ARFIMAResult.is_stationary

property ARFIMAResult.is_stationary: bool

Whether the process is covariance stationary.

Overrides the short-memory test because stationarity here has two conditions, not one: the fractional order must satisfy |d| < 0.5 and the short-memory autoregressive block must be stable. Inheriting the AR-only test would report a unit-root-like d = 0.9 process as stationary purely because its ARMA block is well behaved.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> res = ARFIMA(rng.standard_normal(400), order=(1, 0)).fit()
>>> res.is_stationary == (abs(res.d) < 0.5 and res.stability.is_stable)
True