[R-sig-ME] Fitting Non linear mixed effects models

[Ernest William Mauya]

I am learning how to fit non linear mixed effect models in R . I have data which are in clusters i.e 65 clusters each with 8 plots.
I am fitting the model with four predictors like this :


start <- coef(lm(log(y)~I(log(Maxf))+I(log(PF70))+I(log(TF3))+I(log(ps60)),data=data[data$y>0,]))
start[1] <- exp(start[1])
names(start) <- c("a","b1","b2","b3","b4")
model <- nlme(y~a*Maxf^b1*PF70^b2*TF3^b3*ps60^b4,data=data,start=start)
SSasymp<-start

win.m.1 <- nlme(y~a*Maxf^b1*PF70^b2*TF3^b3*ps60^b4, fixed = a + b1 + b2+b3+b4 ~ 1,
                random = a + b1 + b2+b3+b4~ 1 | clustno, data = data, start = start).


But I get errors any assistance from your side?