r0f1/datascience

This repository is a curated reference list of Python tools and resources for doing data science work. It does not contain code to run, instead it is a collection of links organized by topic, pointing to libraries, tutorials, blog posts, and talks that the maintainer considers worth knowing about. The list covers a wide range of practical needs. Core tools like pandas (for organizing tabular data), scikit-learn (for machine learning), and matplotlib (for charts) are listed first, followed by sections on alternatives and extensions to each. For example, there are faster replacements for pandas such as polars and modin, tools for working with very large datasets that don't fit in memory, and GPU-accelerated options for heavy computation. Sections also cover Jupyter notebook tricks, environment management, extracting text from documents, and working with databases. Beyond the basics, the list branches into machine learning subfields: classical statistics, Bayesian methods, regression, clustering, neural networks, natural language processing, time series forecasting, and computer vision. Each section tends to mix well-known libraries with lesser-known but useful ones, along with links to talks or blog posts that explain how to use them. There are also sections aimed at making data science work more production-ready, including tools for testing data quality, monitoring model behavior over time, building web applications from notebooks, and running experiments in a reproducible way. A section on large language models covers libraries for working with modern AI text models. The list is actively maintained and broad in scope, making it a useful starting point or ongoing reference for anyone working in Python data science, whether just beginning or looking for tools in a specific area. The full README is longer than what was shown.