[R-sig-ME] Independent curves with common variance parameter
Nicholas Lewin-Koh
nikko at hailmail.net
Fri Aug 10 20:10:17 CEST 2007
Hi,
There is hopefully a simple answer to this that I am not seeing. I would
like to fit the models:
y_ij = f(x_ij,B_i) + g{f(x_ij,B_i), v_i}e_ij (1)
and
y_ij = f(x_ij,B_i) + g{f(x_ij,B_i), v_pooled}e_ij (2)
where v=(sigma,d), and I would like to test for a common variance
parameter.
If I am understanding correctly
than the following code will get me model (2) with a pooled variance
parameter?
tt<-groupedData(response~conc|curve,assay.data)
bb<-nlsList(response~SSllogis(conc,A,B,xmid,scal),tt,start=start[[1]])
bb2<-nlme(bb,weights=varPower(), random=list(A~1,B~1,xmid~1,scal~1))
And then to get model (1) I would have to define a gnlsList structure
and fit each curve. How would I go about fitting a model with
v={sigma_pooled, dij} ? is this possible in the nlme framework?
Any suggestions appreciated, and I can provide the data I am
using if requested, it is published.
Nicholas
More information about the R-sig-mixed-models
mailing list