[R] size limit in R?

From: <j-simpson_at_northwestern.edu>
Date: Mon, 21 May 2007 12:12:31 -0500


Hi,

Please see the email exchanges below. I am having trouble generating output that is large enough for our needs, specifically when using the GaussRF function. However, when I wrote Dr. Schlather (the author of the GaussRF function), he indicated that there is also a limit imposed by R itself. Is this something that we can overcome?

Thank you very much for any assistance you may provde. Regards,
Jamesina

==============Original message text=============== From: Martin Schlather <schlath_at_math.uni-goettingen.de> Date: Mon, 21 May 2007 8:50:05 am CDT
Subject: Re: GaussRF function

Dear Jamesina,

> x <- numeric( 200 / 0.025 * 1450 / 0.025) Error: cannot allocate vector of size 3625000 Kb

So, the problem is at least also in R, and I guess that it is even deeper, namely that your system does not support 3GB for one pointer. (My own supports about 1.8 GB, but I never met your problem, since my needs have been more moderat (0.1 GB)).

I suggest to write to the R-help mailing list, mentioning the above problem. If solved then I can check RandomFields .

Cheers,
Martin

j-simpson_at_northwestern.edu wrote:

> Dear Dr. Schlather,
> 
> Thank you for your quick reply.  Please find below the error:
> 

>> library(RandomFields)
>> library(MASS)
>>
>> x <- c(0, 200.0, 0.025)
>> y <- c(0, 1450.0, 0.025)
>>
>> model <- "stable"
>> scale <- 0.1
>> alpha <- 2
>> mean <- 0
>> variance <- 1.0
>> nugget <- 0
>>
>> f <- GaussRF(x=x, y=y, model=model, grid=TRUE, gridtriple=TRUE,
> +               param=c(mean, variance, nugget, scale,alpha))
> Error: cannot allocate vector of size 3625515 Kb
> In addition: Warning messages:
> 1: Reached total allocation of 2560Mb: see help(memory.size) 
> 2: Reached total allocation of 2560Mb: see help(memory.size) 
> 
> Regards,
> Jamesina
> 
> 
> 
> ==============Original message text===============
> On Mon, 21 May 2007 1:03:10 am CDT Martin Schlather wrote:
> 
> 
> Dear Jamesina,
> 
> would you send me the error message
> and where it appears?
> 
> Cheers,
> Martin
> 
> 
> j-simpson_at_northwestern.edu wrote:

>> Dear Dr. Schlather,
>>
>> I am a recent Ph.D. graduate at Northwestern University in Evanston, IL working for a
>> research project with the GaussRF function in the R program. I saw that you are the
>> author of the GaussRF documentation, and am hoping you may be able to be of
> assistance 

>> for one question.
>>
>> We are trying to create an output "f" that is 8000 x 58000 numbers (an ~ 8GB output
> file).  

>> However, it appears that we are limited by either the R program or the GaussRF function
> to 

>> only a matrix creating a 3 GB output. Do you know if there is a way for us to overcome
>> this?
>>
>> Below is the brief code I am using in case it would be of help.
>> Thank you in advance for any assistance you may provide.
>>
>> Best regards,
>> Jamesina
>>
>>
>>
>> library(RandomFields)
>> library(MASS)
>>
>> x <- c(0, 200.0, 0.025)
>> y <- c(0, 1450.0, 0.025)
>>
>> model <- "stable"
>> scale <- 0.1
>> alpha <- 2
>>
>> mean <- 0
>> variance <- 1.0
>> nugget <- 0
>>
>>
>> f <- GaussRF(x=x, y=y, model=model, grid=TRUE, gridtriple=TRUE,
>> param=c(mean, variance, nugget, scale,alpha))
>>
>> write.matrix(f,file='C:\\Documents and Settings\\Jamesina\\My Documents\
>> \gauss_8000x58000cells.dat')
>>
>>
>>
>> .................................
>> Jamesina J. Simpson, Ph.D.
>> Dept. of Electrical Engineering and Computer Science
>> Assistant Master of Shepard Residential College
>> Northwestern University
>> 2145 Sheridan Road, EECS Dept.
>> Evanston, IL 60208
>> 847-332-4750
>>
>>

>
>
-- 
Prof. Dr. Martin Schlather
Institut für Mathematische Stochastik & Zentrum für Statistik
Georg-August-Universität Göttingen
Maschmühlenweg 8-10
D -- 37073 Göttingen

email: schlather_at_math.uni-goettingen.de
http://www.stochastik.math.uni-goettingen.de/research/index.phpphone: +49 (0)551 39 13 509     fax : +49 (0)551 39 13 505
===========End of original message text===========


.................................
Jamesina J. Simpson
Ph.D. candidate, Dept. of Electrical Engineering and Computer Science
Assistant Master of Shepard Residential College
Northwestern University
2145 Sheridan Road, EECS Dept.
Evanston, IL 60208
847-332-4750



______________________________________________ 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.

Received on Mon 21 May 2007 - 17:18:46 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Mon 21 May 2007 - 18:31:37 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help. Please read the posting guide before posting to the list.