[Rd] means in arima0 (PR#754)

Prof Brian D Ripley ripley@stats.ox.ac.uk
Thu, 30 Nov 2000 14:38:12 +0000 (GMT)


On Thu, 30 Nov 2000 arto.luoma@uta.fi wrote:

> Full_Name: Arto Luoma
> Version: 1.1.0
> OS: Windows 98
> Submission from: (NULL) (153.1.53.119)
> 
> 
> In arima0 it is possible to specify whether the mean of the original series is
> included in the model or not. However, it is not possible to specify whether the
> mean of the differenced series is included. It seems that it is not included.

Right, ARMA processes have zero mean, by definition, and the definition is
on the help page.  I don't see any room for ambiguity in what is written
there.

> However, if differencing is used to eliminate trend, the mean of the differenced
> series is not zero. In my opinion, this leads to errors both in estimation and
> prediction problems. The uncertainty in the estimation of the mean of the
> differenced series should be also included  in the standard errors of
> prediction. 

This is not a bug: the function works as described.  If you want it to work
differently, please contribute code to enhance it.  Note that arima0 is
explicitly described as a preliminary version, and it is planned one day to
replace it. So, please supply what you see as a better version as a
candidate for replacement.

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._