How data scientists can incorporate testing into their workflow, with Hypothesis for code and Great Expectations for data.
Code written in a programming language is the most flexible way to accomplish your data science work, and every line we write is technical debt. In this PyData Ann Arbor talk, I use two worked examples to show how testing fits into a data scientist's workflow, and introduce two libraries that help: Hypothesis for property-based testing of your code, and Great Expectations for validating your data. You'll leave with a practical sense of which tests to write and when.
Slides are available here.