Using literate programming is a widespread practice amongst data scientists. This practice not only encourages data scientists to produce transparent, rich and reflective accounts of their analysis without the extra overhead of switching between tools, but also leads to artefacts (i.