MarkovSwitchingARResult.specification#

property MarkovSwitchingARResult.specification: str[source]#

Krolzig code for which blocks switch, e.g. "MSIH(2)-AR(4)".

I marks a switching intercept, A switching autoregressive coefficients, H a switching variance – so the code names the model you actually fitted rather than the family it came from. The constructor guarantees at least one letter is present.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> y = rng.standard_normal(300)
>>> model = MarkovSwitchingAR(y, order=1, n_regimes=3, switching_variance=False)
>>> model.fit(seed=0, n_init=2).specification
'MSI(3)-AR(1)'