There are two distinct and orthogonal questions here: 1) Is it possible to package an R program/system in such a way that it can be distributed as a single executable file (including the R system, any necessary loaded packages, plus the application program) and run without an installation of R (which sets up library directories etc.)?
2) Is it possible to compile R to bytecodes or machine language? The answers to these questions are independent. It is perfectly possible to have an interpreted system which is distributed as a single executable file (using e.g. unexec to save the fully-loaded state). It is also perfectly possible to have a compiled system which requires configuration of library directories etc. I believe the original question on this thread was about (1), but I am not certain. It also appears that the answer to (1) is 'no', but I am not certain about that either. I wouldn't think that it would be impossible to add 'unexec' to R, but I don't know R's internals. -s On Wed, May 13, 2009 at 4:17 PM, Matthew Dowle <mdo...@mdowle.plus.com> wrote: > > Does Ra get close to compiled R ? "The R code is compiled on the fly to > bytecode which is executed internally by an interpreter in C." The timing > tests look impressive. > http://www.milbo.users.sonic.net/ra/ > > ______________________________________________ > R-devel@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-devel > ______________________________________________ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel