StochasticVolatilitySVARResult#

class cultivars.multivariate.structural.stochastic_volatility.StochasticVolatilitySVARResult(*, source, impact_draws, h_draws, phi_draws, sigma2_draws, shock_names, relabel_rate, restriction, n_draws, n_burn, thin)[source]#

The posterior of a stochastic-volatility identification.

Variables:

  • source – The reduced-form result identified.

  • impact_draws – (S, k, k) posterior draws of the impact matrix, already relabelled to one order and sign convention.

  • h_draws – (S, T, k) posterior draws of the structural log variances, zero sample mean per shock.

  • phi_draws – (S, k) draws of each shock’s log-variance persistence.

  • sigma2_draws – (S, k) draws of each shock’s log-variance innovation variance.

  • shock_names – One label per shock column.

  • relabel_rate – Fraction of kept draws whose columns had to be permuted to match the reference labelling.

  • restriction – The identifying assumption, stated.

  • n_draws – Total sampler iterations.

  • n_burn – Burn-in discarded.

  • thin – Post-burn thinning.

Attributes#

source

impact_draws

h_draws

phi_draws

sigma2_draws

shock_names

relabel_rate

restriction

n_draws

n_burn

thin

impact

The pointwise posterior median impact matrix, (k, k).

k_endog

Number of variables.

n_kept

Posterior draws retained.

names

Variable labels, from the reduced form.

Methods#

convergence(*chains[, include, exclude, ...])

Assess the sampler's convergence over the retained draws.

fevd([horizon, quantiles])

Pointwise posterior quantiles of the variance decomposition.

identification_test(*[, n_blocks, window])

Test the identifying condition pair by pair on the posterior-median shocks.

irf([horizon, quantiles, cumulative])

Pointwise posterior quantiles of the impulse responses.

irf_draws([horizon, cumulative])

Structural impulse responses for every posterior draw.

structural_shocks()

Posterior-median structural shocks B**-1 u_t, (T, k).

summary()

The estimation summary, renderable as text, HTML, or a dataframe.

to_point()

The posterior-median impact as a point-identified structural result.

volatility_path(shock, *[, quantiles])

Posterior quantiles of one shock's standard deviation over the sample.