auto_psi about bvar HOT 2 CLOSED

While it is true that Giannone et al. (2015) fit an AR(1) to obtain the innovations for scaling, we followed common practice in the literature and set it to an AR(lags) for the automated setting of psi (see e.g. Banbura et al. (2010) https://doi.org/10.1002/jae.1137, Sims and Zha (1998) https://doi.org/10.2307/2527347, Litterman (1986) https://doi.org/10.2307/2527347 )

However, there is an option to provide values for the mode of psi in form of a vector via the argument mode in bv_psi() manually should there be a need for experimenting without the need of adding another parameter to the functions.

from bvar.

Ok, thank you!

from bvar.