Each week it seems as if there are new articles in Nature or Science about the reproducibility crisis. While modern statistical and software development tools make reproducible research within anyone’s reach, most ecologists are not exposed to software engineering workflows. Yet ecologists can assemble a reproducible workflow with tools like R, and git.

git was developed to manage the Linux kernel—17 million lies of code and a few thousand developers. Few ecologists are involved in such big projects. Even if you are a solitary researcher, a reproducible workflow offers immense benefits to one very important collaborator—your future self! Having a repeatable workflow allows you to return to- and re-engage with- your analysis at any point. In addition to collaborating with your future self, having a reproducible workflow will: 1) enhance any manuscript you submit; 2) allow you to share your analysis with other scientists; and 3) contribute to and build your lab’s knowledge base.

Participants will learn how to use git to support all phases of an R-based analysis, including: data cleaning, data storage, exploratory data analysis, documentation, modeling, and presentation—including static, print, and interactive web-based outputs. We will learn what git is, what it is capable of, how take advantage of its branching model, and how to connect to public code repositories like GitHub and BitBucket. Participants will learn how to use MAKE files to support a workflow. Participants will also learn how to prepare a worked example designed to accompany a submitted, or published manuscript.