blockLambdat: Create digonal block on transposed relative covariance factor

In lme4/lme4pureR: lme4 in pure R

Description

Each random-effects term is represented by diagonal block on the transposed relative covariance factor. blockLambdat creates such a block, and returns related information along with it.

Usage

blockLambdat(nl, nc)

Arguments

nl	Number of levels in a grouping factor for a particular random effects term (the number of levels in the grp argument in Zsection).
nc	Number of columns in a dense model matrix for a particular random effects term (the number of columns in the mm argument in Zsection).

Value

A list with:

• the block

• ititial values of theta for the block

• lower bounds on these initial theta values

• a function that updates the block given the section of theta for this block

Examples

(l <- blockLambdat(2, 3))
within(l, slot(Lambdat, 'x') <- updateLambdatx(as.numeric(10:12)))

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