[Rcpp-devel] Linking problem when wrapping a armadillo-based C++ library in a R package using RcppArmadillo

[julien chiquet]

On Wed, Jul 25, 2018 at 5:57 PM Dirk Eddelbuettel <edd at debian.org> wrote:

>
> On 25 July 2018 at 17:30, julien chiquet wrote:
> | Hi,
> |
> | I am trying to wrap the optim library (https://github.com/kthohr/optim)
> in
> | R via an R package using Rcpp/RcppArmadillo. Indeed, optim relies on
> | armadillo itself. The github repository can be found here
> | https://github.com/jchiquet/optimLibR
> |
> | To do that, I put the headers of optim into inst/include/ and the cpp
> files
> | into src/. Then I tried to "heuristically" pull everything together in
> the
> | src/Makevars files, after my (obviously poor) understanding of the R
> manual
> | pages about Makevars and from the various examples that I copied from the
> | web (especially from the RcppMLPACK1 which, relies on the C++ library
> | MLPACK, itself based on armadillo
> | https://github.com/rcppmlpack/RcppMLPACK1/blob/master/src/Makevars).
> |
> | The package installation (including compilation of optim) successes (at
> | least, the package can be installed). Then I tried to create two small
> Rcpp
> | functions for two examples available in the original optim library,
> | starting with the following lines
> |
> | // [[Rcpp::depends(RcppArmadillo)]]// [[Rcpp::plugins(cpp11)]]
> | #define USE_RCPP_ARMADILLO
> | #include "optim.hpp"
> |
> | (The #define command is an option of the optim library.)
> |
> | However, when I call the corresponding R functions, I get the following
> | Armadillo errors in R which I do not have when running them in 'pure' C++
> | (I am pretty sure it is not a syntax problem, the Rcpp code and the
> | original C++ code are hopefully identical, except for the headers).
> |
> | > library(optimLibR)> ackley_function()error: arma::memory::acquire():
> out of memory
> |
> | > library(optimLibR)> logit_optimLib()error: Mat::init(): requested size
> is not compatible with column vector layout
> |
> | So it seems to me that the header of armadillo are correctly defined but
> I
> | am not sure that I reach the src (cpp) part of armadillo which is
> required
> | at the execution.
> |
> | If anyone can confirm or help pointing the origin of the error, I would
> be
> | debtful...
> |
> | Thanks for your time,
>
> Good news / bad news. I think the good news is that everything builds, so
> you
> have no linking issue. In fact, no RcppArmadillo package links to Armadillo
> as we do everything via the headers.
>
> But you seem to have run-time issues which _could_ stem from the fact that
> RcppArmadillo sets a bunch of #define variables and your use may be
> slightly
> different / conflict with these. Maybe you can check with Keith about how
> his optim(lib) could / should be called from R.
>

Thanks for your prompt and helpful answer. At least I know I do not have to
struggle with the Makevars.

I'll check with Keith.

j


> Dirk
>
> --
> http://dirk.eddelbuettel.com | @eddelbuettel | edd at debian.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20180725/eaa95967/attachment-0001.html>