Python data science tools plus the statistics papers to use them right

What it does This repository is a curated index of resources for practicing data science in Python. It groups libraries, tools, blog posts, tutorials, and academic papers into categories ranging from core libraries and big data frameworks to Jupyter utilities and classical statistics. Think of it as a map to the ecosystem rather than a single piece of software.

The interesting bit Most awesome-lists stop at software catalogs; this one treats statistical rigor as part of the stack. It includes entire sections on p-value fallacies, confidence interval misinterpretations, and domain-specific datasets alongside the usual DataFrame alternatives and GPU accelerators. The list acknowledges that a faster groupby is useless if your inference is broken.

Key highlights

• Covers the full pipeline: core libs like pandas and scikit-learn, alternatives like polars and duckdb, big-data tools like spark and dask, and GPU options like cuDF.
• Includes operational Jupyter tooling: papermill for parameterizing notebooks, nbdime for diffing, and voila for turning notebooks into standalone web apps.
• The Classical Statistics section links to full books (e.g., Lakens on statistical inferences), papers on abandoning p-values, and specialized R datasets for medicine, crime, and education.
• Practical utilities are not forgotten: ruff for linting, uv for dependency management, rainbow-csv for VSCode, and handcalcs for writing equations in Jupyter.
• Even has niche extraction tools like docling and DeepSeek-OCR for document parsing.

Caveats

• It is one massive, flat markdown file with minimal editorializing beyond grouping; expect to do your own filtering.
• Many entries receive only a one-line description, so the list tells you what exists but not always why you should pick one tool over another.
• The README source is truncated in places, leaving some sections partially obscured.

Verdict Grab this if you are a Python data scientist looking for a broad orientation to the ecosystem or a reminder that statistical theory matters as much as the latest DataFrame accelerator. Skip it if you need a guided tutorial or a ranked product comparison; this is a library card catalog, not a syllabus.