Let me tell you about Pandas DataFrames. DataFrames are wonderful pythonic objects that support clever programming and fast execution via numpy. DataFrames can be masked, joined, manipulated, and plotted however you want. DataFrames rock, and Python is fun.
Unfortunately, real businesses run on SQL. Now, my workflow as a data scientist would be to load all my data into memory and work on in Pandas, and if I don't have enough memory, start a bigger instance on AWS. But sometimes you have to use a 70s vintage DBMS, and that means knowing SQL.
Forta's book is a solid introduction to SQL concepts, and how to do selects, filters, and joins. Your exact database will vary, but I find this book to be clear written, a good place to start, and a worthwhile addition to your knowledge base.