HierarchicalBVARResult.k_endog

property HierarchicalBVARResult.k_endog: int

Number of endogenous variables, k.

The width of endog and the length of names; the trailing axis of beta_mean, resid, and fittedvalues, both axes of sigma_u, and the dimension of the inverse-Wishart posterior, whose prior degrees of freedom are k + 2. A large k is what this model exists for: with a proper prior the sample-length constraint is one observation, not one per regressor.