[R] ARMA random effects?

Spencer Graves spencer.graves at pdf.com
Wed Jun 11 04:47:29 CEST 2008


Hi, All: 

      Is there a way to get random effects for ARMA parameters?  
Consider the following example from the 'corARMA' help page: 

fm1Ovar.lme <- lme(follicles ~ sin(2*pi*Time) + cos(2*pi*Time),
                        data = Ovary, random = pdDiag(~sin(2*pi*Time)))
fm5Ovar.lme <- update(fm1Ovar.lme,
                     corr = corARMA(p = 1, q = 1))
fm5Ovar.lme

Linear mixed-effects model fit by REML
  Data: Ovary
  Log-restricted-likelihood: -771.9471
  Fixed: follicles ~ sin(2 * pi * Time) + cos(2 * pi * Time)
       (Intercept) sin(2 * pi * Time) cos(2 * pi * Time)
        12.1248728         -2.9198264         -0.8487095

Random effects:
 Formula: ~sin(2 * pi * Time) | Mare
 Structure: Diagonal
        (Intercept) sin(2 * pi * Time) Residual
StdDev:    2.614435           1.004898 3.733423

Correlation Structure: ARMA(1,1)
 Formula: ~1 | Mare
 Parameter estimate(s):
      Phi1     Theta1
 0.7868896 -0.2793591
Number of Observations: 308
Number of Groups: 11

 > coef(fm5Ovar.lme)
   (Intercept) sin(2 * pi * Time) cos(2 * pi * Time)
4     8.681454          -2.719864         -0.8487095
<snip>

      Evidently, this model estimates only one set of ARMA parameters.  
Is there a way I can get a different set for each Mare in this example? 

      Thanks,
      Spencer Graves



More information about the R-help mailing list