[R-sig-ME] nlme Error message

Andrew Close a.j.close at newcastle.ac.uk
Tue Jan 18 14:56:05 CET 2011

Dear All,

I apologise in advance if this appears to be an obvious mistake on my part but I am hoping someone can help.

I am attempting to describe a decrease in bacteria in substrate following application of different types of treatment.

Initially, I created the group object and generated an nlsList model


CP8E=read.csv("CP8E.csv", header=TRUE, row.names = 1)
CP8EG<-groupedData(Number ~ Time | Treatment, data = CP8E)
plot(CP8EG, aspect = 0.5)

mod2.lis = nlsList(Number ~ SSasympOff(Time, Asym ,lrc, c0), data = CP8EG)

Model: Number ~ SSasympOff(Time, Asym, lrc, c0) | Treatment
Data: CP8EG
           Estimate Std. Error    t value  Pr(>|t|)
Pre-Chill  -74371.96   127981.3 -0.5811157 0.5172986
Control   -144109.97   127725.3 -1.1282809 0.4469065
            Estimate Std. Error   t value   Pr(>|t|)
Pre-Chill -0.8484641  0.2978322 -2.848799 0.06216886
Control   -0.8462067  0.1333199 -6.347189 0.03402831
          Estimate Std. Error  t value   Pr(>|t|)
Pre-Chill 7.278602   2.544110 2.860961 0.06168919
Control   7.593089   1.357320 5.594179 0.04318056
Residual standard error: 139242.4 on 4 degrees of freedom

###### used the nlsList model to prompt an nlme model


###### And the following error message was received

Error in nlme.formula(model = Number ~ SSasympOff(Time, Asym, lrc, c0),  :
  Step halving factor reduced below minimum in PNLS step


I am prepared for the fact that I may be using an inappropriate SelfStarter function for the data, but I encounter similar messages for other SS functions that I have tried (such as SSbiexp).

Can  anyone suggest what it is that I am doing wrong.

With Thanks


More information about the R-sig-mixed-models mailing list