By Will Usher, Senior Researcher: Infrastructure Systems Modeller, University of Oxford
Plagiarism is a serious issue, and we are all familiar with the horror stories of students unceremoniously ejected from courses for copying essays. Any undergraduate degree worth its salt teaches students how to cite work correctly, acceptable bounds on quotation and how to attribute ideas and concepts to their sources. But in the growing world of open-source research software, best practices have yet to be universally understood, as I recently found out.
During my PhD at University College London, I became involved in the heady enthusiasm of the Research Software Programming group, attending and then helping out at Software Carpentry workshops. As a consequence, I was keen to apply my new knowledge of Python, version control and software development to my research. As luck would have it, I discovered an existing Python library on Github, which implemented several Global Sensitivity Analysis routines I could make use of. As I used the library, I started adding bits and pieces, and so by the end of the PhD I had made a considerable contribution to the package.
It's probably safe to say that SALib (sensitivity analysis library) is the go-to Python library for the unfortunately still-far-too-niche use of global sensitivity analysis in modelling, and our…Continue Reading