GibbsBVARResult.forecast_paths

GibbsBVARResult.forecast_paths(steps=8, *, seed=None)

The posterior predictive’s raw simulated paths, one per kept draw.

Each retained draw simulates its own future – its coefficients, its covariance, its innovations – and this method hands the paths over unsummarized: the object a forecast evaluation actually scores, and what a fan chart takes quantiles of. forecast() is these paths’ three-number summary.

Parameters:

• steps (int) – Horizons ahead.

• seed (int | Generator | None) – Seed or generator for the predictive shocks.

Return type:

NDArray[double]

Returns:

An array of shape (n_kept, steps, k).

Raises:

SpecificationError – If steps is not positive.