cmp_R2 function

Compute R2 with a specified C matrix

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)

Returns

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')