The Software Side of Data Science (My Python Package Journey)
Hey guys, it's Matt Dancho here. Today, I'm sharing my experience upgrading a Python package for pandas and polars. This is important because it helps you understand better the "Software Side" of Data Science. I dive into the challenges of transitioning to Pytimetk version 1.0 plus, aiming for improved performance with the latest versions of Pandas and Polars. Join me as I navigate through code changes, failed unit tests, and bug fixes to keep the package up-to-date. Your feedback is valuable!
Pytimetk GitHub (Give Pytimetk a Star): https://github.com/business-science/p...
Pytimetk Docs: https://business-science.github.io/py...
Table of Contents:
00:00 Introduction
01:50 How we use Polars inside my Pytimetk Package
03:35 Transition to Polars 1.0
04:48 Testing with PyTest, Importance of Unit Tests
07:00 GitHub Actions: Testing different OS, Python, Polars, and Pandas Versions
08:40 Summary