R help - Nov 2012 - Error in memory.size(size) when calling R in batch mode (but not in interactive mode)

Dear all

I use a 64 bit Windows7 system with 16 GB memory.
I have run an R script in batch mode with the following command:
      R CMD BATCH  masterassignment_2012_11_09.r --save
The execution was terminated with the following error message:

        Error in memory.size(size) :  don't be silly!: your machine has a
4Gb address limit


I never get this error message in an interactive session (and I don't think
I should, taking into account the system). To be sure, I have just double
checked this again in an interactive session, and it works fine.

There are two possible explanations I can think of:
a) The command line doesn't "know" the system, i.e. it
"thinks" it works on a 32bit system (is this possible at all?)
b) I run R on a remote desktop, and the command line thinks I am calling R from
my laptop

Anyway, I wouldn't know how to solve this.

In case you wonder why I run R in batch mode at all: I have a series of
iterations in my model, each of which is taking huge amount of memory. Although
I regularly impose a garbage collection, after two or three iterations, R
crashes when used interactively. I thought that running and exiting R after each
iteration would be the only effective way to "clean up" memory. If you
have any other solutions, I would be happy to hear about them as well.



Laurent Franckx, PhD
VITO NV
Boeretang 200, 2400 MOL, Belgium
Tel. + 32 14 33 58 22
Skype: laurent.franckx
laurent.franckx at vito.be
Visit our website: www.vito.be/english and http://www.vito.be/transport






________________________________
 [http://www.vito.be/e-maildisclaimer/bsds.png]  <http://bsds.vito.be>

VITO Disclaimer: http://www.vito.be/e-maildisclaimer

Le vendredi 09 novembre 2012 ? 10:43 +0000, Franckx Laurent a ?crit
:
> Dear all
> 
> I use a 64 bit Windows7 system with 16 GB memory.
> I have run an R script in batch mode with the following command:
>       R CMD BATCH  masterassignment_2012_11_09.r --save
> The execution was terminated with the following error message:
> 
>         Error in memory.size(size) :  don't be silly!: your machine
> has a 4Gb address limit
> 
> 
> I never get this error message in an interactive session (and I don't
> think I should, taking into account the system). To be sure, I have
> just double checked this again in an interactive session, and it works
> fine.
> 
> There are two possible explanations I can think of:
> a) The command line doesn't "know" the system, i.e. it
"thinks" it
> works on a 32bit system (is this possible at all?)
> b) I run R on a remote desktop, and the command line thinks I am
> calling R from my laptop
c) When you run "R CMD BATCH", the 32-bit version of R is used, while
when you run it interactively, you start the 64-bit version.

To check this, add sessionInfo() to your batch script, and have a look
at the output.


Hope this helps
> Anyway, I wouldn't know how to solve this.
> 
> In case you wonder why I run R in batch mode at all: I have a series
> of iterations in my model, each of which is taking huge amount of
> memory. Although I regularly impose a garbage collection, after two or
> three iterations, R crashes when used interactively. I thought that
> running and exiting R after each iteration would be the only effective
> way to "clean up" memory. If you have any other solutions, I
would be
> happy to hear about them as well.
> 
> 
> 
> Laurent Franckx, PhD
> VITO NV
> Boeretang 200, 2400 MOL, Belgium
> Tel. + 32 14 33 58 22
> Skype: laurent.franckx
> laurent.franckx at vito.be
> Visit our website: www.vito.be/english and http://www.vito.be/transport
> 
> 
> 
> 
> 
> 
> ________________________________
>  [http://www.vito.be/e-maildisclaimer/bsds.png] 
<http://bsds.vito.be>
> 
> VITO Disclaimer: http://www.vito.be/e-maildisclaimer
> 
> ______________________________________________
> R-help at r-project.org 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.

Dear all,

I use R version 2.15.1 on  x86_64-pc-linux-gnu (64-bit).

I run a simulation where I have to call a series of R scripts in batch mode.

When I run the following two lines, this works well for PERIOD='17' but
not for PERIOD='08':

R CMD BATCH  --no-save "--args PERIOD='08' YEAR='203O'
SC='BAU' " preparecostvectors.R preparecostvectors08_2030.Rout
R CMD BATCH  --no-save "--args PERIOD='17' YEAR='2030'
SC='BAU' " preparecostvectors.R preparecostvectors17_2030.Rout

The error message I get for PERIOD='08' is (but I do not think this is
relevant for my main question):

Error in names(votlist) <- votname :
  'names' attribute [3] must be the same length as the vector [0]

I get this error message for all values of YEAR.

Oddly enough, when I run the script preparecostvectors.R interactively, all
works perfectly fine. However, this is not a structural solution, because I need
to be able to run my simulation for a large number of years.

I was wondering if any tools exist that could help me to reconstruct what went
wrong in batch mode. I think I am familiar with most debugging tools in
interactive mode, but how can I trace the origin of the problem when I get an
error in batch mode?

Just to avoid any misunderstanding: I am not asking any help on the error
message itself, just some suggestions on how I could trace what goes wrong when
running R in batch mode.


Laurent Franckx, PhD
VITO NV
Boeretang 200, 2400 MOL, Belgium
Tel. + 32 14 33 58 22
Skype: laurent.franckx
laurent.franckx at vito.be
Visit our website: www.vito.be/english and http://www.vito.be/transport






VITO Disclaimer: http://www.vito.be/e-maildisclaimer