I am running the cosinor mixed effect, assessing the interaction between compound (2) and agonist (2)
du.model<-create.cosinor.param(time="Hour_of_Day", period=24,data=db.data)
f2<-fit.cosinor.mixed(y="value",
interaction=c("compound","agonist"),
random="1+rrr+sss|sample_id",
data=du.model)
I have 24 samples each in triplicate. so In total I have 72 observations. However, I keep getting an error like this:
Error: number of observations (=72) <= number of random effects (=72) for term (1 + rrr + sss | sample_id); the random-effects parameters and the residual variance (or scale parameter) are probably unidentifiable