[R] vars package - irf() does not work

John Dougherty jwd at surewest.net
Fri Feb 24 01:46:48 CET 2017

On Wed, 22 Feb 2017 09:57:00 +0000
"T.Riedle" <tr206 at kent.ac.uk> wrote:

> Dear all,
> I have not received any response on this email. Is there anybody who
> can help me?
> I want to run an impulse response analysis using the vars() package.
> The code looks as follwows.
> # list of class varest
> varest.USA<-VAR(VAR_analsis_DataUSA, lag.max = 24, ic = "SC", type =
> "both")
> varest.USA
> summary(varest.USA)
> #Run irf analysis
> irf.USAg<-irf(varest.USA, response = "g", n.ahead = 48, boot = TRUE,
> ci=0.95)
> plot(irf.USAg)
> The problem is that R returns an error that the arguments in irf are
> unused. That is, unused arguments (response="g", n.ahead = 48, boot =
> TRUE, ci=0.95) The strangeness is that it sometimes works but most of
> the time it does not. I installed vars() last month and irf() worked
> well but now it does only occasionally. I have just edited the data
> but kept the code unchanged.
> In addition, I have the same problem when I am trying to replicate
> the example on irf() in the vars vignette althoug it also worked well
> when I installed vars and run the example.
> Does anybody have an idea what is wrong?
Two guesses are 1) you are not consistently entering the commands and
thus get variable outcomes, and 2) possibly there are "depends" that
irf() needs that need to updated.  You could also contact the package
creators.  I would also observe that this looks remarkably like you
might be asking for help on your homework.  

I would create a script of the commands you use as they work, each
entered in sequence as you used it. You use it essentially as a lab
notebook that documents procedures.  RStudio works well for this, since
you can copy each line that works from the Console to a script file in
the editor.  You can do the same with a text editor, while using R from
a console, but it is somewhat more clumsy.  



More information about the R-help mailing list