Applied AI Tools for Scientific Programming
Description
AI coding assistants can accelerate learning, improve code quality, support debugging, and streamline scientific programming. This lecture provides an overview of how these tools can be integrated into research workflows. I will demonstrate GitHub Copilot (free for academics) inside Positron, a modern data-science editor developed as the successor to RStudio. Examples will use R and Python, but the concepts apply equally to other editors such as VS Code and are independent of the programming language.
Content
- Core principles and capabilities of AI-assisted programming
- Effective use of code completions and context-aware suggestions
- Chat-based interactions with scripts and objects for writing and debugging code
- Benefits, limitations, and considerations for responsible use
Links
Positron
In general, the Positron webpage is very helpful and contains a lot of useful information. Below are some direct links to get started.
- Download Positron
- Positron Guides to help you get started or migrate from other IDEs like RStudio
- Setup guide for GitHub Copilot in Positron
GitHub Copilot
- Step-by-step guide on how to get Copilot for free for Visual Studio Code or RStudio.
- GitHub Copilot Docs: Useful information and guides on how to use Copilot
- Privacy FAQ of GitHub Copilot
Guidelines, Privacy, Limitations
- A blogpost on data privacy and AI assistants in Positron
- Guideline for Dealing with Artificial Intelligence by Berlin Universities Publishing
- FU KI website with more information and guidelines on AI use at FU Berlin (but not specifically about coding)
- TU Berlin on AI: Information about AI in general but also AI tools in research (not specifically about coding)