我正在尝试使用
RkmeansRcppMLPACK#include <RcppArmadillo.h>
#include <RcppMLPACK.h>
#include <RcppGSL.h>
#include <RcppDist.h>
#include <sstream>
#include <iostream>
#include <fstream>
#include<omp.h>
#include<gsl/gsl_math.h>
#include<gsl/gsl_rng.h>
#include<gsl/gsl_randist.h>
#include<gsl/gsl_sf.h>
// [[Rcpp::depends(RcppProgress)]]
#include <progress.hpp>
#include <progress_bar.hpp>
// [[Rcpp::depends(RcppArmadillo,RcppDist)]]
// [[Rcpp::depends(RcppMLPACK)]]
// [[Rcpp::depends(RcppGSL)]]
// [[Rcpp::plugins(cpp11)]]
// [[Rcpp::plugins(openmp)]]
using namespace mlpack::kmeans ;
using namespace arma;
我的
MakevarsCXX_STD = CXX17
GSL_CFLAGS=`${R_HOME}/bin/Rscript -e "RcppGSL:::CFlags()" 4`
GSL_LIBS=`${R_HOME}/bin/Rscript -e "RcppGSL:::LdFlags()"`
RCPP_LDFLAGS=`${R_HOME}/bin/Rscript -e "Rcpp:::LdFlags()"`
PKG_CXXFLAGS = $(SHLIB_OPENMP_CXXFLAGS) $(GSL_CFLAGS)
PKG_LIBS = $(SHLIB_OPENMP_CXXFLAGS) $(LAPACK_LIBS) $(BLAS_LIBS) $(FLIBS) $(GSL_LIBS) $(RCPP_LDFLAGS)
我正在使用 macOS ventura。当我尝试构建我的
RIn file included from /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/library/RcppMLPACK/include/mlpack/core.hpp:171,
from /Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/library/RcppMLPACK/include/RcppMLPACK.h:4,
from RcppExports.cpp:6:
/Library/Frameworks/R.framework/Versions/4.3-arm64/Resources/library/RcppMLPACK/include/mlpack/prereqs.hpp:46:10: fatal error: boost/math/special_functions/gamma.hpp: No such file or directory
>> 46 | #include <boost/math/special_functions/gamma.hpp>
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
但是，如果我只是在
Rcpp::sourcecppC++附注我使用
gccclangboostmlpack这个主题的记录有点少：
mlpack这就是我所做的（使用mlpack 3.4.2，请参阅下面的mlpack 4.0.1）：
我首先创建了一个 minimal C++ 文件，仅包含两个标头，没有做太多事情。
它看起来像这样，给予或接受：
#include <Rcpp/Rcpp>
#include <mlpack.h>
// [[Rcpp::depends(RcppArmadillo)]]
// [[Rcpp::depends(mlpack)]]
// [[Rcpp::export]]
void foo() {
Rcpp::Rcout << "Foo\n";
}
/*** R
foo()
*/
编译这意味着找到
mlpack这对我来说有点复杂，因为我碰巧（主要）在 Ubuntu 22.10 上工作，它只有一个较旧的 mlpack 3.4.2
作为发行版中方便的系统库。我认为使用较新的
mlpack版本4.*我不需要链接。正如我经常做的那样，我从单元测试中举了一个简单的例子。它有数据，以及调用。现在完整的文件如下：
#include <Rcpp/Rcpp>
#include <mlpack.h>
// Two include directories adjusted for my use of mlpack 3.4.2 on Ubuntu
#include <mlpack/core.hpp>
#include <mlpack/methods/kmeans/kmeans.hpp>
#include <mlpack/methods/kmeans/random_partition.hpp>
#include <mlpack/methods/neighbor_search/neighbor_search.hpp>
// [[Rcpp::depends(RcppArmadillo)]]
// [[Rcpp::depends(mlpack)]]
// This is 'borrowed' from mlpack's own src/mlpack/tests/kmeans_test.cpp
// and src/mlpack/tests/kmeans_test.cpp. We borrow the data set, and the
// code from the first test function. Passing data from R in easy thanks
// to RcppArmadillo, 'and left as an exercise'.
// Generate dataset; written transposed because it's easier to read.
arma::mat kMeansData(" 0.0 0.0;" // Class 1.
" 0.3 0.4;"
" 0.1 0.0;"
" 0.1 0.3;"
" -0.2 -0.2;"
" -0.1 0.3;"
" -0.4 0.1;"
" 0.2 -0.1;"
" 0.3 0.0;"
" -0.3 -0.3;"
" 0.1 -0.1;"
" 0.2 -0.3;"
" -0.3 0.2;"
" 10.0 10.0;" // Class 2.
" 10.1 9.9;"
" 9.9 10.0;"
" 10.2 9.7;"
" 10.2 9.8;"
" 9.7 10.3;"
" 9.9 10.1;"
"-10.0 5.0;" // Class 3.
" -9.8 5.1;"
" -9.9 4.9;"
"-10.0 4.9;"
"-10.2 5.2;"
"-10.1 5.1;"
"-10.3 5.3;"
"-10.0 4.8;"
" -9.6 5.0;"
" -9.8 5.1;");
// [[Rcpp::export]]
arma::Row<size_t> kmeansDemo() {
mlpack::kmeans::KMeans<mlpack::metric::EuclideanDistance,
mlpack::kmeans::RandomPartition> kmeans;
arma::Row<size_t> assignments;
kmeans.Cluster((arma::mat) trans(kMeansData), 3, assignments);
return assignments;
}
/*** R
kmeansDemo()
*/
现在，因为我使用的是mlpack 3.4.2我必须链接，所以我还需要运行Sys.setenv("PKG_LIBS"="-lmlpack")
——并且我必须从为mlpack设置的存储库中获取的示例中稍微调整标题4.1.*.链接步骤将根据您运行此链接的位置而有所不同。
但是这样，我的 R 会话就产生了结果：
> Sys.setenv("PKG_LIBS"="-lmlpack")
> Rcpp::sourceCpp("~/git/stackoverflow/76319284/answer.cpp")
> kmeansDemo()
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14] [,15] [,16] [,17] [,18] [,19] [,20] [,21] [,22] [,23] [,24] [,25] [,26] [,27] [,28] [,29] [,30]
[1,] 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0
>
运行 kmeans (mlpack 4.0.1)mlpack
，事情会变得更好、更容易。在 Ubuntu 上安装 4.0.1 后，头文件包含稍微简化了一点，命名空间发生了一点变化，我添加了对RcppEnsmallen 的 R 包依赖（它提供了优化例程）。最重要的是，我可以在不链接的情况下构建它。 更新了代码（针对 mlpack 4.0.1）
#include <Rcpp/Rcpp>
#include <mlpack.h>
// Adjusted for mlpack 4.0.1
#include <mlpack/core.hpp>
#include <mlpack/methods/kmeans.hpp>
#include <mlpack/methods/kmeans/random_partition.hpp>
#include <mlpack/methods/neighbor_search/neighbor_search.hpp>
// [[Rcpp::depends(RcppArmadillo)]]
// [[Rcpp::depends(RcppEnsmallen)]]
// [[Rcpp::depends(mlpack)]]
// [[Rcpp::plugins(cpp14)]]
// This is 'borrowed' from mlpack's own src/mlpack/tests/kmeans_test.cpp
// and src/mlpack/tests/kmeans_test.cpp. We borrow the data set, and the
// code from the first test function. Passing data from R in easy thanks
// to RcppArmadillo, 'and left as an exercise'.
// Generate dataset; written transposed because it's easier to read.
arma::mat kMeansData(" 0.0 0.0;" // Class 1.
" 0.3 0.4;"
" 0.1 0.0;"
" 0.1 0.3;"
" -0.2 -0.2;"
" -0.1 0.3;"
" -0.4 0.1;"
" 0.2 -0.1;"
" 0.3 0.0;"
" -0.3 -0.3;"
" 0.1 -0.1;"
" 0.2 -0.3;"
" -0.3 0.2;"
" 10.0 10.0;" // Class 2.
" 10.1 9.9;"
" 9.9 10.0;"
" 10.2 9.7;"
" 10.2 9.8;"
" 9.7 10.3;"
" 9.9 10.1;"
"-10.0 5.0;" // Class 3.
" -9.8 5.1;"
" -9.9 4.9;"
"-10.0 4.9;"
"-10.2 5.2;"
"-10.1 5.1;"
"-10.3 5.3;"
"-10.0 4.8;"
" -9.6 5.0;"
" -9.8 5.1;");
// [[Rcpp::export]]
arma::Row<size_t> kmeansDemo() {
mlpack::KMeans<mlpack::EuclideanDistance, mlpack::RandomPartition> kmeans;
arma::Row<size_t> assignments;
kmeans.Cluster((arma::mat) trans(kMeansData), 3, assignments);
return assignments;
}
/*** R
kmeansDemo()
*/
它的构建和运行当然是相同的，现在包含少量的默认日志记录：
> Rcpp::sourceCpp("answer.cpp")
> kmeansDemo()
[INFO ] KMeans::Cluster(): iteration 1, residual 14.8221.
[INFO ] KMeans::Cluster(): iteration 2, residual 1.77636e-15.
[INFO ] KMeans::Cluster(): converged after 2 iterations.
[INFO ] 186 distance calculations.
[,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14] [,15] [,16] [,17] [,18] [,19] [,20] [,21] [,22] [,23] [,24] [,25] [,26] [,27] [,28] [,29] [,30]
[1,] 2 2 2 2 2 2 2 2 2 2 2 2 2 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1
>