getME function crashes

everything's fine for me with the development version of lme4 (github)

library(arm)

Le chargement a n?cessit? le package : MASS
Le chargement a n?cessit? le package : Matrix
Le chargement a n?cessit? le package : lme4
Le chargement a n?cessit? le package : lattice

arm (Version 1.6-09, built: 2013-9-23)

Working directory is D:/Analyses

library(lme4)
set.seed (1)
J <- 15
n <- J*(J+1)/2
group <- rep (1:J, 1:J)
mu.a <- 5
sigma.a <- 2
a <- rnorm (J, mu.a, sigma.a)
b <- -3
x <- rnorm (n, 2, 1)
y <- rbinom (n, 1, invlogit (a[group] + b*x))
mod <- glmer (y ~ x + (1 | group), family=binomial(link="logit"))
getME(mod, "L")

'MatrixFactorization' of Formal class 'dCHMsimpl' [package "Matrix"]
with 10 slots
  ..@ x       : num [1:15] 1.22 1.22 1.1 1.44 1.74 ...
  ..@ p       : int [1:16] 0 1 2 3 4 5 6 7 8 9 ...
  ..@ i       : int [1:15] 0 1 2 3 4 5 6 7 8 9 ...
  ..@ nz      : int [1:15] 1 1 1 1 1 1 1 1 1 1 ...
  ..@ nxt     : int [1:17] 1 2 3 4 5 6 7 8 9 10 ...
  ..@ prv     : int [1:17] 16 0 1 2 3 4 5 6 7 8 ...
  ..@ colcount: int [1:15] 1 1 1 1 1 1 1 1 1 1 ...
  ..@ perm    : int [1:15] 0 1 2 3 4 5 6 7 8 9 ...
  ..@ type    : int [1:4] 2 1 0 1
  ..@ Dim     : int [1:2] 15 15

sessionInfo()

R version 3.0.2 (2013-09-25)
Platform: i386-w64-mingw32/i386 (32-bit)

locale:
[1] LC_COLLATE=French_France.1252  LC_CTYPE=French_France.1252
[3] LC_MONETARY=French_France.1252 LC_NUMERIC=C
[5] LC_TIME=French_France.1252

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] arm_1.6-09      lme4_1.1-1      lattice_0.20-25 Matrix_1.1-0
MASS_7.3-29

loaded via a namespace (and not attached):
[1] abind_1.4-1    coda_0.16-1    fortunes_1.5-1 grid_3.0.2
minqa_1.2.1    nlme_3.1-111
[7] splines_3.0.2  tools_3.0.2


Idem with the 64-bit version of R

Renaud




Le 07/11/2013 17:31, Johannes Radinger a ?crit :

Hi,

Actually I tried to use the sim function from the package "arm" to get
posterior simulations of beta from a merMod.
But somehow this function crashed and I could trace the problem back
to the
function getME(model,"L") which is
the cause for the crash.
For example here an example dataset for a binomial model (glmer) from the
sim function in the 'arm'-package:

set.seed (1)
J<- 15
n<- J*(J+1)/2
group<- rep (1:J, 1:J)
mu.a<- 5
sigma.a<- 2
a<- rnorm (J, mu.a, sigma.a)
b<- -3
x<- rnorm (n, 2, 1)
y<- rbinom (n, 1, invlogit (a[group] + b*x))

mod<- glmer (y ~ x + (1 | group), family=binomial(link="logit"))

When I now run getME(mod, "L") my R session completely crashes with
following error:
*** caught segfault ***
address (nil), cause 'memory not mapped'

Traceback:
  1: .Call(merPredDL, ptr())
  2: PR$L()
  3: getME(mod, "L")

Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace

I am running following setup:

R version 3.0.2 (2013-09-25)
Platform: i686-pc-linux-gnu (32-bit)

locale:
  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
  [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
LC_PAPER=en_US.UTF-8       LC_NAME=C
  [9] LC_ADDRESS=C               LC_TELEPHONE=C
LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] arm_1.6-09      lme4_1.0-5      lattice_0.20-24 Matrix_1.1-0
MASS_7.3-29

loaded via a namespace (and not attached):
[1] abind_1.4-0   coda_0.16-1   grid_3.0.2    minqa_1.2.1   nlme_3.1-111
splines_3.0.2 tools_3.0.2

Can anyone reproduce this problem?

/Johannes

    [[alternative HTML version deleted]]