The Rcpp package integrates R and C++ via R functions and a (header-only) C++ library.
All underlying R types and objects, i.e., everything a
SEXP represents internally in R, are matched to
corresponding C++ objects. This covers anything from vectors, matrices
or lists to environments, functions and more. Each SEXP
variant is automatically mapped to a dedicated C++ class. For example,
numeric vectors are represented as instances of the
Rcpp::NumericVector class, environments are represented as
instances of Rcpp::Environment, functions are represented
as Rcpp::Function, etc … The Rcpp-introduction
vignette (now published as a TAS
paper; an earlier
introduction was also published as a JSS paper provides a
good entry point to Rcpp as do the Rcpp
website, the Rcpp page and
the Rcpp Gallery. Full
documentation is provided by the Rcpp
book.
Other highlights:
The conversion from C++ to R and back is driven by the templates
Rcpp::wrap and Rcpp::as which are highly
flexible and extensible, as documented in the Rcpp-extending
vignette.
Rcpp also provides Rcpp modules, a framework that allows exposing C++ functions and classes to the R level. The Rcpp-modules vignette details the current set of features of Rcpp-modules.
Rcpp includes a concept called Rcpp sugar that brings many R functions into C++. Sugar takes advantage of lazy evaluation and expression templates to achieve great performance while exposing a syntax that is much nicer to use than the equivalent low-level loop code. The Rcpp-sugar gives an overview of the feature.
Rcpp attributes provide a high-level syntax for declaring C++ functions as callable from R and automatically generating the code required to invoke them. Attributes are intended to facilitate both interactive use of C++ within R sessions as well as to support R package development. Attributes are built on top of Rcpp modules and their implementation is based on previous work in the inline package. See the Rcpp-atttributes vignettes for more details.
The package ships with ten pdf vignettes, including a recent introduction to Rcpp now published as a paper in TAS (and as a preprint in PeerJ). Also available is an earlier introduction which was published as a JSS paper.
Among the other vignettes are the Rcpp FAQ and the introduction to Rcpp Attributes. Additional documentation is available via the Rcpp book by Eddelbuettel (2013, Springer); see ‘citation(“Rcpp”)’ for details.
Rcpp follows the C++ motto of “you pay only for what you use” and imposes no run-time performance penalty: Rcpp outperforms related packages in direct comparison, see for example this repo for details.
Compile-time performance can be tuned by selecting components. But it
is also worth noting that use of ccache will (strongly)
dominate all such possible component choices, we have previously
recommended its use.
The Rcpp Gallery showcases over one hundred fully documented and working examples. The package RcppExamples contains a few basic examples covering the core data types.
A number of examples are included, as are well over one thousand unit tests which provide additional usage examples.
An earlier version of Rcpp, containing what we now call the ‘classic Rcpp API’ was written during 2005 and 2006 by Dominick Samperi. This code has been factored out of Rcpp into the package RcppClassic, and it is still available for code relying on the older interface. New development should always use this Rcpp package instead.
Other usage examples are provided by packages using Rcpp. As of early July 2022, there are 2560 CRAN packages using Rcpp, a further 252 BioConductor packages in its current release as well as an unknown number of GitHub, Bitbucket, R-Forge, … repositories using Rcpp. All these packages provide usage examples for Rcpp. The package is in widespread use and has been downloaded over 61 million times (per the partial logs from the cloud mirrors of CRAN).
Rcpp released on CRAN are carefully tested and curated. CRAN ensures they interoperate with all other CRAN package on all test environment. The released and tested versions are available via all mirrors of CRAN network, and can be installed from within R via
install.packages("Rcpp")For the last several releases, we also made interim candidate releases available on the Rcpp Drat Repo. Versions from a drat repo can be installed either by just temporarily setting the drat repo as in
install.packages("Rcpp", repos="https://RcppCore.github.io/drat")or by setting a drat repo more permanently (as described in the documentation of the drat package).
Testing the release candidates prior to actual release help. Please run this if you can.
To install from source, ensure you have a complete package development environment for R as discussed in the relevant documentation; also see questions 1.2 and 1.3 in the Rcpp-FAQ.
If you want to run Rcpp on another (not-tested on CRAN) platform, or on releases older than the previous release, we suggest you do your due diligence and test accordingly. Rcpp is provided by an all-volunteer team with finite resources. We work hard to test Rcpp with several thousand CRAN packages using it—but we cannot test on outdated versions of R or your OS.
The best place for questions is the Rcpp-devel mailing list hosted at R-forge. Note that in order to keep spam down, you must be a subscriber in order to post. One can also consult the list archives to see if your question has been asked before.
Another option is to use StackOverflow and
its ‘rcpp’ tag. Search functionality (use rcpp in
squared brackets as in [rcpp]
my question terms to tag the query) is very valuable as many
questions have indeed been asked, and answered, before.
The issue tickets at the GitHub repo are the primary bug reporting interface. As with the other web resources, previous issues can be searched as well.
Dirk Eddelbuettel, Romain Francois, JJ Allaire, Kevin Ushey, Qiang Kou, Nathan Russell, Iñaki Ucar, Doug Bates, and John Chambers
GPL (>= 2)