Skip to content
Prev 13264 / 20628 Next

Results from GLMM, error bars for predictions

Hello,

Thank you for writing,

I tried what you said, but simple addition of "offset(log(0.25))" did 
not work. Hence I created a factor "area" expressed as m? with the value 
0.25 for all observations:
  data$area <- rep(0.25, 180)
and added offset(log(area)) to the model formula.

Ok, plotting the predicted values for all relevant combinations of the 
fixed effects was also my intention. However, the problem with the 
errorbars occured. Or is it a faulty assumption of mine, that plots of 
predicted values need error-bars?

I'm sorry, i recognized that the answer was only adressed to you, after 
i send the mail. Then i send it again to the Mailing List, i hope it 
won't get to chaotic right now. Now i used "answer all"

kind regards



Am 27.04.2015 um 10:50 schrieb Thierry Onkelinx: