[R] R for Windows crash on new laptop - corrected script

Karen Martien Karen.Martien at noaa.gov
Thu Mar 30 01:56:39 CEST 2006


I realized that the script I included in my original post (see below) 
was written for an older version of rmetasim and will not work with the 
current version.  The only change that need be made to get my script to 
run with the latest release of rmetasim is to change the command 
'simulate.landscape()' to 'sim.landscape()'.  My apologies to anyone who 
tried to use my script and ran into problems.
Cheers, Karen


>> -----Original Message-----
>> From: r-help-bounces at stat.math.ethz.ch [mailto:r-help-
>> bounces at stat.math.ethz.ch] On Behalf Of Karen Martien
>> Sent: Wednesday, March 29, 2006 11:07 AM
>> To: r-help at stat.math.ethz.ch
>> Subject: [R] R for Windows crash on new laptop
>>
>> I'm trying to use a contributed package (rmetasim) to generate simulated
>> genetic datasets.  My scripts work fine when I run them on a Sun
>> workstation running Solaris 7 and when I run them on a ~4 year old
>> laptop PC that I have.  However, when I run them on my new laptop (Dell
>> Latitude D410 purchased in August 2005), the simulation will run for a
>> short (variable) period of time, then R crashes.  When it crashes, I
>> either get an Application Popup that says
>>
>> 'The instruction at "0x5ad71531" referenced memory at "0x00000014".  The
>> memory could not be "read".',
>>
>> or I simply get the popup saying that RGUI has experienced a problem,
>> would I like to send a report to Microsoft.  When the latter happens, an
>> event log is generated that says
>>
>> 'Fauling application rgui.exe, version 2.30.37590.0, fauling module
>> ntdll.dll, version 5.1.2600.2180, fault address 0x0001888f.'
>>
>> In both types of errors, the memory addresses listed are different every
>> time a crash occurs.
>>
>> I experience this problem primarily when attempting to simulate large
>> datasets.  My script saves the 'landscape' (the object being simulated)
>> every 100 time steps.  When R crashes, I can start it back up, reload
>> the last landscape saved before the crash, and continue simulating from
>> there.  Thus, there appears to be nothing wrong with the landscapes I'm
>> simulating.
>>
>> I've been in contact with the author of the contributed package I'm
>> using, and he has been unable to find any bugs in the package or
>> reproduce the error.  The bug seems to be platform specific since, as I
>> mentioned, I'm able to run my scripts crash-free on other computers.
>> However, I've tried running my scripts on another computer that is
>> nearly identical to mine (a Dell D610 purchased in August 2005), and
>> experience the same problems on that computer.  I've tried running my
>> scripts from a command window using 'R --vanilla', and experience the
>> same problems.  I've tried using the development version of R
>> (R-2.3.0alpha) and experience the same problems.
>>
>> In reading other bug reports I found that some people have had crashes
>> caused by buggy video drivers.  I've updated my video drivers, but it
>> didn't help.  However, I was only able to upgrade to the newest version
>> of the video driver; I haven't been able to find an alternative driver
>> to work with my display adapter.  The display adapter used by my
>> computer is the mobile Intel 915GM/GMS,910GML Express Chipset.
>>
>> I'm running R 2.2.1 (though I've tried R 2.0.1 and 2.3.0alpha and had
>> the same problem).  The machine on which I've had problems is a Dell
>> Latitude D410 laptop running Windows XP Pro SP2, 2GHz Pentium processor
>> with 1GB of RAM.  The output of the R command 'version' is;
>>
>> platform i386-pc-mingw32
>> arch     i386
>> os       mingw32
>> system   i386, mingw32
>> status
>> major    2
>> minor    2.1
>> year     2005
>> month    12
>> day      20
>> svn rev  36812
>> language R
>>
>> The exact sequence of commands I type, after launching the Rgui, is as
>> follows:
>>
>>  >library(rmetasim)
>>  >source("ArchI_burnin_NOdensdep-mod.R")
>>
>> The script I'm running (which obviously requires that the package
>> 'rmetasim' be installed) follows.  Beware that the script takes about 15
>> minutes to run on my old computer (the one on which it doesn't crash.
>> For what it's worth, that computer is a Dell Latitude C400 running
>> Windows XP Pro SP2, 866 MHz Pentium III processor, 512 MB of RAM).  Any
>> suggestions would be greatly appreciated.
>>
>> ##########################################################################
>> ##
>> #ArchI_burnin_NOdensdep-mod.R#
>>
>> habitats <- 1
>> carrycap <- 7500
>> stages <- 5
>> rland <- NULL
>> numreps <-  2
>> numsteps <- 50
>> stepsize <- 100
>> runlength <- numsteps*stepsize
>>
>> rland <- new.landscape.empty()
>> rland <- new.intparam.land(rland, h = habitats, s = stages, totgen =
>> runlength)
>> rland <- new.switchparam.land(rland, mp = 1)
>> rland <- new.floatparam.land(rland)
>>
>> #life history matrices at zero population density
>> SZ <- matrix (c(0.730, 0, 0, 0, 0,
>>            0.210, 0, 0, 0, 0,
>>            0, 0.470, 0, 0.946, 0,
>>            0, 0, 0.946, 0, 0,
>>            0, 0.470, 0, 0, 0.954), nrow=5, byrow = T)
>>
>> RZ <- matrix (c(0, 0, 0, .94, 0,
>>            0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 0), nrow = 5, byrow = T)
>>
>> M <- matrix (c(0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 0,
>>            0, 0, 0, 0, 1,
>>            0, 0, 0, 0, 0), nrow = 5, byrow = T)
>>
>> rland <- new.local.demo(rland,SZ,RZ,M)
>>
>> rland <-
>> new.epoch.island(rland,0,c(0,0,0,0,0),c(0,0,0,0,0),0,c(0,0,0,0,0),c(0,0,0,
>> 0,0),
>> 0,c(0,0,0,0,0),c(0,0,0,0,0), carry = rep(carrycap,habitats))
>>
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.001)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.001)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.001)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.002)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.002)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.002)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.003)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.003)
>> rland <- new.locus(rland, type = 1, ploidy = 2, transmission = 0,
>> numalleles = 1, mutationrate = 0.003)
>> rland <- new.locus(rland, type = 2, ploidy = 1, transmission = 1,
>> numalleles = 1, allelesize = 500, mutationrate = 0.005)
>>
>> rland <- new.individuals(rland, as.integer(c(2775,525,1050,975,2175)))
>>
>> rland.start <- rland
>> for (j in 1:numreps) {
>>   rland <- rland.start
>>   for (i in 1:numsteps) {
>>       rland <- simulate.landscape(rland, stepsize)
>>      save(rland, file=paste("ArchI_test_",j,"_",i,".rda",sep=""))
>>   }
>> }
>>
>>
>> --
>> ***********************************
>> Dr. Karen Martien
>> Southwest Fisheries Science Center
>> 8604 La Jolla Shores Drive
>> La Jolla, CA 92037
>> USA
>> Ph: 858-546-7058
>> Fax: 858-546-7003
>>
>> ______________________________________________
>> 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
>>     
>
>
>   

-- 
***********************************
Dr. Karen Martien
Southwest Fisheries Science Center
8604 La Jolla Shores Drive
La Jolla, CA 92037
USA
Ph: 858-546-7058
Fax: 858-546-7003




More information about the R-help mailing list