cmp_R2: Compute R2 with a specified C matrix

In r2glmm: Computes R Squared for Mixed (Multilevel) Models

Description

Compute R2 with a specified C matrix

Usage

cmp_R2(c, x, SigHat, beta, method, obsperclust = NULL, nclusts = NULL)

Arguments

c	Contrast matrix for fixed effects
x	Fixed effects design matrix
SigHat	estimated model covariance (matrix or scalar)
beta	fixed effects estimates
method	the method for computing r2beta
obsperclust	number of observations per cluster (i.e. subject)
nclusts	number of clusters (i.e. subjects)

Value

A vector with the Wald statistic (ncp), approximate Wald F statistic (F), numerator degrees of freedom (v1), denominator degrees of freedom (v2), and the specified r squared value (Rsq)

Examples

library(nlme)
library(lme4)
library(mgcv)
lmemod = lme(distance ~ age*Sex, random = ~1|Subject, data = Orthodont)
X = model.matrix(lmemod, data = Orthodont)
SigHat = extract.lme.cov(lmemod, data = Orthodont)
beta = fixef(lmemod)
p = length(beta)
obsperclust = as.numeric(table(lmemod$data[,'Subject']))
nclusts = length(obsperclust)
C = cbind(rep(0, p-1),diag(p-1))
partial.c = make.partial.C(p-1,p,2)

cmp_R2(c=C, x=X, SigHat=SigHat, beta=beta, obsperclust = obsperclust,
nclusts = nclusts, method = 'sgv')
cmp_R2(c=partial.c, x=X, SigHat=SigHat, beta=beta, obsperclust = obsperclust,
nclusts = nclusts, method = 'sgv')

r2glmm documentation built on May 1, 2019, 9:09 p.m.