## One way to fit the logistic function without conditional
## linearity as in ?nls
library(gnm)
set.seed(1)
DNase1 <- subset(DNase, Run == 1)
test <- gnm(density ~ -1 +
Mult(1, Inv(Const(1) + Exp(Mult(1 + offset(-log(conc)),
Inv(1))))),
start = c(NA, 0, 1), data = DNase1, trace = TRUE)
coef(test)
Run the code above in your browser using DataLab