Ken,
You just need to wrap the rhs of your formula in I() to
get around update()'s parsing of terms.
m2 <- update(m1, . ~
I(Rm * Contrast^ex/(Contrast^fx + sig^fx)),
start = list(Rm = 30, sig = 0.05, ex = 3,
fx = 3.1))
-Peter Ehlers
Ken Knoblauch wrote:
Hi,
I'm getting an error that I don't understand when updating an nls
object. Here is a toy example.
dd <- structure(list(Contrast = c(0.00376, 0.03759, 0.12782, 0.25564,
0.50376, 1), Response = c(0.29915, 6.13248, 29.01709, 30.0641,
29.46581, 27.67094)), .Names = c("Contrast", "Response"), class =
"data.frame", row.names = c(NA,
-6L))
m1 <- nls(Response ~
Rm * Contrast^ex/(Contrast^ex + sig^ex),
data = dd,
start = list(Rm = 30, sig = 0.05, ex = 3))
m2 <- update(m1, . ~
Rm * Contrast^ex/(Contrast^fx + sig^fx),
start = list(Rm = 30, sig = 0.05, ex = 3,
fx = 3.1))
Error in terms.formula(tmp, simplify = TRUE) : invalid power in formula
but there is nothing wrong with fitting the second model directly
with nls
m2a <- nls(Response ~
Rm * Contrast^ex/(Contrast^fx + sig^fx),
data = dd,
start = list(Rm = 30, sig = 0.05, ex = 3,
fx = 3.1))
nor with fitting the model with the data transformed (though it is
not necessarily the way I would like to fit the model in this
case).
m2b <- update(m1, log(.) ~ log(.))
m3 <- update(m2b, . ~
log(Rm * Contrast^ex/(Contrast^fx + sig^fx)),
start = list(Rm = 30, sig = 0.05, ex = 3,
fx = 3.1))
sessionInfo()
R version 2.10.1 Patched (2010-01-25 r51051)
i386-apple-darwin9.8.0
locale:
[1] en_US.UTF-8/en_US.UTF-8/C/C/en_US.UTF-8/en_US.UTF-8
attached base packages:
[1] stats graphics grDevices utils datasets methods
[7] base
Thanks, in advance, for any help.
Ken
--
Peter Ehlers
University of Calgary