How can I divide a NumPy 2D array into smaller 2D arrays?

Slicing 2D Arrays into Smaller Arrays

Problem:
You wish to fragment a two-dimensional (2D) NumPy array into smaller 2D arrays. For instance, you may want to transform a 2x4 array into two 2x2 arrays.

Solution:
A combination of reshape and swapaxes functions allows you to divide your array into "blocks." Here's a Python implementation that achieves this:

def blockshaped(arr, nrows, ncols):
    h, w = arr.shape
    assert h % nrows == 0, f"{h} rows is not evenly divisible by {nrows}"
    assert w % ncols == 0, f"{w} cols is not evenly divisible by {ncols}"

    return (arr.reshape(h//nrows, nrows, -1, ncols)
               .swapaxes(1,2)
               .reshape(-1, nrows, ncols))

In this solution:

• nrows and ncols define the dimensions of each smaller array.
• You must ensure that the input array is evenly divisible by these values.
• The reshape function rearranges the array into blocks.
• The swapaxes function swaps the axes to obtain the desired shape.
• The final reshape flattens the array into a sequence of smaller blocks.

Example:

Consider the following input array:

c = np.arange(24).reshape((4, 6))

print(c)

[out]:
[[ 0  1  2  3  4  5]
 [ 6  7  8  9 10 11]
 [12 13 14 15 16 17]
 [18 19 20 21 22 23]]

Using blockshaped with nrows=2 and ncols=3, you can fragment this array into the following blocks:

print(blockshaped(c, 2, 3))

[out]:
[[[ 0  1  2]
  [ 6  7  8]]

 [[ 3  4  5]
  [ 9 10 11]]

 [[12 13 14]
  [18 19 20]]

 [[15 16 17]
  [21 22 23]]]

This demonstration illustrates how you can slice a 2D array into smaller rectangular arrays of specified dimensions.

The above is the detailed content of How can I divide a NumPy 2D array into smaller 2D arrays?. For more information, please follow other related articles on the PHP Chinese website!