[R] Prediction from Arima model

[Rui Barradas]

Às 18:54 de 31/08/2024, Christofer Bogaso escreveu:
> Hi,
> 
> I have run following code to obtain one step ahead confidence interval
> from am arima model
> 
> library(forecast)
> 
> set.seed(100)
> 
> forecast(Arima(rnorm(100), order = c(1,0,1), xreg = rt(100, 1)), h =
> 1, xreg = 10)
> 
> However this appear to provide the Prediction interval, however I
> wanted to get the confidence interval for the new value.
> 
> Is there any way to get the confidence interval for the new value?
> 
> I also wanted to get the estimate of SE for the new value which is
> used to obtain the confidence interval of the new value. Is there any
> method available to obtain that?
> 
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
Hello,

To get the se use ?predict.Arima instead.


library(forecast)

set.seed(100)

model <- Arima(rnorm(100), order = c(1,0,1), xreg = rt(100, 1))
# in predict.Arima, se.fit defaults to TRUE
pred <- predict(model, n.ahead = 1, newxreg = 10)

pred$se
c(pred$se)

# with more points ahead
predict(model, n.ahead = 2, newxreg = c(10, 12))



Hope this helps,

Rui Barradas


-- 
Este e-mail foi analisado pelo software antivírus AVG para verificar a presença de vírus.
www.avg.com