[R] lme4/Matrix: Call to .Call("mer_update_y"...) and LMEoptimize gives unexpected side effect...

Søren Højsgaard Soren.Hojsgaard at agrsci.dk
Thu Mar 16 22:43:53 CET 2006


Dear all
I want to compute Monte Carlo p-values in lmer-models based on sampled data sets. To speed up calculations, I've tried to use internal functions from the Matrix package (as suggested ealier on the list by Doug Bates). 

So I did:

 fm2 <- lmer(resistance ~ ET + position + (1|Grp), Semiconductor,method='ML')
 simdata<-simulate(fm2,nsim=1)
 ynew <- simdata[,1]

 mer <- fm2
 .Call("mer_update_y", mer, ynew, PACKAGE = "Matrix") 
 mer1u <- LMEoptimize(mer, lmerControl(mer))

What puzzles me is that this call alters my original model fm2 as some kind of side effect. In fact, after the call fm2 is the same as mer1u. Is this side effect intentional and is it possible to avoid?

A detail is that "LMEoptmize" and "LMEoptimize<-" are not exported from the namespace in Matrix, so I simply copied the LMEoptimize function and made it an ordinary function as shown below.

Thanks in advance
Søren



LMEoptimize <- function(x, value)
             {
                 if (value$msMaxIter < 1) return(x)
                 nc <- x at nc
                 constr <- unlist(lapply(nc, function(k) 1:((k*(k+1))/2) <= k))
                 fn <- function(pars)
                     deviance(.Call("mer_coefGets", x, pars, 2, PACKAGE = "Matrix"))
                 gr <- if (value$gradient)
                     function(pars) {
                         if (!isTRUE(all.equal(pars,
                                               .Call("mer_coef", x,
                                                     2, PACKAGE = "Matrix"))))
                             .Call("mer_coefGets", x, pars, 2, PACKAGE = "Matrix")
                         .Call("mer_gradient", x, 2, PACKAGE = "Matrix")
                     }
                 else NULL
		 optimRes <- nlminb(.Call("mer_coef", x, 2, PACKAGE = "Matrix"),
                                    fn, gr,
                                    lower = ifelse(constr, 5e-10, -Inf),
                                    control = list(iter.max = value$msMaxIter,
                                    trace = as.integer(value$msVerbose)))
                 .Call("mer_coefGets", x, optimRes$par, 2, PACKAGE = "Matrix")
                 if (optimRes$convergence != 0) {
                     warning(paste("nlminb returned message",
                                   optimRes$message,"\n"))
                 }
                 return(x)
             }

lmerControl <-
  function(maxIter = 200, # used in ../src/lmer.c only
           tolerance = sqrt(.Machine$double.eps),# ditto
           msMaxIter = 200,
           ## msTol = sqrt(.Machine$double.eps),
           ## FIXME:  should be able to pass tolerances to nlminb()
           msVerbose = getOption("verbose"),
           niterEM = 15,
           EMverbose = getOption("verbose"),
           PQLmaxIt = 30,# FIXME: unused; PQL currently uses 'maxIter' instead
           gradient = TRUE,
           Hessian = FALSE # unused _FIXME_
           )
{
    list(maxIter = as.integer(maxIter),
         tolerance = as.double(tolerance),
         msMaxIter = as.integer(msMaxIter),
         ## msTol = as.double(msTol),
         msVerbose = as.integer(msVerbose),# "integer" on purpose
         niterEM = as.integer(niterEM),
         EMverbose = as.logical(EMverbose),
         PQLmaxIt = as.integer(PQLmaxIt),
         gradient = as.logical(gradient),
         Hessian = as.logical(Hessian))
}




More information about the R-help mailing list