Hi Sven ,
question 1 solved, thanks for the help.
Regarding question 2 , I created a column vector  and I receive another error message. 
Matrices not conformable for operation
*** error within loop in function fcast_setup

matrix ED = A_inter + D + E

Yes, I can replicate that, I'm afraid this is simply a bug. (Actually, looking at the code I found two separate ones.) To be honest, the case of having exogenous variables in forecasting apparently hasn't really been tested yet in the BVAR package, so you are implicitly our guinea pig, even though we didn't plan for that. But that's why we often start with version numbers like 0.1 when we release something for the first time.

I'm quite optimistic that we (Luca and I) can fix that and update the BVAR package to version 0.3 soon. However, right now I can't tell you how to work around this issue in the meantime, sorry.

(BTW, Luca told me he didn't get your message to the mailing list, it's not clear why. Otherwise he might have commented as well.)

cheers

sven