Growth of sitka spruce trees over three growing seasons.

SUMMARY:

The SpruceGrpd data come from a balanced design. There are 23 measurements on 79 trees, with two dummy time points at 366 and 731 to mark the border between growing seasons. Each time series is the size of a young sitka spruce tree measured as the log(square of the diameter X height). There are 79 trees over a three-year period. The time is days since January 1, 1987. The first 54 trees were grown in an atmosphere with introduced ozone at 70ppb, 27 trees in each of two growing chambers. The remaining 25 trees are controls, 12 and 13 trees in each of two chambers.

DATA DESCRIPTION:

ARGUMENTS:

y
is the log(square of the diameter X height) at each time point for one tree.
Time
days since January 1, 1987.
Subject
an identifying number for each tree.
group
a factor, treatment or control.

SOURCE:

EXAMPLES:

## Make between seasons independent. 
data<-recordDesign(cluster="Subject",data=SpruceGrpd,record="Time")
blk<-data.frame(layer.id=c(1,1,1), begin.row = c(1, 12,18), 
  begin.col = c(2,13,19), end.row = c(9,15,24), end.col = c(10,16,25))
corDes<-corDesign(size=25,type.layer=list(type="contAR",x.layer="Time"),
   block=blk,data=data[data$clusterID==1,])
gee(y~Time + group, cluster = cbind(clusterID,recordID), family="gaussian", 
  link="identity", correlation=corDes, variance="glm.scale", data=data)