TVPSVARResult.impact

TVPSVARResult.impact(t)

The posterior mean structural impact matrix at date t.

Parameters:

t (int) – Row of the estimation sample.

Return type:

NDArray[double]

Returns:

A lower-triangular (k, k) matrix.

Raises:

SpecificationError – If t is out of range.

Example

Under stochastic volatility the impact drifts from date to date while staying lower triangular:

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> y = np.zeros((140, 2))
>>> for t in range(1, 140):
...     y[t] = 0.5 * y[t - 1] + rng.standard_normal(2)
>>> source = TVPVARSV(y, order=1, training=40).fit(n_draws=40, n_burn=15, seed=0)
>>> svar = TVPSVAR(source).identify()
>>> first, last = svar.impact(0), svar.impact(98)
>>> bool(np.allclose(first, np.tril(first))), bool(np.any(first != last))
(True, True)
>>> svar.impact(99)
Traceback (most recent call last):
cultivars.exceptions.SpecificationError: t must be in 0..98; got 99.