People use Jupyter Notebook for lots of different needs; it's a great tool for exploration, especially when learning new APIs as well as debugging APIs. A typical workflow to follow when working with new APIs:
This process includes reviewing error messages, the documentation, updating the code, and trying again. Time may be wasted trying this more than once before admitting. That's when one realizes what should have been done first: get the API working from a Jupyter Notebook. That takes a lot less time, resulting in a much deeper understanding of how the API works, then copying and pasting working code into my program. Best of all, that worksheet can be saved for future reference.
We'll show a few examples of doing this and answer questions about this methodology.