UnobservedComponentsResult.simulate#

UnobservedComponentsResult.simulate(n=200, *, seed=None, burn=0)[source]#

A fresh sample path at the fitted variances.

The unit-root components – level, slope, seasonal – start at zero rather than from the approximate-diffuse initial covariance the filter uses, so the sample is a random walk from the origin and not from a draw of standard deviation a thousand; the cycle starts from its stationary law. burn initial periods are discarded.

Parameters:

  • n (int) – Observations kept.

  • seed (int | Generator | None) – Seed or generator.

  • burn (int) – Periods discarded from the start.

Return type:

NDArray[double]

Returns:

An array of shape (n,).

Raises:

SpecificationError – If the counts are not usable.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> y = np.cumsum(0.2 * rng.standard_normal(200)) + rng.standard_normal(200)
>>> res = UnobservedComponents(y).fit()
>>> path = res.simulate(100, seed=1)
>>> path.shape, bool(abs(path[0]) < 3.0)
((100,), True)
>>> bool(np.array_equal(path, res.simulate(100, seed=1)))
True
>>> res.simulate(0)
Traceback (most recent call last):
cultivars.exceptions.SpecificationError: n must be positive and burn non-negative; ...