Datasets#

The datasets listed below are bundled with the package. Each page describes the features and targets, gives the source with a ready-to-use citation, states the license and the changes made to the data, and shows how to load it.

Only datasets with an explicit CC0 1.0 or CC BY 4.0 license are bundled, see License and citation for the reason and for how this is checked.

Dataset

Task

Samples

Features

Missing values

License

Abalone

Regression

4177

8

No

CC BY 4.0

Adult

Classification

48842

14

Yes

CC BY 4.0

Airfoil Self-Noise

Regression

1503

5

No

CC BY 4.0

Auto MPG

Regression

398

7

Yes

CC BY 4.0

Bank Marketing

Classification

45211

16

No

CC BY 4.0

Breast Cancer Wisconsin (Diagnostic)

Classification

569

30

No

CC BY 4.0

Combined Cycle Power Plant

Regression

9568

4

No

CC BY 4.0

Concrete Compressive Strength

Regression

1030

8

No

CC BY 4.0

Default of Credit Card Clients

Classification

30000

23

No

CC BY 4.0

Optical Recognition of Handwritten Digits

Classification

1797

64

No

CC BY 4.0

Energy Efficiency

Regression

768

8

No

CC BY 4.0

Forest Fires

Regression

517

12

No

CC BY 4.0

German Credit

Classification

1000

20

No

CC BY 4.0

Iris

Classification

150

4

No

CC BY 4.0

Mushroom

Classification

8124

22

Yes

CC BY 4.0

Palmer Penguins

Classification

344

7

Yes

CC0 1.0

Real Estate Valuation

Regression

414

6

No

CC BY 4.0

Superconductivity Data

Regression

21263

81

No

CC BY 4.0

Wine

Classification

178

13

No

CC BY 4.0

Wine Quality

Regression

6497

11

No

CC BY 4.0

Missing values are stored as NaN. A dataset says whether it has any before you load it: get_info(name).has_missing.