[19177] in s-news-athena
[S] arima.diag residuals and arima.filt
daemon@ATHENA.MIT.EDU (John Balding)
Thu Sep 9 19:33:37 1999
Message-Id: <19990909232904.77004.qmail@hotmail.com>
From: "John Balding" <jbalding@hotmail.com>
To: s-news@wubios.wustl.edu
Date: Thu, 09 Sep 1999 16:29:04 PDT
Mime-Version: 1.0
Content-Type: text/plain; format=flowed
I have a (probably simple) question about ARIMA modeling in Splus 2000. How
are the residuals from arima.diag calculated? Adding the residuals to the
original series gives values that are different from the predictions from
arima.filt. For example:
>al.mod <- list(list(order=c(0,1,1)), list(order=c(0,1,1),
period=12))
>fit <- arima.mle(ship, model=al.mod)
>f1 <- arima.diag(fit, resid=T, plot=F)$resid + ship
>f2 <- arima.filt(ship, fit$model)$pred
In this case f1 is not equal to f2. Looking at the source for arima.diag,
the residuals appear to be the input series minus filt$pred... Why aren't
the two the same?
I'm trying to do transfer function modeling and need to filter my output
series using the ARIMA model obtained for my covariate series (prewhitening
before calculating the cross-correlations). The discrepancy between the
arima.diag predictions and the arima.filt predictions has got me wondering
what I'm doing wrong...
Any help would be appreciated.
Thanks,
John
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
