Computational Workflows

Doctoral Programme in Science and Engineering (DSSE)

Practicalities

Instructor: Jack S. Hale (jack.hale@uni.lu). Office hours: 1500-1600 Wednesdays (by appointment).

Course Room: MNO 1.030.

Date: Thursday 10th November 2022 0900 - 1200 and 1315 - 1700.

Description

This one day course covers the basics of Computational Workflows for Scientific Computing. During the course we will discuss acheivable best practice for performing Scientific Computing, and we will discover some tools (containers, unit testing and continuous integration) that can help.

Outline

• Introduction to Good Practice in Scientific Computing.
• Specifying and running software anywhere with Docker.
• Unit testing software with Python and pytest.
• Continuous integration using GitHub Actions.

Useful information

• Assumed knowledge: basic git, basic Python, basic Linux administration (shell, installing software from package managers etc.). Please refresh your knowledge prior to the course if you are feeling unsure.

• The Etherpad. We will use the Etherpad for technical questions and for sharing information during the course. The Etherpad is public, please use a pseudonym and do not reveal any personal information.

Prior to the day of the course

1. Please download and read the paper Good enough practices in Scientific Computing. We will use and discuss the paper during the class.

2. Install Docker on your personal computer following the instructions here. Windows, Mac and Linux are supported. Please verify that you can run the following command before the day of the course:

   docker run -it hello-world
   

   If you are having difficulties please contact me via the Etherpad.

3. See if you have the version control system git installed:

   git --version
   

   If not, then please install it following the instructions here.

4. Get an account at GitHub.

Assessment

A coursework will be distributed at the end of the class. To pass the course and receive the ECTS credits you must complete the coursework.