An Introduction to R Package Development with Rcpp

Resources

• http://rcpp.org/
• http://adv-r.had.co.nz/
• http://r-pkgs.had.co.nz/

Installation

1. Setup R for development
   • Mac OSX: Install XCode Developer tools from the App Store
   • Windows: Install RTools from https://cloud.r-project.org/bin/windows/Rtools/
   • Linux: Install the core software development utilities sudo apt-get install r-base-dev
2. Then install the R Packages Rcpp and RcppArmadillo install.packages(c("Rcpp", "RcppArmadillo"))
3. devtools::has_devel()

Getting started

cppFunction() allows you to write C++ functions in R.

library(Rcpp)
cppFunction(
  'int add(int x, int y, int z) {
    int sum = x + y + z;
    return(sum);
  }'
)

add

## function (x, y, z) 
## .Primitive(".Call")(<pointer: 0x107bf45a0>, x, y, z)

add(1, 2, 3)

## [1] 6

Rcpp types

• Scalars: int, double, String, bool
• Vectors: IntegerVector, NumericVector, CharacterVector, LogicalVector
• Matrices: IntegerMatrix, NumericMatrix, CharacterMatrix, LogicalMatrix
• R output: List, DataFrame

Still use [] for accessing elements in Rcpp vectors, but () for accessing alements in Rcpp matrices.

sourceCpp

Can use standalone C++ files with extension .cpp and source them into R using sourceCpp(). If you do this, must include

#include <Rcpp.h>
using namespace Rcpp;

at the top of your C++ file and for each function that you want to use in R, prefix it with

// [[Rcpp::export]]

Then, you can source your function using

sourceCpp("code/timesTwo.cpp")
timesTwo(1:10)

##  [1]  2  4  6  8 10 12 14 16 18 20

Package Skeleton

Workflow

After editing/creating your package,

1. Generate the necessary documentation changes with devtools::document()
2. Build and reload your packages

RcppArmadillo

Armadillo is a C++ linear algebra library aiming towards a good balance between speed and ease of use. We can use the R package RcppArmadillo to interface with this library.

• DESCRIPTION: LinkingTo: Rcpp, RcppArmadillo
• .cpp file:

#include <RcppArmadillo.h>
// [[Rcpp::depends(RcppArmadillo)]]
using namespace Rcpp; using namespace arma;