In [1]:

    

import numpy as np


    

In [2]:

    

a_3d = np.arange(24).reshape(2, 3, 4)
print(a_3d)


    

    




[[[ 0  1  2  3]
  [ 4  5  6  7]
  [ 8  9 10 11]]

 [[12 13 14 15]
  [16 17 18 19]
  [20 21 22 23]]]

In [3]:

    

print(a_3d.shape)


    

    




(2, 3, 4)

In [4]:

    

a0, a1 = np.dsplit(a_3d, 2)


    

In [5]:

    

print(a0)


    

    




[[[ 0  1]
  [ 4  5]
  [ 8  9]]

 [[12 13]
  [16 17]
  [20 21]]]

In [6]:

    

print(a0.shape)


    

    




(2, 3, 2)

In [7]:

    

print(a1)


    

    




[[[ 2  3]
  [ 6  7]
  [10 11]]

 [[14 15]
  [18 19]
  [22 23]]]

In [8]:

    

print(a1.shape)


    

    




(2, 3, 2)

In [9]:

    

a0, a1 = np.dsplit(a_3d, [1])


    

In [10]:

    

print(a0)


    

    




[[[ 0]
  [ 4]
  [ 8]]

 [[12]
  [16]
  [20]]]

In [11]:

    

print(a1)


    

    




[[[ 1  2  3]
  [ 5  6  7]
  [ 9 10 11]]

 [[13 14 15]
  [17 18 19]
  [21 22 23]]]

In [12]:

    

a = np.arange(16).reshape(4, 4)
print(a)


    

    




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

In [13]:

    

# np.dsplit(a, 2)
# ValueError: dsplit only works on arrays of 3 or more dimensions