[R] Loop Autoregression

[Jonas Ulbrich]

Hello everybody, I have to confess that I am relatively new to R.

My problem is the following one:

I have a data set of 500 values. For each value I wanna make prediction 
of four steps ahead via autogregression. Because I have to repeat this 
procedure on three other data sets I wanna make a loop. (Manual 
calculation would take ages.)
My idea which does not work looks is as follows:

x<-NULL
for (i in 1:500){
   
x[i]<-predict(ar(y[i],order.max=1,method="yule-walker",intercept=TRUE, 
n.used=500, n.ahead=4))
}

Doing this I only get an error request that the value of "order.maX" 
needs to be smaller than "n.used". But this is the case. Isn't it?

Best Regards