[R] postscript (eps) / latex / par(mfg=...) / problem!

[Dan Bolser]

Should I post this to 'bugs'?

On Tue, 26 Apr 2005, Dan Bolser wrote:

>
>The same problem I am having has been reported here 
>
>http://tolstoy.newcastle.edu.au/R/devel/04a/0344.html
>
>
>Namely that using par(mfg=...) with a postscript (eps) for inclusion with
>latex makes the figure appear upside down and back to front (flipped)!
>
>Converting the dvi to ps makes matters worse (the eps seems to be broken),
>however, it appears fine with gv.
>
>Here is (basically) the code I am using...
>
>>dat <- read.table("x.dmp", header=1)
>>t(dat)
>t(dat)
>            1   2   3   4  5  6 7  8 9 10 11 12 13 14 15 16 17 18 19 20
>CHAINS      1   2   3   4  5  6 7  8 9 10 11 12 13 14 16 20 23 24 26 28
>FREQUENCY 886 792 136 201 16 58 6 21 3  9  3  9  1  4  3  1  1  1  1  1
>>
>>postscript(
>+           "x.eps",
>+           width = 6.0,
>+           height = 6.0,
>+           horizontal = FALSE,
>+           onefile = FALSE,
>+           paper = "special",
>+           )
>>
>>par(mfg=c(1,1))
>>par(mar=c(3,4,1,2))
>>plot(dat,type='b')
>>
>>par(mfg=c(2,1))
>>par(mar=c(4,4,0,2))
>>plot(dat,type='b', log='y')
>>
>>dev.off()
>
>
>Including the resulting file in a latex document like this...
>
>begin{figure}
>\centering
>\includegraphics[width=\textwidth]{x.eps}
>\caption[X]
>{
>Hello!
>}
>\label{xFig}
>\end{figure}
>
>The result is an upside down (flipped) version of my plot. I tried
>rotating 180 degrees (based on similar problems people were having on the
>list), but then it just gets worse (most of the plot is off the page). If
>I convert the dvi to ps (dvips -Ppdf my.tex.dvi -o my.tex.ps) it gets
>worse (a tiny speck where the image should be).
>
>After removing the two mfg commands (which I use to add grid lines (not
>shown for clarity)) everything is fine! Some how mfg is snarling things
>up.
>
>OK, I just had a brain wave (dont laugh). Here is a diff of the working
>eps vs the broken eps...
>
>diff broken working
>
>78a79,80
>> %%Page: 1 1
>> bp
>229c231
>< 57.60 43.20 403.20 201.60 cl
>---
>> 57.60 57.60 403.20 216.00 cl
>417c419
>< %%Pages: 0
>---
>> %%Pages: 1
>
>
>Does that help anyone debug my problem? Like I said, both look identical
>via gv, and are 'conceptually' identical in R.
>
>Here are my vitals 
>
>Linux 2.4.20-31.9 i686 athlon i386 GNU/Linux
>R 2.0.0 (2004-10-04).
>GNU Ghostscript 7.05 (2002-04-22)
>
>Anything else you need?
>
>Please help!
>
>______________________________________________
>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
>