Hi,
I have similar memory problems. I am using 5 multiply imputed data sets.
I have tried both fixes suggested in the How-to link you provided. I do not
have access to a Unix system. What else can people with large or multiple
data sets do on a Windows system?
Thanks,
Sean Richey
-----Original Message-----
From: owner-zelig(a)latte.harvard.edu
[mailto:owner-zelig@latte.harvard.edu]On Behalf Of Kosuke Imai
Sent: Thursday, July 22, 2004 10:54 PM
To: David Mermelstein
Cc: zelig(a)latte.harvard.edu
Subject: Re: [zelig] Memory problems using Zelig-Bprobit
Hi David,
Have you tried everything on the following page?
http://gking.harvard.edu/zelig/docs/How_do_I2.html
If you have an access to Unix server, then it would be best if you could
run your model on that server rather than Windows.
Kosuke
---------------------------------------------------------
Kosuke Imai Office: Corwin Hall 041
Assistant Professor Phone: 609-258-6601
Department of Politics eFax: 973-556-1929
Princeton University Email: kimai(a)Princeton.Edu
Princeton, NJ 08544-1012 http://www.princeton.edu/~kimai
---------------------------------------------------------
On Thu, 22 Jul 2004, David Mermelstein wrote:
> Hi all,
>
> I'm trying to find some help with some memory problems.
> I've been reading the FAQ archive on this topic and I could see that I'm
not
> the first one with this issue. The solutions included in previous posts
did
> not work with my model and dataset.
>
> I am trying to estimate a bivariate probit model. The fact is that my
model
> includes 27 variables (considering the explanatory ones included in the
two
> equations) and about 60.000 records (but I also tryied with a sub-sample
of
> about 20.000).
>
> The mesagge i obtain is the following:
>
> Error: cannot allocate vector of size 287850 Kb
> In addition: Warning message:
> Reached total allocation of 904Mb: see help(memory.size)
>
> My PC has 512 MB RAM and is running with a P-IV and and W2K(Proffesional)
>
> Does anyone know what could be the upper limit that i would be able to
> manage in terms of dataset size?
>
> And, more important, with my model and dataset, and having tryied the
> suggested alternatives in the FAQ, does anyone have another suggestion on
> this?
>
> Many thanks in advance.
>
> David Mermelstein
>
> _________________________________________________________________
> STOP MORE SPAM with the new MSN 8 and get 2 months FREE*
> http://join.msn.com/?page=features/junkmail
>
>
> -
> Zelig Mailing List, served by Harvard-MIT Data Center
> Send messages: zelig(a)latte.harvard.edu
> [un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
> Zelig program information: http://gking.harvard.edu/zelig/
>
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)latte.harvard.edu
[un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)latte.harvard.edu
[un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
Hi all,
I'm trying to find some help with some memory problems.
I've been reading the FAQ archive on this topic and I could see that I'm not
the first one with this issue. The solutions included in previous posts did
not work with my model and dataset.
I am trying to estimate a bivariate probit model. The fact is that my model
includes 27 variables (considering the explanatory ones included in the two
equations) and about 60.000 records (but I also tryied with a sub-sample of
about 20.000).
The mesagge i obtain is the following:
Error: cannot allocate vector of size 287850 Kb
In addition: Warning message:
Reached total allocation of 904Mb: see help(memory.size)
My PC has 512 MB RAM and is running with a P-IV and and W2K(Proffesional)
Does anyone know what could be the upper limit that i would be able to
manage in terms of dataset size?
And, more important, with my model and dataset, and having tryied the
suggested alternatives in the FAQ, does anyone have another suggestion on
this?
Many thanks in advance.
David Mermelstein
_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE*
http://join.msn.com/?page=features/junkmail
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)latte.harvard.edu
[un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
Dear Sean,
There are several ways to deal with memory issues. Please refer to:
http://gking.harvard.edu/zelig/docs/How_do_I2.html
The best way to deal with memory limits, however, is to use linux R rather
than Windows R. It looks like your system has only 64MB of memory, so you
might want to up that to soemthing reasonable -- like 256MB or higher.
Yours,
Olivia.
p.s. -- traceback is used as traceback() -- It's a function.
On Wed, 21 Jul 2004, Richey, Sean wrote:
> Hi,
>
> I get a memory limit error when trying to run multiple data. How can I
> increase R's memory, or what else can I do? Below is the z.out line, error
> message, and traceback.
>
> Thank you,
>
> Sean Richey
>
>
> z.out <- zelig(V2 ~ V1 + V5 + V6 + V9 + V11 + V12 + V13 + V15 + V16 + V17,
> model = "logit", data = list(data1, data2, data3, data4, data5))
> Error: cannot allocate vector of size 19164 Kb
> In addition: Warning message:
> Reached total allocation of 63Mb: see help(memory.size)
> > traceback
> function ()
> {
> if (exists(".Traceback", env = .GlobalEnv))
> .Traceback <- get(".Traceback", env = .GlobalEnv)
> else .Traceback <- NULL
> if (is.null(.Traceback) || length(.Traceback) == 0)
> cat("No traceback available\n")
> else {
> n <- length(.Traceback)
> for (i in 1:n) {
> label <- paste(n - i + 1, ": ", sep = "")
> if ((m <- length(.Traceback[[i]])) > 1)
> label <- c(label, rep(substr(" ", 1,
> nchar(label)), m - 1))
> cat(paste(label, .Traceback[[i]], sep = ""), sep = "\n")
> }
> }
> invisible()
> }
> <environment: namespace:base>
>
-
Zelig Mailing List, served by Harvard-MIT Data Center
Send messages: zelig(a)latte.harvard.edu
[un]subscribe Options: http://lists.hmdc.harvard.edu/?info=zelig
Zelig program information: http://gking.harvard.edu/zelig/
Dear Asghar,
Please look at:
http://gking.harvard.edu/zelig/docs/Special_Cases_When.html
I suggest that you:
library(foreign)
tmp <- read.spss(...) # As you have been doing.
write.table(file = "tmp.tab", tmp)
data <- read.table("tmp.tab", na.strings = "9", header = TRUE)
unlink(tmp.tab) # Cleaning up.
I would suggest emailing the r-help list if you have further questions on loading data. You might suggest that they add the na.strings option directly into read.spss().
Yours,
Olivia
Yours,
Olivia
----- Original Message -----
From: Asghar Minaei
To: Olivia Lau
Sent: Saturday, July 07, 2001 11:35 PM
Subject: question
Dear Lau
Hello
My thank to you for your kind answer to my email. I have a problem with R. I have a SPSS dat file. I want to analysis my data with R. I have chosen 9 as missing value. Could you tell me how I can to introduce 9 as missing value to R, please?
Best regards
Asghar Minaei
