Log in to DST

From Remote Desktop download to a running RStudio

Published

September 9, 2026

This page guides you from the very first login to having RStudio running and being ready to execute code.

This guide assumes you are working with R. When you log in to DST, you choose “Python/R” - select R/RStudio.

Connect to the server

Step 1 - Download Microsoft Remote Desktop (Mac only)

Open the App Store on your Mac and search for “Microsoft Remote Desktop”. Download and install for free.

Windows users have Remote Desktop built in - search for “Remote Desktop Connection” in the Start menu.

Step 2 - Log in

  1. Go to remote.dst.dk in your browser
  2. Log in with your ident (4 letters) and the one-time code you receive
  3. Click Python/R
  4. Log in with username dstfse\IDENT+PROJECTNUMBER (e.g. dstfse\abcd708421) and your chosen password

You can also choose “FSE Windows” instead of Python/R - from there you land on a selection desktop and can click through to e.g. Python/R.

Step 3 - Find/create your own project folder

Open File Explorer and navigate to the workspaces folder under your project. Create a folder with your name or ident if it does not already exist:

E:/workdata/[projectnumber]/workspaces/[your-name]/

Recommended structure within your folder:

[your-name]/
  [your-project-name]/
    R/              ← all your scripts (.R files)
    datasets/       ← intermediate results (.rds files, local to DST only)
    output/         ← tables and figures ready for repatriation

Step 4 - Mac keyboard fix (do this at every connection)

Special characters such as `, $, \, @, { and } do not work correctly in Remote Desktop by default on Mac.

Solution: Switch to Unicode mode with ⌃⌘U (Ctrl + Cmd + U) inside the Remote Desktop window.

Character Mac shortcut in Unicode mode
Pipe \| Option + i
[ Option + 8
] Option + 9
{ Option + Shift + 8
} Option + Shift + 9

In RStudio, code is run with Ctrl+Enter (not Cmd+Enter).

DARTER-specific setup (project 708421)

Project 708421 uses fastreg (on CRAN) to read the registers with read_register("name") - just like the general guide.

See the complete guide: DARTER - overview and pipeline →

Create your R project

An R project collects all your scripts, output files and settings in one folder. Each study should have its own project.

What does a project actually do? A project is first of all just a folder - you still save your own files in it (in R/, datasets/, output/). The point is the .Rproj file: when you open the project, RStudio automatically sets the working directory to the folder where .Rproj lives. So getwd() always points to your project folder, and you can use relative paths (datasets/cohort.rds) instead of writing the full path or setwd("E:/...") at the top of every script. Each project also keeps its own settings and history separate, so studies don’t get mixed up.

  1. Open RStudio
  2. File → New Project → New Directory → New Project
  3. Give the project a name and choose a location under your workspace folder
  4. Click Create Project

Run getwd() now and you get the path to your project, e.g. E:/workdata/[projectnumber]/workspaces/[your-name]/[your-project-name]

Do not save workspace RStudio asks at close: “Save workspace image?” - always choose No. Set it permanently: Tools → Global Options → General → “Save workspace to .RData” → Never.

Make RStudio start faster on DST Disable the packages panel: Tools → Global Options → Packages → uncheck “Enable packages pane”. You use library() and install.packages() directly in the code anyway.

The rules you sign up to, in short

You certify to these before you get access, and breaking them carries sanctions. The four that catch people out in ordinary daily work:

  • No images of the screen: the ban covers photos and screenshots alike, so you cannot snap a console output to paste into a message to a colleague.
  • Nothing leaves except through repatriation: not by email, not on paper, not by retyping it. See Phase 14 - Export and repatriation.
  • Never put real data in a support question: when you write to a data manager or to DST, invent the numbers. A pasted error message with a real pnr in it is a rule breach.
  • Work somewhere private: your institution or your home office. Not a train, a café or a library, and never leave the screen unattended. Screen sharing is allowed only with people who have access to that same project.

The full text is short and worth reading once: Rules for working with microdata →.

Log off correctly when you are done

Closing the RStudio window or simply disconnecting is not enough. You are sitting on two desktops nested inside each other plus a browser session - close all three, otherwise your session keeps running and holds the server’s RAM (see Mind the RAM).

Log off in this order:

  1. Sign out of the Python/R desktop (the inner desktop where RStudio runs) - use Sign out in the Start menu, not just closing the window.
  2. Sign out of the FSE Windows desktop (the outer choice desktop) - Sign out again. (skip this step if you went straight to the Python/R desktop)
  3. Log out of remote.dst.dk in your browser.

So: Sign out twice (once per desktop) plus log out in the browser.

Next steps

You are logged in and set up. Next is the data itself: which files you meet, and what opens them.

→ Phase 4 - File types and loading

Project 708421: DARTER - Register paths and datastores

Back to top