C++11 Header Files for 'cereal'

To facilitate using 'cereal' with Rcpp. 'cereal' is a header-only C++11 serialization library. 'cereal' takes arbitrary data types and reversibly turns them into different representations, such as compact binary encodings, XML, or JSON. 'cereal' was designed to be fast, light-weight, and easy to extend - it has no external dependencies and can be easily bundled with other code or used standalone. Please see < http://uscilab.github.io/cereal> for more information.

This package provides R with access to cereal header files. cereal is a header-only C++11 serialization library. cereal takes arbitrary data types and reversibly turns them into different representations, such as compact binary encodings, XML, or JSON. For more information, please visit the official website of cereal project: http://uscilab.github.io/cereal/

This package can be used via the LinkingTo: field in the DESCRIPTION field of an R package and the Rcpp::depends in the Rcpp-attributes. The R and Rcpp infrastructure tools will know how to set include flags properly.

Please use the devtools::install_github to install the latest version of Rceral and use Rcereal::update_version to install the content of the header files of cereal.


In this project, we will not explain how to use cereal in c++ because the official cereal project has already provides a complete documentation. Please visit the Quick Start page to learn how to use cereal.

The following example briefly shows how to use the Rcereal in Rcpp-attributes to serialize a user defined c++ structure into raw vector and deserialize from the raw vector.

#include <sstream>
#include <cereal/archives/binary.hpp>
#include <Rcpp.h>
struct MyClass
  int x, y, z;
  // This method lets cereal know which data members to serialize 
  template<class Archive>
  void serialize(Archive & archive)
    archive( x, y, z ); // serialize things by passing them to the archive 
using namespace Rcpp;
RawVector serialize_myclass(int x = 1int y = 2int z = 3) {
  MyClass my_instance;
  my_instance.x = x;
  my_instance.y = y;
  my_instance.z = z;
  std::stringstream ss;
    cereal::BinaryOutputArchive oarchive(ss); // Create an output archive 
  ss.seekg(0, ss.end);
  RawVector retval(ss.tellg());
  ss.seekg(0, ss.beg);
  ss.read(reinterpret_cast<char*>(&retval[0]), retval.size());
  return retval;
void deserialize_myclass(RawVector src) {
  std::stringstream ss;
  ss.write(reinterpret_cast<char*>(&src[0]), src.size());
  ss.seekg(0, ss.beg);
  MyClass my_instance;
    cereal::BinaryInputArchive iarchive(ss);
  Rcout << my_instance.x << "," << my_instance.y << "," << my_instance.z << std::endl;
/*** R
raw_vector <- serialize_myclass(1, 2, 4)

To compile the cpp file, the user must enable the support of c++11 before using Rcpp::sourceCpp.

Rcpp::sourceCpp("<the path to the cpp file>")

To use cereal with Rcpp in the following way, the user must remember two points:

  • The supports of c++11 must be enabled.
  • The Rcereal must be install properly.

If you see the compiler reports the missing header files, please use the Rcereal::update_version() to update the content of cereal from github. You can manual check whether a directory named sereal is in the folder system.file("include", package = "Rcereal").

If you see lots of error during compiling, please check the version of the compiler and the content of the environment variable PKG_CXXFLAGS. As far as I know, cereal will fail if the gcc-4.6 is used.

os x


Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.2.1 by Wush Wu, a month ago


Report a bug at https://github.com/wush978/Rcereal/issues

Browse source code at https://github.com/cran/Rcereal

Authors: Wush Wu, Randolph Voorhies, and Shane Grant

Documentation:   PDF Manual  

BSD_2_clause + file LICENSE license

Suggests Rcpp, git2r, httr, testthat, tools, devtools, knitr, rmarkdown

System requirements: C++11

See at CRAN