This is a follow up on the blog post on pandas, Polars and Dask. In this post we look at some of the libraries that either build on top of pandas, Polars and Dask or they sit underneath them. This starts from distributed task libraries like Ray to the wrapper nature of Modin and to the massive infrastructure from Spark.

We are looking at the three most common dataframe libraries for Python: Pandas, Polars, and Dask. We compare them on an architectural level, i.e. which one do we use for which situation and where do they differ in concept.