changepoints_to_labels

changepoints_to_labels(changepoints: ndarray, n_samples: int, labels: ndarray | None = None) → ndarray

Convert changepoint indices to per-sample segment labels.

Parameters:

changepointsnp.ndarray

Changepoint indices, shape (n_changepoints,).

n_samplesint

Number of samples in the time series.

labelsnp.ndarray | None, default=None

Segment labels, shape (n_changepoints + 1,). If None, auto-generates [0, 1, 2, …].

Returns:

np.ndarray

Segment labels, shape (n_samples,). Each sample assigned its segment label.

Examples

>>> changepoints = np.array([50, 100])
>>> labels = changepoints_to_labels(changepoints, n_samples=150)
>>> labels.shape
(150,)
>>> np.unique(labels)
array([0, 1, 2])