lmerCorr.fit: lmer.fit function for the single correlation template model

In lme4/lme4pureR: lme4 in pure R

Description

lmer.fit function for the single correlation template model

Usage

lmerCorr.fit(y, mmFE, corr, grp, weights, offset = numeric(n), REML = TRUE)

Arguments

y	response vector
mmFE	model matrix for the fixed effects
corr	template correlation matrix for the single scalar random effect
grp	grouping factor for the random effect (levels correspond to dimnames of corr
weights	weights
offset	offset
REML	should restricted maximum likelihood be used?

Examples

library(lme4pureR)
library(lme4)
library(minqa)
library(subscript)
library(rmv)
set.seed(1)
n <- 100
x <- rnorm(n)
X <- cbind(1, x)
grp <- gl(n/5,5)
ugrps <- unique(grp)
q <- length(ugrps)
corr <- rcov(q+1,q)
dimnames(corr) <- rep(list(as.character(ugrps)), 2)
b <- as.numeric(rmv(1,corr) %*% as(grp, "sparseMatrix"))
y <- as.numeric(X%*%rnorm(ncol(X)) + b + rnorm(n))

m <- lmerCorr.fit(y, X, corr, grp)
m$par

lme4/lme4pureR documentation built on May 21, 2019, 7:34 a.m.