Hi Folks,
The following gives me the exact plot I want using trellis.
xyplot(V56 ~ V12 | subset(frequency, V17 < 12000) , data=kdata,
ylab="Fine Structure Depth [dB]",
xlab="QSIN score",
panel=function(x,y,...) {
panel.xyplot(x[HL=="N"],y[HL=="N"],
pch=16,col="green")
panel.xyplot(x[HL=="SN"],y[HL=="SN"],pch=16,
col="red")
panel.lmline(x[HL=="N"],y[HL=="N"],
col="green")
panel.lmline(x[HL=="SN"],y[HL=="SN"],
col="red")
}
)
Now I want to add to each panel by writing the result of
cor(V56[HL=="N"],V12[HL=="N"], use="complete")
cor(V56[HL=="SN"],V12[HL=="SN"], use="complete")
in the specific colors.
I have tried things such as:
panel.text(4,20,round(cor(V56[HL=="N"],V12[HL=="N"],
use="pairwise.complete"),digits=3),col="green4")
But this does not seem to compute correlation based on the grouping
variable in each panel.
I am not sure that I have given you adequate information. Any help
will be much appreciated.
Regards,
Sumit
[[alternative HTML version deleted]]