[R] svm in e1071 package segment fault in redhat 7.2

Martyn Plummer plummer at iarc.fr
Tue Mar 12 11:44:38 CET 2002


I can replicate this bug on Red Hat 7.2.  It seems to be a compiler
problem. Here is the relevant output from gdb

(gdb) frame 2
#2  0x403a2ab6 in SVC_Q::get_Q (this=0xbfffdae0, i=41, len=100) at svm.cpp:1131
1131                    if((start = cache->get_data(i,&data,len)) < len)
(gdb) p cache
$10 = (Cache *) 0x8b083a8
(gdb) p i
$11 = 41
(gdb) p &data
$12 = (Qfloat **) 0xbfffd984
(gdb) p len
$13 = 100
(gdb) frame 1
#1  0x4039d90e in Cache::get_data (this=0x8b083a8, index=146466881, 
    data=0xbfffd984, len=100) at svm.cpp:109
109             if(h->len) lru_delete(h);

The index argument to Cache::get_data is corrupted (should be 41
not 14646881).

I see that Red Hat have released a package for gcc3.  You can
download this from 

ftp://updates.redhat.com/7.2/en/os/i386/

or a mirror near you.  I recommend that you try this.

Martyn

On 11-Mar-2002 jason zhao wrote:
> Hi, Guys
> I run one simple R script like this
>        
> library(e1071) 
> data(iris)
> attach(iris)
> model <- svm (Species~., data=iris)
> 
> The R version is 1.4.1. The OS is standard redhat 7.2.
> Then I got a segmentation fault like this (run through
> R CMD BATCH scriptName)
> /usr/local/R/lib/R/bin/BATCH: line 54: 28513 Done     
>               (    
> echo "invisible(options(echo = TRUE))"; cat ${in};
> echo "proc.time()" )
>      28514 Segmentation fault      | ${R_HOME}/bin/R
> ${opts} >${out} 2>&1
> 
> There is no problem if the R-binary is running redhat
> 6.2.
> 
> Does anyone see this problem?  Is there a workaround
> for this?
>   
> Help is appreciated.
> 
> Jason
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._



More information about the R-help mailing list