Question: liveProject: How to Think about Manipulating Data -- DataFrame

[Alex Khvatov]

Thank you for answering my previous question.
I have another one:

If pandas library is single threaded, working partitioned data would not bring any benefits; the data would still be accessed sequentially and not in parallel manner, right?

Basically I am trying to find a good use case to use pandas vs Spark.

Regards,
Alex

[A. Bell]

If you have a small-ish dataset then pandas is just as good to use -- small-ish probably means that it should all be able to fit into memory.