Port of the credit-card dataset from UCI (link here ). See details there and use carefully.

Basic preprocessing done by the imodels team in this notebook .

The target is the binary outcome default.payment.next.month .

Sample usage

Load the data:

from datasets import load_dataset

dataset = load_dataset("imodels/credit-card")
df = pd.DataFrame(dataset['train'])
X = df.drop(columns=['default.payment.next.month'])
y = df['default.payment.next.month'].values

Fit a model:

import imodels
import numpy as np

m = imodels.FIGSClassifier(max_rules=5)
m.fit(X, y)
print(m)

Evaluate:

df_test = pd.DataFrame(dataset['test'])
X_test = df.drop(columns=['default.payment.next.month'])
y_test = df['default.payment.next.month'].values
print('accuracy', np.mean(m.predict(X_test) == y_test))