Summary

With Jupyter, Spark, and ODBC as some of the most ubiquitous utilities in data science, it only makes sense to cover Arrow from the perspective of its integration with these tools. Many of you will likely not use Arrow directly in these cases, but rather benefit from the work being done by others utilizing Arrow. But, if you're a library or utility builder, or just want to tinker a bit to see whether you can improve the performance of some different tasks, this chapter should have given you a lot of information to chew on and hopefully a bunch of ideas to try out, such as converting Arrow on the fly to populate an Elasticsearch index but maintain a consistent interface.

I don't want to give you all the answers, mostly because I don't have them. There's a wealth of people all over experimenting with Arrow in a large number of different use cases, some of which we'll cover in other chapters. Hopefully, this chapter, and the chapters to come after it...