🎙️ Do your tests verify that your code doesn’t get slower as data grows?
Episode 282 of the Real Python Podcast covers two essential topics for Python developers:
🧪 Unit Testing for Performance (Big-O):
- How to write automated tests that validate algorithmic complexity
- Catching performance regressions before they reach production
tproftool: a profiler focused on specific functions
🐼 What’s New in pandas 3.0:
- New dedicated
strdata type (faster, cleaner) - Copy-on-Write (CoW): predictable default copying behavior
- Cleaner column-based operations
📰 Also covered:
- PEP 819/820/822 (new Python proposals)
- 8 versions of UUID and when to use each
python-fire: auto-generate CLIs from any Python objectgazetteer: offline reverse geocoding library
💡 Explanation in a nutshell#
A performance test is like quality control for your code’s speed: instead of checking that it gives the right result, it checks that it doesn’t become unacceptably slow as data grows. It’s an extremely valuable practice that few teams implement consistently.
More information at the link 👇
Also published on LinkedIn.
