ssfortran.ARIMA

class ssfortran.ARIMA(order=(1, 0, 0), seasonal_order=(0, 0, 0, 0), series=0, enforce_stationarity=True, enforce_invertibility=True, at_observations=False)

ARIMA(p, d, q)(P, D, Q)_s component (DK §3.4).

The differences are states (diffuse), as in DK §3.4 and statsmodels’ SARIMAX; the ARMA part is stationary. Parameters: the AR, seasonal AR, MA and seasonal MA coefficients, then the innovation variance.

Parameters:

ordertuple of int

(p, d, q).

seasonal_ordertuple of int

(P, D, Q, s), with D at most 1.

seriesint

The series (0-based).

enforce_stationarity, enforce_invertibilitybool

Keep the AR polynomials stationary and the MA polynomials invertible (Monahan 1984).

at_observationsbool

Act on the observations rather than the signals.