On Mon, 21 Aug 2006, Pai-Hsien Hung wrote: > > > Thanks for your reply. > > I have already read the manual again, and use the R.dll (Ch 8.2.2) (Am I > right?).
I don't know, as it depends on the (unstated) version of R where in the manual it is. If I were using VC++, I would be more inclined to use the (D)COM interface. > Running the example in the section is OK. But, when I try to add > "do_matprod" function in the same example, I don't know what args should put > in. The function description in array.c is SEXP attribute_hidden > do_matprod( SEXP call, SEXP op, SEXP args, SEXP rho), and where I can get > the definition and description of each variables, call, op, args, .... That is a private function not in the API nor in the header files. Indeed, in upcoming versions of R it is not exported from R.dll (that's what 'attribute_hidden' entails). Since it is private you will not find the information you are seeking in the documentation. You asked about using code in packages, and this is not in a package. The time-series functions are in R code, not C code, so I was answering the question you actually asked. > Thanks, > > -----Original Message----- > From: Prof Brian Ripley [mailto:[EMAIL PROTECTED] > Sent: Friday, August 18, 2006 10:32 PM > To: Pai-Hsien Hung > Cc: r-devel > Subject: Re: [Rd] How to use R and add-on packages as library in C/C++ > > This is indeed in the manual you mention. To use R functions and not just > standalone libRmath, you need to embed R. > > On Fri, 18 Aug 2006, Pai-Hsien Hung wrote: > > > Hi, > > I have a question for building a standalone exe file via C/C++ with R > > and add-on packages in Windows (Compiler is VC). It may looks like > > ... > > > > #include <R.h> > > #include <Rmath.h> > > > > int main() { > > double *x, *y, z; > > > > // ... (Variables initial) > > > > z = ar( x, ...); > > // In this example, I want use the ar (time series) function (in stats > package) > > // as a library funciton. > > > > // ... (Other statement) > > > > return 0; > > } > > > > In the "Writing R Extensions" manual, just demo with numerical > > functions, not statistical functions. I don't know is it possible to > > use statistical functions (Time series at least) or other functions in > > the add-on package, and how to make it. Thanks for your patient to > > reading this. ^^ > > > > > > -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595 ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel
