[R] Include C++ DLL, error in ...: C symbol name not in load table
Duncan Murdoch
murdoch.duncan at gmail.com
Wed Apr 20 22:24:13 CEST 2011
On 20/04/2011 4:06 PM, Sascha Vieweg wrote:
> Hello R experts
>
> I am googling and reading around, however, I can't get it working
> (perhaps because I do not understand much C, however, I'll give it
> a try). I am trying to include C++ code into an R routine, where
> the C++ code looks:
>
> #include<iostream>
> using namespace std;
> void foo (double* x, double* y, double* out)
> {
> out[0] = x[0] + y[0];
> }
>
> Back in R, the command
>
> R CMD SHLIB --preclean -o xplusy
>
> works fine resulting in two new files, xplusy.o and xplusy.so. The
> wrapper in R is:
>
> dyn.load("xplusy.so")
> xplusy<- function(x, y){
> .C("foo", as.double(x), as.double(y), out=double(1))$out
> }
> xplusy(1, 2)
> dyn.unload("xplusy.so")
>
> Now, dyn.load() works and xplusy also shows up in getLoadedDLLs().
> However, when invoking the function, xplusy(1, 2), R complains:
>
> Error in .C("foo", as.double(x), as.double(y), out = double(1)): C
> symbol name "foo" not in load table
>
> I found some hints concerning Fortran code producing this error
> message, but no help concerning C code.
You have C++ code, not C code. C++ normally mangles names of exports.
To get this to work, you should surround your declarations with
extern "C" {
}
Another possibility is to use the Rcpp package; it writes the interface
code for you.
Duncan Murdoch
>
> Any help tips?
>
> Thanks, *S*
>
> $. uname -a
> Darwin * 10.7.3 Darwin Kernel Version 10.7.3[...]
>
> $: c++ --version
> i686-apple-darwin10-g++-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666)
> (dot 3)
>
> platform i386-apple-darwin9.8.0
> arch i386
> os darwin9.8.0
> system i386, darwin9.8.0
> status
> major 2
> minor 12.2
> year 2011
> month 02
> day 25
> svn rev 54585
> language R
> version.string R version 2.12.2 (2011-02-25)
>
More information about the R-help
mailing list