Vincent Warmerdam is a data scientist and Python educator known for exploring practical data tools, developer workflows, and thoughtful software design. He regularly writes about Python, machine learning, tooling experiments, and lessons learned from real-world data work.
9 articles from this blog
A technical benchmark comparing Python, NumPy, Numba, Rust, and Mojo for simulating permutation cycles, with corrected results.
Explores performance optimizations for scikit-learn's GridSearchCV by using closed-form solutions and warm starts for specific linear models.
Using zsh's compdef to enable file autocompletion for Python command-line tools like marimo via uvx.
Testing a prompt technique inspired by 'The Office' to get more concise and detailed AI-generated explanations of technical concepts like Huber regression.
A technical guide explaining how to use wget with recursive options to download entire websites for offline viewing, including a breakdown of key command-line flags.
Benchmarking diskcache with zlib compression for JSON data in Python, showing potential disk space savings for text-heavy caches.
Using Ruff to ban SQLAlchemy dialect-specific imports, ensuring database portability in Python projects.
TaskiPy is a Python tool that automates tasks by defining them in pyproject.toml, serving as a Makefile alternative for project workflows.
A Python developer discusses their new LLM course, ergonomic keyboards, the diskcache project, and coding agents on the Talk Python podcast.
