From: Jonathan Smith <jon3smith_at_hotmail.com>

Date: Tue 01 Aug 2006 - 00:08:22 EST

R-help@stat.math.ethz.ch 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. Received on Tue Aug 01 00:16:38 2006

Date: Tue 01 Aug 2006 - 00:08:22 EST

I am writing as I am still having trouble trying to define my own covariance
matrix. My code is displayed below. I am defining the covariance matrix in
the form of an AR1 process so it can be easily checked if working correctly.
Another question I have is if it is possible to define the matrix without
giving p a specific value and leaving it in as a coefficient. For the
reason that it can be estimated when model is run. I figure that is the way
AR1 works in GLS.

Thank you

Jon Smith

I would appreciate any help s I am stuck here and need to figure this out
prior to continuing my research.

function ()

{

library(nlme)

tim<-c(1,2,3,4,1,2,3,4,1,2,3,4,1,2,3,4)

peep<-c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4)
y<-c(11.78,9.53,11.03,9.89,10.80,8.74,10.25,10.69,5.60,7.27,6.81,4.56,7.01,5.64,6.30,8.31)

#This y data was created from and AR1 model with correlation coefficient

equaling 0.7.

timMat<-matrix(c(tim),ncol=1,nrow=16)

peepMat<-matrix(c(peep),ncol=1,nrow=16)

yMat<-matrix(c(y),ncol=1,nrow=16)

dataframe<-data.frame(yMat,timMat,peepMat)
p=0.7

g=1

tester2<-corSymm(value = c(p^(1),p^(2),p^(3),p^(1),p^(2),p^(1)),form = ~
timMat|peepMat)

tester2<-Initialize(tester2, data = dataframe)
testMat2<-corMatrix(tester2)

print(testMat2)

# this appears to be working correctly

#smanGls<-gls(yMat~timMat,data = dataframe, corr = corAR1(form =

~timMat|peepMat))

# works perfectly

smanGls<-gls(yMat~timMat,data = dataframe, corr = corSymm(tester2))

arsum<-summary(smanGls)

print(arsum)

#this is what message I get when I try to use the covarince matrix I

defined.

#Error in gls(yMat ~ timMat, data = dataframe, corr = corSymm(tester2)) :

# false convergence (8)

}

R-help@stat.math.ethz.ch 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. Received on Tue Aug 01 00:16:38 2006

Archive maintained by Robert King, hosted by
the discipline of
statistics at the
University of Newcastle,
Australia.

Archive generated by hypermail 2.1.8, at Tue 01 Aug 2006 - 02:16:53 EST.

*
Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help.
Please read the posting
guide before posting to the list.
*