A horrendous hack is employed in order to do arbitrary likelihood ratio tests: the model matrix is built, the names possibly mangled, then fed in as a symbolic formula to glmer/lmer. This is necessary because there is no (easy) way to specify an arbitrary fixed-effect model matrix in glmer.
# S4 method for LMERlike update(object, formula., design, ...) # S4 method for LMERlike vcov(object, which, ...) # S4 method for LMERlike coef(object, which, singular = TRUE, ...) # S4 method for LMERlike logLik(object)
LMERlike
formula
data.frame
vcov
, ignored. In the case of update
, passed to model.matrix
.character
, one of 'C', 'D'.logical
. Should NA coefficients be returned?see the section "Methods (by generic)"
update
: update the formula or design matrix
vcov
: return the variance/covariance of component which
coef
: return the coefficients. The horrendous hack is attempted to be undone.
logLik
: return the log-likelihood
pseudoMM
strictConvergence
logical
return results even when the optimizer or *lmer complains about convergenceoptimMsg
character
record warnings from lme. NA_character_
means no warnings.