segmentation

pyts.utils.segmentation(ts_size: int, window_size: int, overlapping: bool = False, n_segments: int | None = None) → tuple[NDArray[int64], NDArray[int64], int]

Compute the indices for Piecewise Agrgegate Approximation.

Parameters:

ts_sizeint

The size of the time series.

window_sizeint

The size of the window.

overlappingbool (default = False)

If True, overlapping windows may be used. If False, non-overlapping are used.

n_segmentsint or None (default = None)

The number of windows. If None, the number is automatically computed using window_size.

Returns:

startarray

The lower bound for each window.

endarray

The upper bound for each window.

sizeint

The size of start.

Examples

>>> from pyts.utils import segmentation
>>> start, end, size = segmentation(ts_size=12, window_size=3)
>>> print(start)
[0 3 6 9]
>>> print(end)
[ 3  6  9 12]
>>> size
4