>>> from collections import Counter
>>> from sklearn.datasets import make_classification
>>> from imblearn.under_sampling import ClusterCentroids
>>> X, y = make_classification(n_classes=2, class_sep=2,
... weights=[0.1, 0.9], n_informative=3, n_redundant=1, flip_y=0,
... n_features=20, n_clusters_per_class=1, n_samples=1000, random_state=10)
>>> print('Original dataset shape %s' % Counter(y))
Original dataset shape Counter({1: 900, 0: 100})
>>> cc = ClusterCentroids(random_state=42)
>>> X_res, y_res = cc.fit_resample(X, y)
>>> print('Resampled dataset shape %s' % Counter(y_res))
...
Resampled dataset shape Counter({...})