[R] tseries(arma) vs. stats(arima)

[Prof Brian Ripley]

arima() certainly does.  You missed the 'fixed' argument that allows ARMA 
coefficients to be set to any value, including 0.

Note that arma() does not do ML estimation.  For that you have to worry 
about invertibility, and it can be much harder to do the optimization with 
constrained parameters.

On Fri, 21 Mar 2008, Richard Saba wrote:

> Hello,
> The "arma"  function in the "tseries"  package allows estimation of models
> with specific "ar" and  "ma" lags  with its "lag" argument.
> For example:  y[t] = a[0] + a[1]y[t-3] +b[1]e[t-2] + e[t]  can be estimated
> with the following specification :   arma(y, lag=list(ar=3,ma=2)).
>
> Is this possible with the "arima" function in the "stats" or in other time
> series packages like fArima, forecast, or FinTS? They all take a "lag"
> argument. I  would like to have the ability to estimate models like the one
> above while utilizing the "xreg" argument available in the other arima
> functions .
> Thanks,
> Richard Saba
> sabaric at auburn.edu
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

-- 
Brian D. Ripley,                  ripley at 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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595