from sklearn.datasets import make_classification
X, y = make_classification(n_samples=1000, n_features=20,
              n_informative=5, n_redundant=2, n_classes=4)