Create a Sliding Window function using NumPy.

sliding_window.py

# Create a function to reshape a ndarray using a sliding window.
# NOTE: The function uses numpy's internat as_strided function because looping in python is slow in comparison.

# Adopted from http://www.rigtorp.se/2011/01/01/rolling-statistics-numpy.html
import numpy as np

# Reshape a numpy array 'a' of shape (n, x) to form shape((n - window_size), window_size, x))
def rolling_window(a, window, step_size):
    shape = a.shape[:-1] + (a.shape[-1] - window + 1 - step_size + 1, window)
    strides = a.strides + (a.strides[-1] * step_size,)
    return np.lib.stride_tricks.as_strided(a, shape=shape, strides=strides)

FYI internet

There is off-by-one bug in this code. Last window is getting dropped.

I recommend this implementation that works correctly -> https://stackoverflow.com/a/6811241

Just edit the line #9 and add a +1 for the step_size

    shape = a.shape[:-1] + (a.shape[-1] - window + 1 - step_size + 1, window)

Thanks for the change. I'll update the code accordingly.

I tested the code but still wrong:

x = np.asarray([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
rolling_window(x, 3, 2)

Result:

array([[ 1, 3, 5], [ 2, 4, 6], [ 3, 5, 7], [ 4, 6, 8], [ 5, 7, 9], [ 6, 8, 10], [ 7, 9, 11], [ 8, 10, 12], [ 9, 11, 13], [10, 12, 14], [11, 13, 15], [12, 14, 0]])