Computational Modelling Group

Test and Rest

Started
1st March 2014
Research Team
Evander DaCosta, James Graham, Oliver Laslett
Investigators
Hans Fangohr

Example output from running tests using the py.test tool.

Computational modelling has a number of errors, including model errors, approximation errors, numerical errors, and coding errors. This projects is concerned with the last.

How can we be (as) sure (as possible) that we have not got any bugs in our simulation codes? How can we ensure that we do not introduce bugs as the code matures, develops and changes? How can this be maintained despite new people being added to a project and the original author moving on (as is often the case in research)?

This project and interest group tries to establish, try and offer training to others. The main tool to be studied is the creation and use of regression and system tests, combined with tools (such as Jenkins and Vagrant) that allow the automatic execution of this tests, in a number of different environments.

Please contact Hans Fangohr if you are interested in joining this activity.

Categories

Software Engineering Tools: CVS, Git, Jenkins, Mercurial, Sourcesafe, SVN

Transdisciplinary tags: Complex Systems, e-Research, NGCM, Scientific Computing, Software Engineering