Skip to contents

Install rJava from source

Source: vignettes/install-rjava-from-source.qmd

The basics of using rJavaEnv are covered in the Quick Start Guide, which demonstrates how to install Java in one line of code. This tutorial will show you how to install rJava packag from source if you want to or have to.

Assume your project directory is currently in a temporary directory. Feel free to skip that, if you are already working in a desired project directory where you would like to install Java. In the example outputs below you will see paths that point to a temporary directory, but in a real project you would see your project directory instead.

project_dir <- tempdir()
setwd(project_dir)

1. Install Java

You can use detailed step-by-step process from the step-by-step vignette to download, unpack, install and link Java JDK of your desired version, but for simplicity, let’s use shortcut functions

Either quick install Java into current project directory:

java_quick_install(version = 21)

Note

You might also do java_home <- java_quick_install(version = 21), as this will save the path to the installed Java home directory into java_home variable for later use, but the function to set the build environment actually uses the JAVA_HOME environment variable set by java_quick_install() by default, so it’s not necessary.

Or you can simply do:

This will essentially do the same as java_quick_install(), but only sets the JAVA_HOME and PATH environment variables in the current R session, without modifying your .Rprofile file or copying/linking the installation folder into your project directory.

2. Set the environment for building rJava

Now simply use:

That is it, java_build_env_set will detect your platform and architecture and set the correct environment variables. If you want advanced control, you can use java_home to set custom JAVA_HOME path, or use where='project' to set the environment variables in your .Rprofile file in the current project directory. See ?java_build_env_set for details.

Once java_build_env_set() completes, it will print out OS-specific notes. In summary,

  • On Linux, you may need to install several system dependencies. For Debian/Ubuntu, this can be done with: bash sudo apt-get update && sudo apt-get install -y --no-install-recommends libpcre2-dev libdeflate-dev libzstd-dev liblzma-dev libbz2-dev zlib1g-dev libicu-dev && sudo rm -rf /var/lib/apt/lists/*

  • On Windows, Rtools is required. You can download it from https://cran.r-project.org/bin/windows/Rtools/.

  • On macOS, you’ll need the Xcode Command Line Tools. Install them by running: bash xcode-select --install

3. Install rJava from source

Now you are ready to install rJava from source:

install.packages("rJava", type = "source")

If you are on a Linux distribution with pre-configured Posit Package Manager (PPM) repository that serves pre-compiled binaries even if type="source" (or in a similar situation on Windows or macOS), you can use repos = "https://cloud.r-project.org" to force installation from source:

install.packages("rJava", type = "source", repos = "https://cloud.r-project.org")

And that is it. You have successfully installed rJava from source. If you restart the R session, the build-related environment variables will be reset. To use your newly installed rJava package just do:

library(rJavaEnv)
use_java(21) # or whatever version you installed
library(rJava)