[R] lmer2 error under Mac OS X on PowerPC G5 but not on Dual-Core Intel Xeon

[Prof Brian Ripley]

On Mon, 29 Jan 2007, Michael Kubovy wrote:

> On Jan 28, 2007, at 9:39 PM, Benilton Carvalho wrote:
>
>> This seems to be due to the fact that you didn't have enough memory
>> when running lmer2.
>>
>> I might be wrong, but I think Calloc tries to get contiguous
>> memory, so this might the problem.
>>
>> If you are positive that you have enough memory, a gc() might help.

Generally, it almost never helps.  R will itself do a gc() before giving 
up on memory allocation, and R is not allocating memory with Calloc, the 
OS services are being used.  Nothing is going to help allocating 3.6Gb on 
a 32-bit OS.  (One case where it might help is doing a large allocation 
with Calloc after freeing a large R object, on OSes where gc() will 
actually give memory back to the pool Calloc uses.)

What makes no sense is that it says using as.double gave an error message 
from Calloc.  No known as.double method calls Calloc.  This suggests that 
this is a symptom of memory corruption.

>
> I have 2 GB memory on this machine. Should be enough, no?
>
> > gc()
>           used (Mb) gc trigger (Mb) max used (Mb)
> Ncells 1008175 27.0    1476915 39.5  1368491 36.6
> Vcells  540055  4.2    1031040  7.9  1031026  7.9
> > (fm1 <- lmer2(Reaction ~ Days + (Days|Subject), sleepstudy))
> Error in as.double(start) : Calloc could not allocate (903190944 of
> 4) memory
>
>
>> On Jan 28, 2007, at 8:35 PM, Michael Kubovy wrote:
>>
>>>> (fm1 <- lmer2(Reaction ~ Days + (Days|Subject), sleepstudy))
>>> Error in as.double(start) : Calloc could not allocate (888475968 of
>>> 4) memory
>>> *************************
>>>> sessionInfo()
>>> R version 2.4.1 (2006-12-18)
>>> powerpc-apple-darwin8.8.0
>>>
>>> locale:
>>> C
>>>
>>> attached base packages:
>>> [1] "grid"      "datasets"  "stats"     "graphics"  "grDevices"
>>> "utils"     "methods"
>>> [8] "base"
>>>
>>> other attached packages:
>>>          lme4       Matrix       xtable latticeExtra      lattice
>>> gridBase         MASS
>>> "0.9975-11"   "0.9975-8"      "1.4-3"      "0.1-4"    "0.14-16"
>>> "0.4-3"     "7.2-31"
>>>           JGR       iplots       JavaGD        rJava
>>>      "1.4-15"      "1.0-5"      "0.3-5"     "0.4-13"
>>> *************************
>>> lmer runs the example w/o a problem
>>>
>>> I just tried to run it on on Intel-based MacPro, and lmer2 ran
>>> without a hitch.
> _____________________________
> Professor Michael Kubovy
> University of Virginia
> Department of Psychology
> USPS:     P.O.Box 400400    Charlottesville, VA 22904-4400
> Parcels:    Room 102        Gilmer Hall
>         McCormick Road    Charlottesville, VA 22903
> Office:    B011    +1-434-982-4729
> Lab:        B019    +1-434-982-4751
> Fax:        +1-434-982-4766
> WWW:    http://www.people.virginia.edu/~mk9y/
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595