | Generator | None = None)[source]¶

Standard normal distribution for use with ARCH models.

Parameters:¶

seed: int | RandomState | Generator | None = None¶

Random number generator instance or int to use. Set to ensure reproducibility. If using an int, which must not be negative, the argument is passed to np.random.default_rng. If not provided, default_rng is used with system-provided entropy.

Raises:¶

• TypeError – If seed is not None, an integer, a Generator or a RandomState.

• ValueError – If seed is a negative integer.

Methods

bounds(resids)	Construct parameter bounds for use in optimization.
cdf(resids[, parameters])	Compute the cumulative distribution function.
constraints()	Construct arrays to use in constrained optimization.
loglikelihood(parameters, resids, sigma2[, ...])	Compute the log-likelihood of normally distributed residuals.
moment(n[, parameters])	Compute the moment of order n.
parameter_names()	Return the names of the distribution shape parameters.
partial_moment(n[, z, parameters])	Compute the order n lower partial moment from -inf to z.
ppf(pits[, parameters])	Compute the inverse cumulative density function (ICDF).
simulate(parameters)	Simulate i.i.d.
starting_values(std_resid)	Construct starting values for use in optimization.

Properties

generator	The NumPy Generator or RandomState attached to the distribution.
name	The name of the distribution.