Datasets¶
Registry of datasets used across the course. Small datasets are committed here; large datasets are downloaded via script or linked, never committed to Git.
Registry¶
| Dataset | Used in | Source | License / notes |
|---|---|---|---|
| Penguins | EDA case study (W8), visualization | seaborn built-in (sns.load_dataset("penguins")) |
CC0 (Palmer Station LTER) |
| Tips | Matplotlib/Seaborn demos | seaborn built-in | CC BY (Bryant & Smith) |
| Flights | Time series / pivot demos | seaborn built-in | public domain |
| Iris | Classification demo (k-NN, trees) | sklearn / UCI | public domain |
| Titanic | Cleaning, classification lab | seaborn built-in / Kaggle | CC BY-SA / Kaggle rules apply |
| California housing | Regression (Lab 18) | sklearn.datasets.fetch_california_housing |
public domain (StatLib) |
| Student records (dirty) | Cleaning & EDA (Assignment 1) | generated by scripts/make_assignment1_dataset.py (seed 2026) |
synthetic; CC0 |
| Mall customers (or similar) | Clustering (Lab 21) | Kaggle | Kaggle rules apply |
| Weather (Open-Meteo archive) | Data acquisition (W6), Assignment 2 (Streamlit app) | live API via requests; cached in app data/ |
CC BY 4.0; check per-API terms |
Conventions¶
- Add each committed dataset with a small
README.mdin its subfolder stating provenance, license, and how it was obtained. - Never commit files > 10 MB; store a download script in
scripts/or link to the source instead. - Prefer CC0 / CC BY datasets for student-facing materials.
Work in progress: sources.md and dataset files will be added with each lab
and assignment.