// the find
atilsamancioglu/PythonForDataScienceNotebooks
A stack of numbered Jupyter notebooks covering the standard numpy/pandas/matplotlib/seaborn progression, ending in basic feature engineering (encoding, balancing, EDA) with a couple of toy datasets bundled in. It's course companion material, almost certainly paired with the author's Udemy class, aimed at someone learning the pandas/numpy stack from zero.
The numbering makes the intended learning order obvious, and each notebook targets one concept (concat/merge, apply, box plots) instead of dumping everything into one giant file. Datasets are committed alongside the notebooks they're used in, so nothing breaks from an external download link going stale. The progression from raw numpy indexing through to encoding and class balancing is a reasonable on-ramp that most intro tutorials skip past too fast.
No README, no environment file, no pinned library versions — you're guessing at pandas/matplotlib/seaborn versions and hoping the API hasn't drifted since March 2025. There's no code outside the notebooks: no reusable functions, no tests, nothing to actually import, so it's read-only reference material rather than something you build on. Naming like '13-WineQT.csv' next to '13-EDAExercise.ipynb' works until you have to find something later; there's no folder structure separating data from notebooks. Last real content push was over a year ago with no indication anyone's maintaining it against newer pandas/seaborn breaking changes.