The research machine
Users of the microdata schemes can choose to do their work on Statistics Denmark’s research machine – a server that can accommodate several separate projects.
Here you can find guides for set-up, login and use of Statistics Denmark’s research machine:
- Connection to Researcher servers at DST - Windows users
- Connection to Researcher servers at DST - Mac users
- Connection to Research Servers at DST - Linux Users
- How to change your password
- How to reset your password
- Connection to research servers at Statistics Denmark from abroad
In the below overviews of disk space (in Danish), you can see how much space your research project has taken up on the research servers in recent years.
On the research machine and hosted servers, An AI-based code assistance tool has been developed. The purpose of the tool is to help users of Denmark’s Data Portal with various programming tasks. The tool allows you to prompt a model that helps you write and understand a wide range of programming languages, including SAS, SPSS, Python, Stata, and R, which are the five languages supported on the research machine.
How to get access
- Log in to the research machine: The AI code assistance tool is located on the research machine. To gain access start by logging in to the research machine via remote.dst.dk.
- Open the AI-tool: Click on the icon named “AI-code help”, which is on the desktop.
- Log in to the AI tool: After opening the AI-tool, a page will appear asking for login information. Here you must use your project-IDENT (without @dstfse.local) and your Windows password.
Once you have done this, you will have access to the AI tool’s main page, and you can begin prompting. Please note that even though the tool runs locally on the research machine, you must not enter sensitive personal data or any other information that may be considered confidential. The tool is intended solely for code assistance.
The model underlying the tool is the publicly available qwen2.5-coder, produced by the company Ollama. If you want more information about the model’s functionality, we refer you to Ollama’s website.