In [1]:

    

import numpy as np


    

In [2]:

    

a1 = np.ones((2, 3), int)
print(a1)


    

    




[[1 1 1]
 [1 1 1]]

In [3]:

    

a2 = np.full((2, 3), 2)
print(a2)


    

    




[[2 2 2]
 [2 2 2]]

In [4]:

    

print(np.block([a1, a2]))


    

    




[[1 1 1 2 2 2]
 [1 1 1 2 2 2]]

In [5]:

    

print(np.block([[a1], [a2]]))


    

    




[[1 1 1]
 [1 1 1]
 [2 2 2]
 [2 2 2]]

In [6]:

    

print(np.block([[a1, a2], [a2, a1]]))


    

    




[[1 1 1 2 2 2]
 [1 1 1 2 2 2]
 [2 2 2 1 1 1]
 [2 2 2 1 1 1]]

In [7]:

    

print(np.block([[[a1]], [[a2]]]))


    

    




[[[1 1 1]
  [1 1 1]]

 [[2 2 2]
  [2 2 2]]]

In [8]:

    

print(np.block([[[a1]], [[a2]]]).shape)


    

    




(2, 2, 3)

In [9]:

    

a3 = np.full(6, 3)
print(a3)


    

    




[3 3 3 3 3 3]

In [10]:

    

print(np.block([[a1, a2], [a3]]))


    

    




[[1 1 1 2 2 2]
 [1 1 1 2 2 2]
 [3 3 3 3 3 3]]

In [11]:

    

# print(np.block([[a1, a2], a3]))
# ValueError: List depths are mismatched. First element was at depth 2, but there is an element at depth 1 (arrays[1])


    

In [12]:

    

# print(np.block([[a1, a2, a3]]))
# ValueError: all the input array dimensions except for the concatenation axis must match exactly