R help - May 2009 - Cgee: error: logistic model

Hello,



Much like Charlie Wills a year ago, I am trying to run the APE program  
COMPAR.GEE with a model containing a categorical response variable and  
a continuous variable.



My command code is :



compar.gee(costusdata$Syndrome ~ costusdata$Stamen, phy = costustree,  
family = "binomial")



I receive the following output with an error:

Beginning Cgee S-function, @(#) geeformula.q 4.13 98/01/27

running glm to get initial regression estimate

           (Intercept)   costusdata$Stamen18 costusdata$Stamen20.5  
costusdata$Stamen22.5 costusdata$Stamen25.5 costusdata$Stamen27.5    
costusdata$Stamen28

          2.056607e+01         -3.750463e-08          
-5.619142e-08         -5.644906e-08         -3.849581e-08          
-5.622864e-08         -3.777595e-08

   costusdata$Stamen30   costusdata$Stamen31   costusdata$Stamen32  
costusdata$Stamen32.5   costusdata$Stamen35 costusdata$Stamen36.5    
costusdata$Stamen37

         -3.954765e-08         -3.749803e-08         -4.113214e 
+01         -1.917977e+01         -3.748004e-08         -4.113214e 
+01         -5.282325e-08

costusdata$Stamen37.5 costusdata$Stamen38.5   costusdata$Stamen39  
costusdata$Stamen41.3 costusdata$Stamen42.5   costusdata$Stamen45  
costusdata$Stamen47.5

         -5.584079e-08         -4.113214e+01         -4.113214e 
+01         -5.272208e-08         -2.016060e+01         -4.113214e 
+01         -4.113214e+01

   costusdata$Stamen50 costusdata$Stamen52.5   costusdata$Stamen55     
costusdata$Stamen?

         -4.113214e+01         -4.113214e+01          
-5.309908e-08         -2.264551e+01

Error in gee(costusdata$Syndrome ~ costusdata$Stamen, c(1, 1, 1, 1,  
1,  :

   Cgee: error: logistic model for probability has fitted value very  
close to 1.

estimates diverging; iteration terminated.



Simon Blumberg answered Charlie's issue with "My guess is that this  
combination of variables produces separation in the data: Too many  
(all?) of the response 1's are in at level of VAR3, and the 0's are at  
the other level (or vice versa)."

I only have the two variables.

I'm wondering if anyone has encountered this problem and how you went  
about solving it. I believe that the problem appears when one has to  
low (phylogenetic) replication within each (pollination mode) factor  
level.  Is there a way to run compar.gee if this is in fact the  
problem.  From the data,  I want to know if there a decrease in the  
continuous trait is correlated with a change in the categorical trait.



Any insight is much appreciated.



Thank you,



Jenn

  
	[[alternative HTML version deleted]]