adjusted_rand_index

adjusted_rand_index(y_true: ArrayLike, y_pred: ArrayLike) → float

Compute the adjusted Rand index for two segmentations.

Similar to Rand index but adjusted for chance. Wraps sklearn.metrics.adjusted_rand_score.

Parameters:

y_truearray-like of shape (n_samples,)

True segment labels, as returned by predict().

y_predarray-like of shape (n_samples,)

Predicted segment labels, as returned by predict().

Returns:

float

Adjusted Rand index in [-1, 1]. Higher is better. 1.0 = perfect agreement, ~0.0 = random labeling.

Examples

>>> adjusted_rand_index([0, 0, 0, 1, 1, 1], [0, 0, 0, 1, 1, 1])
1.0