| 📅 Today’s Picks |
Build Faster Tests with pytest Session Fixtures
Problem:
pytest fixtures provide reusable test data, but they reload for every test function by default.
When your fixture loads a large DataFrame, every test reloads the same data, wasting time and delaying your development workflow.
Solution:
Session-scoped fixtures load data once at the start and reuse it across all test functions.
Apply this pattern to:
- Load large datasets once instead of reloading for each test function
- Share a database connection across all tests without passing it as a parameter
- Automatically set random seeds for reproducible train/test splits
|
latexify: Turn Python Functions Into Clean Math Formulas
Problem:
It is not ideal to present mathematical formulas written in Python code to executives and stakeholders as they are often not familiar with Python code.
However, writing LaTeX manually to show the formulas is time-consuming and tedious.
Solution:
latexify transforms Python functions into clean mathematical notation with a single decorator. No manual LaTeX required.
Key features:
- Automatic LaTeX generation from Python functions
- Functions remain executable for calculations
- Compatible with various notebooks such as Jupyter, Colab, and Marimo
| ☕️ Weekly Finds |
ty
Python Utils
An extremely fast Python type checker and language server, written in Rust
giotto-tda
ML
A high-performance topological machine learning toolbox in Python built on top of scikit-learn
vibekit
MLOps
Run Claude Code, Gemini, Codex — or any coding agent — in a clean, isolated sandbox with sensitive data redaction and observability baked in
