Better course for all
Limited to 20 participants
Software Engineering for Physicists (CANCELED)
Pranav Sampathkumar (KIT)
Abstract
Knowledge transfer of existing code bases and scripts is an essential part of scientific endeavours. As scientists create more and more open code bases to share and collaboratively develop, it becomes very important we learn software engineering fundamentals so that our development paradigm aligns with industry standards, which allows for more maintainable and transferable code. It helps a lot in reducing onboarding and offboarding time from projects and speeds up scientific development.
In this course, we will present strategies to use well-known tools in a "clean" manner, that are based on current industry standards which can allow for creation of code bases which live well beyond the work times of individual contributors. The topics we cover include
- Basics of Software Engineering: Technical Debt, Code Rot/Entropy, Development vs Engineering, Code vs Software.
- Git: How to structure a repository, How to create good commits and have a useful history, How to organise a project on GitHub and other integration with other collaboration tools.
- Programming paradigms and styles: Test/Behaviour driven development, Signs to look out for in bad code (code "smells") and how to fix them, rules of thumb for neat code.
- Strategies in Software Development Life Cycle (SDLC) Eg: Waterfall model, Continuous Integration, Agile
- Tools to help have a better programming structure (Formatters, Linters etc.)
- Hands-on Session in some of these principles.
Most of this course will be taught with Python as examples so a working knowledge of Python is expected. Basics of Git are also recommended as the course deals only with meta-level strategies and not the basics.
