Software Setup

Get your laptop ready before Day 1. Plan 30–60 minutes the day before — not the morning of. The optional Day 0 session is the place to bring any install problems you can’t solve yourself.

Important

Bring your own laptop. The workshop is hands-on — every participant works on their own laptop (at least 16 GB RAM and ~10 GB free disk). There are no lab machines, so set everything up below before you arrive.

Tip

What you’ll have at the end:

  1. R and RStudio installed and working
  2. VS Code and Quarto installed
  3. A working terminal (a POSIX shell)
  4. The command-line build tools your operating system needs to compile R packages

That’s all you need to do on this page. Installing the R / Bioconductor packages happens in Module 00 — see Refresh your skills (§4) below.

1. Required tools

  • R ≥ 4.4 — download
  • RStudio Desktop (current release) — download
  • VS Code (current release) — download
  • Quarto (current release) — download (the workshop tutorials are Quarto .qmd notebooks; RStudio bundles a copy, but install the standalone release too so VS Code and the command line can render)
  • LaTeX — install the lightweight TinyTeX distribution by running quarto install tinytex in a terminal once Quarto is installed (needed only if you render notebooks to PDF; HTML output needs no LaTeX)
Note

What each of these actually is. It helps to know which tool does what:

  • R and LaTeX are scripting/markup languages — you write code in them and a program executes it. R turns data-analysis code into results and figures; LaTeX turns markup into a typeset document (it’s what produces PDF output).
  • Quarto is a formatting and publishing system — it weaves your prose together with the code chunks (run through R) and renders the result into a finished HTML page, slide deck, or PDF. It is the layer that turns a .qmd notebook into the pages on this site.
  • RStudio and VS Code are Integrated Development Environments (IDEs) — the workbenches you actually sit in. They give you an editor, a place to run code and see output, and tools to manage projects. VS Code can also connect over SSH to remote clusters such as Talapas — that connection is covered in Chapter 09 — VS Code & SLURM Basics on Talapas. For this setup page, install VS Code on your local laptop only.

In short: R and LaTeX are languages, Quarto stitches and publishes them, and RStudio / VS Code are where you do the work. During the workshop you’ll use them locally (Wed–Thu); Friday’s Module 09 adds the Talapas Remote-SSH connection.

2. Find or install a terminal

The command-line portions of the workshop assume a POSIX (Unix-style) shell. macOS and Linux ship one; Windows does not, so Windows users install one. Set this up once for your operating system. The command line itself — navigating, pipes, grep — is covered in the P1 reading (Computer Systems & the Command Line) and Appendix C, and the optional Day 0 session walks through it.

macOS — you already have one. Open the built-in Terminal app (Applications → Utilities → Terminal, or press ⌘-Space and type “Terminal”). It runs zsh, a standard POSIX shell — nothing to install. If you’d like a nicer one, iTerm2 is a popular free upgrade, but the built-in Terminal is all you need.

Windows — Windows’ built-in Command Prompt and PowerShell are not POSIX shells, so you install a small Linux environment with WSL2 (Windows Subsystem for Linux) and run a real Linux terminal inside it. Two ideas to keep separate:

  • WSL2 is the machinery. It’s a lightweight Linux virtual machine built into Windows. You don’t interact with WSL2 directly — it’s the engine that runs a Linux distribution for you.
  • Ubuntu is the Linux you actually use. When you install WSL2 you also install a distribution (we use Ubuntu), and Ubuntu is what gives you the terminal, the bash shell, and Linux commands.

To install both at once:

  1. Open PowerShell as Administrator (right-click the Start button → Terminal (Admin) or Windows PowerShell (Admin)).
  2. Run wsl --install -d Ubuntu and press Enter. This turns on WSL2 and downloads Ubuntu. Restart when prompted.
  3. After the restart, launch Ubuntu from the Start menu (search “Ubuntu”). The first launch finishes the install and asks you to choose a Linux username and password — these are separate from your Windows login. The window you’re now in is your terminal for the rest of the workshop.
  4. Update the Ubuntu packages once: sudo apt update && sudo apt upgrade -y (it will ask for the password you just set).

From then on, opening “Ubuntu” from the Start menu is how you open the terminal. Full Microsoft guide: https://learn.microsoft.com/en-us/windows/wsl/install.

TipSeeing your Windows files from the Ubuntu terminal

WSL2 and Windows share files, so you don’t need two copies of anything.

  • Your Windows drives are mounted under /mnt/. Inside the Ubuntu terminal, your C: drive is /mnt/c, so your Desktop is /mnt/c/Users/<YourWindowsName>/Desktop and Downloads is /mnt/c/Users/<YourWindowsName>/Downloads. For example:

    cd /mnt/c/Users/<YourWindowsName>/Desktop
ls

  • Going the other way: your Linux files appear in Windows File Explorer under the Linux entry in the left sidebar (or by typing \\wsl$ in the address bar), so you can open them with Windows apps too.

  • Where to keep workshop files: working inside the Linux home directory (~, i.e. /home/<your-ubuntu-name>) is noticeably faster than working under /mnt/c. A good habit is to keep the project in your Linux home and reach over to /mnt/c only when you need a file that lives on the Windows side.

Linux — you already have one. Open your distribution’s built-in Terminal application; it runs bash (or a similar POSIX shell). Nothing to install.

3. Install tools for compiling

Several R / Bioconductor packages compile from source the first time you install them, which needs a working compiler toolchain. Install it once for your operating system; if a package later fails to build, the FAQ lists the usual fixes.

macOS — install Apple’s Xcode command-line tools from the Terminal you opened in §2:

xcode-select --install

This provides clang, git, make, and the headers R needs to build packages. Some packages also need a Fortran compiler — if you later hit gfortran not found, install gfortran for macOS. Re-launch RStudio after it finishes.

Windows — install Rtools, the toolchain that lets native Windows R build packages from source, from https://cran.r-project.org/bin/windows/Rtools/. Pick the version matching your R (e.g. Rtools44 for R 4.4). Rtools is for the Windows R/RStudio you installed in §1; it is separate from the Ubuntu environment in §2 (Ubuntu compiles with build-essential, below, if you ever run R there).

Linux — install the build toolchain from your package manager. On Ubuntu/Debian (including WSL2 Ubuntu):

sudo apt update && sudo apt install -y build-essential gfortran

build-essential provides gcc, g++, and make; gfortran covers packages with Fortran code. On Fedora/RHEL, use sudo dnf groupinstall "Development Tools" plus gcc-gfortran.

4. Refresh your skills

New to the command line or R — or just rusty? The optional Tuesday (Day 0) session brings you up to speed on your local laptop, and is where we install the R / Bioconductor packages together. Two modules on the Materials page:

Important

Can’t make the Tuesday session? Work through the two readings above, then open Tutorial 00 — Get Up and Running and run its package-install block before Day 1. The first install can take 20–60 minutes, so do it the day before — not the morning of.

Something didn’t work?

The FAQ documents the issues that come up most often: Mac compiler errors, BiocManager mirror selection, RStudio caching old library paths, etc.

Back to top