MarkovSwitchingSVAR.identify#

MarkovSwitchingSVAR.identify()[source]#

Identify every regime under one declaration and package the answers.

Return type:

MarkovSwitchingSVARResult

Returns:

The per-regime structural result. When no identifications were supplied, each regime is factorized recursively under the same ordering – the restriction pattern held fixed while the parameters switch.

Raises:

  • SpecificationError – If a declared ordering is not a permutation of names.

  • NumericalError – If a regime’s innovation covariance is not positive definite.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> s = np.repeat([0, 1, 0, 1], 100)
>>> y = np.array([[-1.0, -0.5], [1.5, 1.0]])[s] + 0.4 * rng.standard_normal((400, 2))
>>> res = MarkovSwitchingVAR(y, order=1, n_regimes=2, names=("a", "b")).fit(seed=0)
>>> svar = MarkovSwitchingSVAR(res, order=("b", "a")).identify()
>>> svar.shock_names, [float(svar.impact(m)[1, 1]) for m in range(2)]
(('b', 'a'), [0.0, 0.0])
>>> MarkovSwitchingSVAR(res, order=("b", "c")).identify()
Traceback (most recent call last):
cultivars.exceptions.SpecificationError: order must be a permutation of the variable ...