R devel - Jan 2010 - Using multicore with an open pdf device results in corrupt pdf (PR#14186)

The attached code produces corrupted pdfs (test2.pdf, test4.pdf and
test5.pdf). The resulting pdf depends on how many cores are available on
the machine. 

I don't see why there should be any difference between the pdfs (exept for
the timestamp). Doing many operations involving mclapply can increase the
size of the resulting pdf by ten times!

Thank you for checking this.

require(multicore)

pdf('test.pdf')
y <- unlist(lapply(1:50, identity))
plot(y)
print(y)
dev.off()

options(cores = 3)

pdf('test2.pdf')
y <- unlist(mclapply(1:50, identity))
plot(y)
print(y)
dev.off()

pdf('test3.pdf')
y <- unlist(lapply(1:50, identity))
plot(y)
print(y)
dev.off()

options(cores = 2)

pdf('test4.pdf')
y <- unlist(mclapply(1:50, identity))
plot(y)
print(y)
dev.off()

options(cores = 8)

pdf('test5.pdf')
y <- unlist(mclapply(1:50, identity))
plot(y)
print(y)
dev.off()


--please do not edit the information below--

Version:
 platform = x86_64-unknown-linux-gnu
 arch = x86_64
 os = linux-gnu
 system = x86_64, linux-gnu
 status = 
 major = 2
 minor = 10.1
 year = 2009
 month = 12
 day = 14
 svn rev = 50720
 language = R
 version.string = R version 2.10.1 (2009-12-14)

Locale:
LC_CTYPE=de_CH.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=de_CH.UTF-8;LC_MONETARY=C;LC_MESSAGES=de_CH.UTF-8;LC_PAPER=de_CH.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=de_CH.UTF-8;LC_IDENTIFICATION=C

Search Path:
 .GlobalEnv, package:skewt, package:rgl, package:ggplot2, package:reshape,
package:plyr, package:proto, package:VGAM, package:stats4, package:splines,
package:latticeExtra, package:RColorBrewer, package:doMC, package:multicore,
package:foreach, package:codetools, package:iterators, package:abind,
package:seqinr, package:mvbutils, mvb.session.info, package:tools,
package:robust, package:rrcov, package:pcaPP, package:mvtnorm,
package:robustbase, package:MASS, package:glmmML, package:playwith,
package:grid, package:gWidgetsRGtk2, package:cairoDevice, package:lattice,
package:gWidgets, package:graphics, package:grDevices, package:datasets,
package:fortunes, package:sfsmisc, package:stats, package:utils,
package:methods, Autoloads, package:base

How is this a bug in R? First, multicore is not R. Second, you're  
running multicore with GUI code loaded which it explicitly tells you  
that it won't work. Third, the code you provided does produce correct  
PDFs (tested on the same platform you provided) in a clean session  
(unsurprisingly).

Cheers,
Simon


On Jan 15, 2010, at 9:15 , kollerma at stat.math.ethz.ch wrote:
> The attached code produces corrupted pdfs (test2.pdf, test4.pdf and
> test5.pdf). The resulting pdf depends on how many cores are  
> available on
> the machine.
>
> I don't see why there should be any difference between the pdfs  
> (exept for
> the timestamp). Doing many operations involving mclapply can  
> increase the
> size of the resulting pdf by ten times!
>
> Thank you for checking this.
>
> require(multicore)
>
> pdf('test.pdf')
> y <- unlist(lapply(1:50, identity))
> plot(y)
> print(y)
> dev.off()
>
> options(cores = 3)
>
> pdf('test2.pdf')
> y <- unlist(mclapply(1:50, identity))
> plot(y)
> print(y)
> dev.off()
>
> pdf('test3.pdf')
> y <- unlist(lapply(1:50, identity))
> plot(y)
> print(y)
> dev.off()
>
> options(cores = 2)
>
> pdf('test4.pdf')
> y <- unlist(mclapply(1:50, identity))
> plot(y)
> print(y)
> dev.off()
>
> options(cores = 8)
>
> pdf('test5.pdf')
> y <- unlist(mclapply(1:50, identity))
> plot(y)
> print(y)
> dev.off()
>
>
> --please do not edit the information below--
>
> Version:
> platform = x86_64-unknown-linux-gnu
> arch = x86_64
> os = linux-gnu
> system = x86_64, linux-gnu
> status > major = 2
> minor = 10.1
> year = 2009
> month = 12
> day = 14
> svn rev = 50720
> language = R
> version.string = R version 2.10.1 (2009-12-14)
>
> Locale:
> LC_CTYPE 
> = 
> de_CH 
> .UTF 
> -8 
> ;LC_NUMERIC 
> = 
> C 
> ;LC_TIME 
> = 
> en_US 
> .UTF 
> -8 
> ;LC_COLLATE 
> = 
> de_CH 
> .UTF 
> -8 
> ;LC_MONETARY 
> = 
> C 
> ;LC_MESSAGES 
> = 
> de_CH 
> .UTF 
> -8 
> ;LC_PAPER 
> = 
> de_CH 
> .UTF 
> -8 
> ;LC_NAME 
> = 
> C 
> ;LC_ADDRESS 
> =C;LC_TELEPHONE=C;LC_MEASUREMENT=de_CH.UTF-8;LC_IDENTIFICATION=C
>
> Search Path:
> .GlobalEnv, package:skewt, package:rgl, package:ggplot2,  
> package:reshape, package:plyr, package:proto, package:VGAM,  
> package:stats4, package:splines, package:latticeExtra,  
> package:RColorBrewer, package:doMC, package:multicore,  
> package:foreach, package:codetools, package:iterators,  
> package:abind, package:seqinr, package:mvbutils, mvb.session.info,  
> package:tools, package:robust, package:rrcov, package:pcaPP,  
> package:mvtnorm, package:robustbase, package:MASS, package:glmmML,  
> package:playwith, package:grid, package:gWidgetsRGtk2,  
> package:cairoDevice, package:lattice, package:gWidgets,  
> package:graphics, package:grDevices, package:datasets,  
> package:fortunes, package:sfsmisc, package:stats, package:utils,  
> package:methods, Autoloads, package:base
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>