powerfulpython.com

A life changer

I'm half-way through the Powerful Python Academy material. I'm a mid-career geoscientist with no comp-sci background but a passion for reproducibility and documentation. When I started the program, I only had taken Python for Everybody (Coursera), and I had watched a few Scipy conference tutorials.I always wanted to know how Python developers were able to write such amazing and extensive libraries, but I couldn't find the right fit. Most material out there is either too broad (books) or too narrow (blog post). PPA offered an opinionated approach to software development in Python, with lots of exercises and direct feedback from Aaron (PPA creator). In this case, opinionated is an excellent approach as it guides you to learn what matters to get your job done, without worrying about obscure methods or practices that are almost never used.The academy moved me from Python enthusiast to Pythonista. From the individual classes in PPA, I value: OOP: I learned the principles of OOP software design. Now I can read other people's codebase and be able to understand what they are doing and why (most of the time). I've also rewritten some of my work using these principles, and the software clarity and brevity gained is unbelievable compared to the way I wrote them before knowing OOP, i.e. using functions only (procedural programming). Before the class, I had very little knowledge of OOP. Knowing OOP is crucial in Python, and this class lays the foundations clearly and succinctly, which is perfect, just focusing on the critical aspects. I learned about inheritance, dunder methods, member variables, and methods, how and when to use @staticmethod and @classmethod, separation of concern, access control, setter, and getter methods, check class arguments on instantiation... TDD: Untested code is broken code. The biggest realization for me in this class was that to build or contribute to a large codebase, one must rely on tests. Also, tests will save us from embarrassing mistakes in simple calculations, so whether big or small, I learned to always test my code. Scaling with generators: Here lies a key to memory-efficient code called scalable composability. It was nice to learn the syntax for most common generator types, but the real money is in learning how to use them to prevent your code from freezing by leveraging lazy evaluation and composability. It also really helps to see how a Python expert developer writes code, as he walks you through the concepts and reasons to write in such way, e.g. think of the many ways one can read a file into a Python object, which one is best? Why?If you are looking to up your Python game, and like I got frustrated trying to understand how things work just by reading the docs, please do yourself a favor a set the time to go over PPA. It will change your life.

Date of experience: Jun 23, 2020