Setting up Julia

sdbuildR uses Julia for ensemble simulations and working with units in models. This is supported by the R package JuliaConnectoR. This guide will help you install Julia and configure it to work with sdbuildR.

Step 1: Install Julia

Download and install Julia from https://julialang.org/install/. If you already have Julia installed, you can go to step 2.

We recommend using juliaup, Julia’s official version manager, which makes it easy to install and switch between Julia versions. Some useful commands are listed below.

# See installed versions and current default
juliaup status

# Install latest stable version
juliaup add release

# Install specific version
juliaup add 1.11.3

# Set default version
juliaup default 1.11.3

# Update juliaup and all Julia versions
juliaup update

# Remove a version
juliaup remove 1.10.0

# List all available versions
juliaup list

Alternatively, you can download a standalone Julia installer from https://julialang.org/downloads/.

Step 2: Check if Julia is accessible

After installing Julia, check if R can find it:

# Check if Julia is on your PATH
Sys.which("julia")

# Check if JULIA_BINDIR is set
Sys.getenv("JULIA_BINDIR")

# Verify JuliaConnectoR can find Julia
JuliaConnectoR::juliaSetupOk()

If both Sys.which("julia") and Sys.getenv("JULIA_BINDIR") return an empty string (""), you’ll need to add Julia to your PATH. If the Julia installation can be found, try to start a Julia session:

# Verify JuliaConnectoR can start Julia
JuliaConnectoR::startJuliaServer()
JuliaConnectoR::stopJulia()

If this runs successfully, you can skip to step 6. If this throws an error, go to the next step.

Step 3: Find your Julia installation path

You need to locate the bin directory of your Julia installation (not the Julia executable itself).

From Julia

If you can start Julia from your terminal/command prompt, run this inside Julia:

juliaBase.julia_cmd()[1]

This returns something like:

• Windows: "C:\\Users\\YourName\\.julia\\juliaup\\julia-1.11.3+0.x64.w64.mingw32\\bin\\julia.exe"
• macOS: "/Applications/Julia-1.11.app/Contents/Resources/julia/bin/julia"
• Linux: "/usr/bin/julia"

Important: You need the bin directory containing the executable (remove julia.exe or julia from the end).

From your terminal/command prompt

Windows:

where julia

macOS/Linux:

which julia
# or
whereis julia

Take note of the path, removing the filename to get just the bin directory.

Step 4: Add Julia to your PATH permanently

To make Julia accessible to R across all sessions, add it to your .Renviron file.

Open .Renviron

Run this in R to open your .Renviron file:

# Install usethis if needed
if (!require("usethis")) install.packages("usethis")

# Open .Renviron for editing
usethis::edit_r_environ()

This will open .Renviron in your text editor. If the file doesn’t exist, it will be created.

Add Julia to PATH

Add one of the following lines to .Renviron, replacing the path with your actual Julia bin directory:

Windows (use forward slashes):

PATH="C:/Users/YourName/.julia/juliaup/julia-1.11.3+0.x64.w64.mingw32/bin;${PATH}"

macOS:

PATH="/Applications/Julia-1.11.app/Contents/Resources/julia/bin:${PATH}"

Linux:

PATH="/usr/bin:${PATH}"

If using juliaup (recommended), you can point to the juliaup directory instead of a specific version:

Windows:

PATH="C:/Users/YourName/.julia/juliaup;${PATH}"

macOS/Linux:

PATH="~/.julia/juliaup/bin:${PATH}"

This allows juliaup to manage which Julia version is used.

Save and restart R.

1. Save the .Renviron file
2. Restart R (Session → Restart R in RStudio, or close and reopen R)

Step 5: Verify the setup

After restarting R, verify that Julia is accessible:

# Should show the path to julia executable
Sys.which("julia")

# Check Julia version
system("julia --version")

# Verify JuliaConnectoR can find Julia
JuliaConnectoR::juliaSetupOk()

If Sys.which("julia") returns a path and juliaSetupOk() returns TRUE, R can find your Julia installation!

Check whether a Julia session can be started from R:

# Verify JuliaConnectoR can start Julia
JuliaConnectoR::startJuliaServer()
JuliaConnectoR::stopJulia()

Troubleshooting

Julia still not found after restarting R

1. Double-check the path you added to .Renviron
2. Make sure you’re pointing to the bin directory, not the Julia executable
3. On Windows, use forward slashes (/) not backslashes (\)
4. Verify you saved .Renviron and restarted R

View your current PATH

To see all directories in your PATH:

# View PATH split by separator for readability
strsplit(Sys.getenv("PATH"), .Platform$path.sep)[[1]]

Look for your Julia bin directory in this list.

Advanced: Using multiple Julia versions

If you have multiple Julia versions installed, the first one found in your PATH will be used by default.

Switch versions with juliaup

If using juliaup, you can change the default version without modifying .Renviron:

# From terminal/command prompt
juliaup default 1.11.3

Step 6: Using Julia with sdbuildR

To set up the Julia environment for sdbuildR, including all necessary packages, run use_julia(). First time use can take quite a while (i.e., 10-30 minutes).

library(sdbuildR)

use_julia()

The Julia environment needs to be set up for each new R session, but should now be much quicker.

Using a specific version with sdbuildR

By default, sdbuildR uses your default Julia version. To use a different version:

use_julia(JULIA_HOME = "C:/Users/YourName/.julia/juliaup/julia-1.11.3+0.x64.w64.mingw32/bin")

Note: This only affects the current R session and is not permanent.

Close Julia session

To close the Julia session:

use_julia(stop = TRUE)