Beyond Pandas: Spark, Dask, Vaex and other big data technologies

Very thorough analysis, and very impressive perf by vaex, with simple (i.e. pandas-like) API.

API and performance comparison on a billion rows dataset. What should you use?