757 distinct datasets originally from the R ecosystem are provided through a unified Python interface. The collection spans multiple disciplines including sociology, medicine, and economics, with each entry accompanied by its original documentation and metadata.
Use Cases
- Train multi-class classification models using the Species column in the 'iris' dataset
- Perform multivariate regression to predict mpg using wt and hp features from the 'mtcars' table
- Analyze historical passenger trends using the monthly 'AirPassengers' time-series data
- Evaluate survival probabilities using the status and time variables in the 'cancer' dataset
Strengths
- Contains 757 individual datasets including classic benchmarks like 'iris' and 'Titanic'
- Includes a searchable index of all items accessible via the data() command
- Provides detailed metadata and descriptive text for every dataset through the show_doc argument
- Supports direct loading into pandas DataFrames for immediate analysis