I'm not familiar with the "lm" test. I used to work out the order
starting from plots of the autocorrelation and partial autocorrelation
functions, see "acf" and "pacf" in R. Traditional
estimation can be
handled with the "arima" command.
Beyond this, have you seen the time series discussion on Venables
and Ripley (2002) Modern Applied Statistics with S (Springer)? Also,
have you reviewed the posting guide,
"http://www.R-project.org/posting-guide.html"? Following the steps
there may help you answer many of your own questions and improve the
quality of the response you get from this list when those steps do not
yield a satisfactory answer.
In particular, have you tried "www.r-project.org" -> search
-> "R
site search"? You might find the package "dse" particularly
useful,
especially regarding state space modeling and more recent time series
techniques.
hope this helps. spencer graves
s viswanath wrote:
>hello R experts,
>
>my question is regarding arma modelling and specification.
>
>in another older, statistics package , after determining stationarity, i
would try to work out the number of ar and ma lags using an lm test.
>to do this i would
>
>1. regress my dependant variable on an intercept term then
>2. use LM test for serial correlation, and finally
>3. use the p value of the ols residuals to get the maximum lags for the arma
specification.
>
>I am interested to know how to do this LM test in R say using a function,
using perhaps the fseries package?
>
>Thank you in advance,
>
>Sri
>
>______________________________________________
>R-help at stat.math.ethz.ch mailing list
>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide!
http://www.R-project.org/posting-guide.html
>
>