R help - May 2009 - R for arma mdel with constraints on parameters

Hi,

 

i am learning R recently and find it very helpful in time series model.

 

In ARMA model,  given (p,q) it can get the estimation of a[i] and b[j] easily
with arima() function.

 

X[t] = a[1]X[t-1] + ... + a[p]X[t-p] + e[t] + b[1]e[t-1] + ... + b[q]e[t-q]
 
but in my recent data model, i met a problem. In the ARMA model, p and q are
fixed, but there are some constraints in the parameters a[i] and b[j], such as
for some i (i<p), a[i]=0. my problem is how to get these parameters'
estimation for these constrants with arima() function? or i need to write
functions to realize it.
 
Best regards,
 
Tongen 
from Beijing
 
_________________________________________________________________


	[[alternative HTML version deleted]]

See the 'fixed' argument, which allows you to fix ARIMA parameters at 
any value, including zero.

On Tue, 26 May 2009, FangTonggen wrote:
> Hi,
>
> i am learning R recently and find it very helpful in time series model.
>
>
>
> In ARMA model, given (p,q) it can get the estimation of a[i] and 
> b[j] easily with arima() function.
>
> X[t] = a[1]X[t-1] + ... + a[p]X[t-p] + e[t] + b[1]e[t-1] + ... + b[q]e[t-q]
>
> but in my recent data model, i met a problem. In the ARMA model, p 
> and q are fixed, but there are some constraints in the parameters 
> a[i] and b[j], such as for some i (i<p), a[i]=0. my problem is how 
> to get these parameters' estimation for these constrants with 
> arima() function? or i need to write functions to realize it.
>
> Best regards,
>
> Tongen
> from Beijing
-- 
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