[R] X11/ graphics problem

Rick Bilonick rab at nauticom.net
Wed Nov 19 16:26:34 CET 2003

Stephen Henderson wrote:

>I'm moving from using R 1.8 for Windows to using Linux (Redhat version 9)
>and I cannot get any graphics. However everything else maths, models
>equations is fine-- and much quicker. I built from the source file following
>the commands in the install manual.

>I've also had a look at the requirements under R/bin/linux/redhat/9/i386--
>but I'm having some trouble mapping these to the gui installation names,
>though Xfree86 which I presume contains libX11.so.6 is installed. This
>document also mentions preference for 'gnorpm'. Whats that?
>Has anyone else had this problem? have an idea what is likely missing? or
>how to reconfigure R?
I have R 1.8 installed on 3 different systems running RH 9 using the 
precompiled binaries. R makes plots on all of them. One of the systems 
is a laptop (Toshiba Satellite 2535cds). For some reason on this system 
there is a readline problem and I have to use:

 > R --no-readline

in order to run R but the graphics work fine. Before running 1.8 on this 
machine, I had compiled 1.7.1 (it took many many hours). R would run BUT 
I still had to use the --no-readline option AND the "x11" function would 
not work. (As a work around, I just used "pdf" and used the "system" 
function with gv to view the plots.) Since then I've installed 1.8 using 
the precompiled rpm available from the CRAN mirrors and it works except 
for the readline problem.

Maybe if you install the rpm your graphics will work. If not, "pdf" 
should work. (I'm assuming you have XFree86's Xwindows running.)

BTW, gnorpm is a graphical frontend to rpm. Although to install the one 
rpm for R 1.8, all you need is to do as root:

 > rpm -i R-1.8.0-1.i386.rpm


 > rpm -U R-1.8.0-1.i386.rpm

if upgrading from an earlier version of R. I always like to do:

 > rpm -U --test R-1.8.0-1.i386.rpm

first, just to see if there are any problems before actually upgrading. 
I'm not sure why anyone would go through the trouble and time needed for 
compiling when the appropriate rpm is readily available.

Rick B.

More information about the R-help mailing list