Python新手求助：如何对指定3D数组进行重塑？

Hey there! Let's work through this 3D array reshaping problem you're dealing with. First, let's break down what we're starting with and what we need to end up with:

Original Array (Shape: (3, 5, 3)) #

[
 [[ 1, 16, 31], [ 2, 17, 32], [ 3, 18, 33], [ 4, 19, 34], [ 5, 20, 35]],
 [[ 6, 21, 36], [ 7, 22, 37], [ 8, 23, 38], [ 9, 24, 39], [10, 25, 40]],
 [[11, 26, 41], [12, 27, 42], [13, 28, 43], [14, 29, 44], [15, 30, 45]]
]

Target Array (Shape: (3, 3, 5)) #

[
 [[ 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, 30]],
 [[31, 32, 33, 34, 35], [36, 37, 38, 39, 40], [41, 42, 43, 44, 45]]
]

Solution with NumPy (Most Efficient) #

NumPy makes this kind of multidimensional array manipulation straightforward with axis transposition. Here's how to do it:

1. Import NumPy and define your original array:

import numpy as np

original_array = np.array([
    [[ 1, 16, 31], [ 2, 17, 32], [ 3, 18, 33], [ 4, 19, 34], [ 5, 20, 35]],
    [[ 6, 21, 36], [ 7, 22, 37], [ 8, 23, 38], [ 9, 24, 39], [10, 25, 40]],
    [[11, 26, 41], [12, 27, 42], [13, 28, 43], [14, 29, 44], [15, 30, 45]]
])

2. Transpose the axes to reorder the dimensions:
   We need to shift the third dimension (the inner-most values like 1,16,31) to the first position, while keeping the original first and second dimensions in the new second and third positions. This is done with transpose(2, 0, 1):

transformed_array = original_array.transpose(2, 0, 1)

3. Check the result:
   Printing transformed_array will give you exactly the target structure you need.

Pure Python Solution (No Libraries) #

If you prefer not to use NumPy, you can achieve the same result with nested list comprehensions:

original_list = [
    [[ 1, 16, 31], [ 2, 17, 32], [ 3, 18, 33], [ 4, 19, 34], [ 5, 20, 35]],
    [[ 6, 21, 36], [ 7, 22, 37], [ 8, 23, 38], [ 9, 24, 39], [10, 25, 40]],
    [[11, 26, 41], [12, 27, 42], [13, 28, 43], [14, 29, 44], [15, 30, 45]]
]

# Build the transformed list by grouping inner elements across blocks
transformed_list = [
    [
        [sub[elem_idx] for sub in block]
        for block in original_list
    ]
    for elem_idx in range(3)
]

print(transformed_list)

This works by iterating over each position in the inner-most subarrays (0, 1, 2), then collecting those values across all blocks and subarrays to form the new structure.

内容的提问来源于stack exchange，提问作者Luca Fichera