Problem with GLMM due to zero inflation

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Problem with GLMM due to zero inflation

vheim
Hey,

I tried to set-up a GLMM looking at how the nr. of boats at a dive site influences the food uptake of observed wildlife.

This did not work since I have a lot of food uptake values that are 0 kg.

My idea was to add a second column to asses if an individual is feeding or not. This column is binomial with 1 = feeding and 0 = non-feeding.

I then tried to run a model on this with the code:

library(nlme)
library(lme4)

###

mR <- glmer(eating ~ boats + (1 | individual), data = df2, family = "binomial")
m0 <- glm(eating ~ boats + provision + month, family = binomial, data = df2)

When I run the code I get the error-message:

> mR <- glmer(eating ~ boats + (1 | individual), data = df2, family = "binomial")
Warning messages:
1: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv,  :
  Model failed to converge with max|grad| = 0.0628276 (tol = 0.001, component 1)
2: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv,  :
  Model is nearly unidentifiable: very large eigenvalue
 - Rescale variables?

I am a beginner to R and stuck with this problem.
I am really appreciative for your help!