r.squared.glmm: Marginal and conditional r-squared for glmm given fixed and...
In Sz-Tim/sevcheck: It's pronounced 'sev-check'

r.squared.glmm

R Documentation

Marginal and conditional r-squared for glmm given fixed and random variances

Description

This function is based on Nakagawa and Schielzeth (2013). It returns the marginal and conditional r-squared, as well as the AIC for each glmm. Users should call the higher-level generic "r.squared", or implement a method for the corresponding class to get varF, varRand and the family from the specific object

Usage

## S3 method for class 'glmm'
r.squared(
  varF,
  varRand,
  varResid = NULL,
  varDisp = NULL,
  family,
  link,
  mdl.aic,
  mdl.class,
  null.fixef = NULL
)

Arguments

`varF`	Variance of fixed effects
`varRand`	Variance of random effects
`varResid`	Residual variance. Only necessary for "gaussian" family
`family`	family of the glmm (currently works with gaussian, binomial and poisson)
`link`	model link function. Working links are: gaussian: "identity" (default); binomial: "logit" (default), "probit"; poisson: "log" (default), "sqrt"
`mdl.aic`	The model's AIC
`mdl.class`	The name of the model's class
`null.fixef`	Numeric vector containing the fixed effects of the null model. Only necessary for "poisson" family