Learn Python by Building Data Science Applications

Bringing your own package

Most of the packages that we've taken a look at so far are very generic—for instance, requests works with HTTP on the client side, while pandas is focused on data analysis. Why should you write your own package with a very specific task in mind that's not necessarily useful or applicable to another person's use case? Simply put, because you'll be the first to benefit!

The main goal that packages solve is to create a deliverable; any new user on a new machine will be able to install the package, as well as all the packages it requires to work properly, at once. As a byproduct, packages remain isolated from the code you write every day. This isolation is a good thing—it makes you, as a package designer, build your ...