packages like testthat has been really useful for writing packages, but oftentimes in data science there is no need to write a package for each project. It will greatly enhance data science projects' reproducibility if we have a package that set up data science project directories, test data and code, and ensure that the end results can always be generated from the raw data using code in the project set up.
You may be interested in looking into the {drake} or {targets} packages. The do not set up a data science projects for you, but they allow you to create data science projects that are reproducible, scalable, and inherently function-oriented (which means they share a lot of similarities to designing packages).
1 Like
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.