FIGARCHResult.simulate

FIGARCHResult.simulate(n=200, *, seed=None, burn=_SIMULATION_BURN)

Refused: a long-memory variance has no finite-state recursion to start from.

The FIGARCH variance is an infinite sum over past squared shocks whose weights decay hyperbolically, so no burn-in of fixed length makes a sample forget its start; a truncated simulation would carry a bias the fitted model does not describe.

Parameters:

• n (int) – Ignored.

• seed (int | Generator | None) – Ignored.

• burn (int) – Ignored.

Raises:

SpecificationError – Always.

Return type:

NDArray[double]

Example

>>> rng = np.random.default_rng(0)
>>> FIGARCH(rng.standard_normal(600)).fit().simulate(10)
Traceback (most recent call last):
    ...
cultivars.exceptions.SpecificationError: FIGARCH results do not simulate: ...