Hi,
I used the package vars for my project but when the impulse response plots
were produced, some of the levels are out of the confidence band. I was
wondering what the problem is and also, does anyone know what type of
bootstrapping the package is using: is it parametric, case resampling,
gauss
Hi, I have a similar question:
ir <- irf(varsumm, impulse=c("prod", "rea", "rpo") n.ahead=20, runs=500,
ci=0.95)
will calculate the orthogonalized impulse responses from "prod", "rea", and
"rpo", i.e. a (1, 1, 1)' vector. What do I need to do to make the impulse
(-1, 1, 1)', i.e. I want the the
2 matches
Mail list logo