R help - Sep 2012 - Questions about the functions ar.ols and auto.arima when fitting an AR model

Hi,

  I am trying to fit an AR model, maximum order =4, order selection
criterion is aic. I wonder why these two give different results:

   m1<-ar.ols(x, aic=TRUE, method="ols", order.max=4)

   m1<-auto.arima(x,d=0, D=0, max.p=4, max.P=0, max.q=0, max.Q=0,
ic="aic")

   Could they both use the function predict to do forecasting? Is there any
function that works better?

        temp1<-predict(m1, n.ahead=4)
        PI.fore[i,j]<-temp1$pred[4]

    Thanks

Miao

	[[alternative HTML version deleted]]