Dear all,
Probably a very stupid question.
I am trying to reproduce the example of Pinheiro & Bates pp.278--287 on
R 3.0.2 (Windows).
I'd like to run :
###
mod.nls <- nls(conc ~ SSbiexp(time, A1, lrc1, A2, lrc2), data=Indometh)
mod.lis <- nlsList(conc ~ SSbiexp(time, A1, lrc1, A2, lrc2), data=Indometh)
mod1 <- nlme(model=mod.lis, data=Indometh,
random = pdDiag(A1 + lrc1 + A2 + lrc2 ~ 1))
>mod1
Nonlinear mixed-effects model fit by maximum likelihood
Model: conc ~ SSbiexp(time, A1, lrc1, A2, lrc2)
Data: Indometh
Log-likelihood: 54.59671
Fixed: list(A1 ~ 1, lrc1 ~ 1, A2 ~ 1, lrc2 ~ 1)
A1 lrc1 A2 lrc2
2.8275372 0.7736221 0.4614716 -1.3441022
Random effects:
Formula: list(A1 ~ 1, lrc1 ~ 1, A2 ~ 1, lrc2 ~ 1)
Level: Subject
Structure: Diagonal
A1 lrc1 A2 lrc2 Residual
StdDev: 0.5714106 0.1580778 0.1115978 7.3195e-06 0.08149341
Number of Observations: 66
Number of Groups: 6
###
which works well. But I'd like to directly specify the model with nlme,
not going through "nls" or "nlsList" (please don't ask
why...). Below
the result. I don't get what I am missing?
###
>mod2 <- nlme(model=conc ~ SSbiexp(time, A1, lrc1, A2, lrc2),
data=Indometh,
+fixed = A1+lrc1+A2+lrc2~1,
+random = pdDiag(A1 + lrc1 + A2 + lrc2 ~ 1))
Error in nlme.formula(model = conc ~ SSbiexp(time, A1, lrc1, A2, lrc2), :
subscript out of bounds
###
[[alternative HTML version deleted]]