MarkovSwitchingARResult.label_ordering

property MarkovSwitchingARResult.label_ordering: str

Which quantity regimes were sorted by, ascending.

A mixture likelihood is invariant to relabelling regimes, so the fit imposes an ordering to make two runs of the same specification comparable. The intercept is used when it switches; otherwise the variance; otherwise the first autoregressive coefficient. Reported because the meaning of “regime 0” depends on it: under a variance-switching-only model, regime 0 is the quiet regime, not a low-mean one.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> y = rng.standard_normal(300)
>>> model = MarkovSwitchingAR(y, order=1, n_regimes=2, switching_mean=False)
>>> res = model.fit(seed=0, n_init=2)
>>> res.label_ordering, bool(res.variances[0] <= res.variances[1])
('variance', True)