Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
def sliding_window(data, size, stepsize=1, padded=False, axis=-1, copy=True):
"""
Calculate a sliding window over a signal
Parameters
----------
data : numpy array
The array to be slided over.
size : int
The sliding window size
stepsize : int
The sliding window stepsize. Defaults to 1.
axis : int
The axis to slide over. Defaults to the last axis.
copy : bool
Return strided array as copy to avoid sideffects when manipulating the
output array.
Returns
-------
data : numpy array
A matrix where row in last dimension consists of one instance
of the sliding window.
Notes
-----
- Be wary of setting `copy` to `False` as undesired sideffects with the
output values may occurr.
Examples
--------
>>> a = numpy.array([1, 2, 3, 4, 5])
>>> sliding_window(a, size=3)
array([[1, 2, 3],
[2, 3, 4],
[3, 4, 5]])
>>> sliding_window(a, size=3, stepsize=2)
array([[1, 2, 3],
[3, 4, 5]])
See Also
--------
pieces : Calculate number of pieces available by sliding
"""
if axis >= data.ndim:
raise ValueError(
"Axis value out of range"
)
if stepsize < 1:
raise ValueError(
"Stepsize may not be zero or negative"
)
if size > data.shape[axis]:
raise ValueError(
"Sliding window size may not exceed size of selected axis"
)
shape = list(data.shape)
shape[axis] = numpy.floor(data.shape[axis] / stepsize - size / stepsize + 1).astype(int)
shape.append(size)
strides = list(data.strides)
strides[axis] *= stepsize
strides.append(data.strides[axis])
strided = numpy.lib.stride_tricks.as_strided(
data, shape=shape, strides=strides
)
if copy:
return strided.copy()
else:
return strided
@delaram19

This comment has been minimized.

Copy link

@delaram19 delaram19 commented Jun 22, 2017

Tnx for sharing, it helped me a lot!

@dojeda

This comment has been minimized.

Copy link

@dojeda dojeda commented Nov 21, 2019

Hi!
I wanted to report that there is a problem with 2D input data that one can reproduce as:

>>> x = np.arange(30).reshape(3, -1)
>>> x
array([[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
       [20, 21, 22, 23, 24, 25, 26, 27, 28, 29]])

>>> sliding_window(x, 3, 2)
array([[[ 0,  1,  2],
        [ 2,  3,  4],
        [ 4,  5,  6],
        [ 6,  7,  8]],

       [[10, 11, 12],
        [12, 13, 14],
        [14, 15, 16],
        [16, 17, 18]],

       [[20, 21, 22],
        [22, 23, 24],
        [24, 25, 26],
        [26, 27, 28]]])

which is particularly weird because the epochs have more rows than the input. The correct output should be, in my opinion:

>>> epoch(x, 3, 2)
array([[[ 0,  1,  2],
        [10, 11, 12],
        [20, 21, 22]],

       [[ 2,  3,  4],
        [12, 13, 14],
        [22, 23, 24]],

       [[ 4,  5,  6],
        [14, 15, 16],
        [24, 25, 26]],

       [[ 6,  7,  8],
        [16, 17, 18],
        [26, 27, 28]]])

Which I generated with: https://gist.github.com/dojeda/2250ad463a925a38d7179ec913167ebc
but I am not 100% sure it works on all other cases as well.

@nils-werner

This comment has been minimized.

Copy link
Owner Author

@nils-werner nils-werner commented Nov 25, 2019

Nowadays I recommend using skimage.util.view_as_windows() instead:

skimage.util.view_as_windows(x, (3, 3), step=2)
# array([[[[ 0,  1,  2],
#          [10, 11, 12],
#          [20, 21, 22]],
# 
#         [[ 2,  3,  4],
#          [12, 13, 14],
#          [22, 23, 24]],
# 
#         [[ 4,  5,  6],
#          [14, 15, 16],
#          [24, 25, 26]],
# 
#         [[ 6,  7,  8],
#          [16, 17, 18],
#          [26, 27, 28]]]])
@dojeda

This comment has been minimized.

Copy link

@dojeda dojeda commented Nov 25, 2019

Nice, I didn't known sklearn had this util function.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.