◀Table of Contents
GraalVM R Runtime
GraalVM provides a GNU-compatible R runtime to run R programs directly or in the REPL mode. It can run R code at unparalleled performance, and seamlessly integrates with the GraalVM ecosystem. The project name behind GraalVM’s R runtime development is FastR.
The GraalVM R runtime is an experimental GraalVM component.
Installing R
The R language runtime is not provided by default, and can be added to GraalVM with the GraalVM Updater, gu
, tool:
gu install r
After this step, the R
and Rscript
launchers will become available in the GRAALVM_HOME/bin
directory.
The R language home directory, which will be further referenced as $R_HOME
, is located in languages/R
.
Prerequisites
GraalVM’s R runtime requires zlib and the OpenMP runtime library. Zlib is default part of MacOS and most modern Linux distributions. The following commands should install the OpenMP runtime library:
- Ubuntu 18.04 and 19.10:
apt-get install libgomp1
- Oracle Linux 7 and 8:
yum install libgomp
- macOS:
libgomp
should be already installed
As of version 20.1.0 and later, GraalVM’s R runtime on Linux supports and bundles the GFortran version 3 runtime libraries. On macOS it bundles and supports the GFortran version 8.3.0 runtime libraries. It is not necessary to install the runtime libraries. However, note that a runtime library is only compatible with the GFortran compiler of that same library version or later.
On a Linux system, $R_HOME/bin/configure_fastr
can be used to check that the necessary libraries are installed, and if not, it will suggest how to install them.
Moreover, to install R packages that contain C/C++ or Fortran code, compilers for those languages must be present on the target system. The following packages satisfy the dependencies of the most common R packages:
- Ubuntu 18.04 and 19.10:
apt-get install build-essential gfortran libxml2 libc++-dev
- Oracle Linux 7 and 8:
yum groupinstall 'Development Tools' && yum install gcc-gfortran bzip2 libxml2-devel
- macOS
brew install gcc
Note: If the gfortran
executable is not on your system path, you will need to configure
the full path to it in $R_HOME/etc/Makeconf
, the FC
variable.
Search Paths for Packages
The default R library location is within the GraalVM installation directory.
In order to allow installation of additional packages for users who do not have write access to the GraalVM installation directory, edit the R_LIBS_USER
variable in the $GRAALVM_HOME/etc/Renviron
file.
Running R
Run R code directly or in the REPL mode with the R
and Rscript
commands:
R [polyglot options] [R options] [filename]
Rscript [polyglot options] [R options] [filename]
The GraalVM R runtime uses the same polyglot options as other GraalVM languages runtimes and the same R options as GNU R, e.g., bin/R --vanilla
.
Use --help
to print the list of supported options. The most important options include:
--jvm
: to enable Java interoperability--polyglot
: to enable interoperability with other GraalVM languages--vm.Djava.net.useSystemProxies=true
: to pass any options to the JVM; this will be translated to-Djava.net.useSystemProxies=true
.
Note: Unlike other GraalVM languages runtimes, R does not yet ship with a Native Image version of its runtime.
Therefore the --native
option, which is the default, will still start Rscript
on top of the JVM,
but for the sake of future compatibility the Java interoperability will not be available in this case.
You can optionally build the native image:
gu rebuild-images R
The native launcher for R is intended only for curious users and experiments.
There are known issues and limitations. Once the native launcher is built, you can use
the --jvm
flag to run R again in the JVM mode.
GraalVM Integration
The R language integration with the GraalVM ecosystem includes:
- seamless interoperability with other GraalVM languages and with Java
- debugging with Chrome DevTools
- CPU and memory profiling
- VisualVM integration
To start debugging R code, start the launcher with the --inspect
option:
Rscript --inspect myScript.R
Note: The GNU-compatible debugging using, for example, debug(myFunction)
, is also supported.