R help - May 2014 - Fitting multiple nonlinear regression lines with nlrob()

Greetings R community,


I would like to regress a nonlinear trend onto several subsets of data
(representing different treatments) within a dataset.  In my case, I would like
to fit a nonlinear trend to several different "Tillage" treatments.


model.global = nlrob(FinalBiomass ~
Wm[Tillage]*(1-exp((DVS/a[Tillage])^b[Tillage])),
                                      data=DATA,
                 start = list(Wm=c(350,350,350),
a=c(1.4,1.4,1.4),b=c(2.3,2.3,2.3) ),
                                        trace=TRUE)?


I can do this no problem with nls(), however, when I try to do this with nlrob()
I get the error message:


Error in nls(formula, data = data, start = start, algorithm = algorithm,  :
  parameters without starting value in 'data': Wm, a, b

?However, I thought I had defined the starting values for my parameters in my
code (I have 3 "Tillage" treatments I am trying to fit).

I do get an output, but it is identical to the output I get when I run nls(). Is
nlrob capable of performing this analysis??

Any thoughts would be appreciated.
Best,

John



John Laborde
Graduate Research Assistant
University of Nebraska-Lincoln
Department of Agronomy & Horticulture
jlaborde@huskers.unl.edu


	[[alternative HTML version deleted]]