>> y_score = np.array([[0.75, 0.5, 1], [1, 0.2, 0.1]])
    >>> label_ranking_average_precision_score(y_true, y_score)
    0.416...
    FÚ