Re: [Rcpp-devel] Filling a list with iterator, or setting names from CharacterVector/SEXP?

2012-08-30 Thread Christian Gunning
To clarify a little, I can avoid most of this hassle by constructing my named lists in R. I would like to compute on the names of those lists on the C++ end, though -- things akin to R's which(names(mylist)=='myname') and easily looping through names. -Christian On Thu, Aug 30, 2012 at 9:52 PM, C

[Rcpp-devel] Filling a list with iterator, or setting names from CharacterVector/SEXP?

2012-08-30 Thread Christian Gunning
I'm trying to implement a class lets me access elements by name in the actual C++ code. I'm handing in a matrix from R, and using the colnames and rownames of the matrix to generate a named list containing the relevant indices. My main interest here is legible code -- I'm trying to set up user-modi

[Rcpp-devel] [ANN] RcppArmadillo 0.3.3.91 at R-Forge

2012-08-30 Thread Dirk Eddelbuettel
Earlier today, Conrad has released 3.3.91 aka "v3.4 beta 1", which I wrapped into RcppArmadillo 0.3.3.91. The NEWS entry below details some of the changes (performance gains, beginning sparse matrix support, small enhancements to fastLm). Because CRAN now prefers to get only about one upload per

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Douglas Bates
On Thu, Aug 30, 2012 at 4:29 PM, Rodney Sparapani wrote: > On 08/30/2012 04:14 PM, Douglas Bates wrote: >> >> It is quite possible that you would have gotten a buggy version if you >> checked out a copy of the SVN archive at some random time. Making >> such a change is something I would tend to f

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Rodney Sparapani
On 08/30/2012 04:14 PM, Douglas Bates wrote: It is quite possible that you would have gotten a buggy version if you checked out a copy of the SVN archive at some random time. Making such a change is something I would tend to forget until tests started failing. I can confirm that it is not a bu

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Douglas Bates
Somehow you have obtained a buggy version of RcppEigen_0.3.1.tar.gz I just downloaded that file from a CRAN mirror and this version has the file RcppEigen/inst/include/Eigen/CholmodSupport starting with #ifndef EIGEN_CHOLMODSUPPORT_MODULE_H #define EIGEN_CHOLMODSUPPORT_MODULE_H #include "SparseC

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Rodney Sparapani
On 08/30/2012 03:24 PM, Rodney Sparapani wrote: # R CMD INSTALL RcppEigen_0.3.1.tar.gz * installing to library '/opt/local/lib64/R/library' * installing *source* package 'RcppEigen' ... ** package 'RcppEigen' successfully unpacked and MD5 sums checked ** libs g++ -I/opt/local/lib64/R/include -I/

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Rodney Sparapani
On 08/30/2012 02:24 PM, Douglas Bates wrote: But the RcppEigen package should not access Eigen 3.1.1 headers other than the ones that it provides. Yes, in a sane world we would be able to install Debian packages of libraries like Eigen 3.1.1, Suitesparse, etc. and access the code from an R pack

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Douglas Bates
On Thu, Aug 30, 2012 at 12:41 PM, Rodney Sparapani wrote: > On 08/29/2012 04:42 PM, Douglas Bates wrote: >> >> Perhaps you have an out-of-date version of the RcppEigen package (or >> maybe the Mac version hasn't been built or ...). In any case >> RcppEigenForward should include RcppEigenCholmod.h

Re: [Rcpp-devel] portable -Wall option in $PKG_CPPFLAGS

2012-08-30 Thread Dirk Eddelbuettel
On 29 August 2012 at 20:52, Peng Yu wrote: | Hi, | | I use -Wall option in $PKG_CPPFLAGS. | | PKG_CPPFLAGS=`echo "Rcpp:::CxxFlags()"|R --vanilla --slave` -Wall -I. | | But when I check the package, it says | | * checking for portable compilation flags in Makevars ... WARNING | Non-portable fla

Re: [Rcpp-devel] Advice on installing CHOLMOD/Suitesparse

2012-08-30 Thread Rodney Sparapani
On 08/29/2012 04:42 PM, Douglas Bates wrote: Perhaps you have an out-of-date version of the RcppEigen package (or maybe the Mac version hasn't been built or ...). In any case RcppEigenForward should include RcppEigenCholmod.h, not cholmod.h and the correct version would be at /opt/local/lib64/R/