anova.glmm.admb: Model Comparison Table

In bbolker/glmmadmb: Generalized Linear Mixed Models using 'AD Model Builder'

Description

Compute a model comparison (likelihood ratio test) table for a series of fitted models.

Usage

## S3 method for class 'glmmadmb'
anova(object, ...)

Arguments

object	a fitted glmmadmb model.
...	More complex fitted models. These may differ in fixed or random effects, but note that the likelihood ratio test is generally conservative for testing among models with differing random terms, because of boundary effects.

Details

Basic table for testing the fixed effects, using likelihood ratio tests (may be extended later).

Value

An object of class "anova", inheriting from class "data.frame".

Note

The models must be nested and arranged in order of increasing complexity.

See Also

glmmadmb, glmmADMB-package.

Examples

  data(epil2)
  epil2$subject <- factor(epil2$subject)
## Not run: 
  ## currently (June 2013) having trouble with Windows checks on r-forge, CRAN
  f1 <- glmmadmb(y~Base*trt+Age+Visit+(Visit|subject),
                data=epil2, family="nbinom")
  f2 <- glmmadmb(y~Base*trt+Age+(Visit|subject),
                data=epil2, family="nbinom")
  anova(f2,f1)

## End(Not run)

bbolker/glmmadmb documentation built on May 11, 2019, 9:29 p.m.