skdh.utility.windowing.get_windowed_view

skdh.utility.windowing.get_windowed_view(x, window_length, step_size, ensure_c_contiguity=False)

Return a moving window view over the data

Parameters:

xnumpy.ndarray

1- or 2-D array of signals to window. Windows occur along the 0 axis. Must be C-contiguous.

window_lengthint

Window length/size.

step_sizeint

Step/stride size for windows - how many samples to step from window center to window center.

ensure_c_contiguitybool, optional

Create a new array with C-contiguity if the passed array is not C-contiguous. This may result in the memory requirements significantly increasing. Default is False, which will raise a ValueError if x is not C-contiguous

Returns:

x_winnumpy.ndarray

2- or 3-D array of windows of the original data, of shape (…, L[, …])