In [1]:

    

import numpy as np


    

In [2]:

    

>>> x = np.array([[[1],[2],[3]], [[4],[5],[6]]])
>>> x.shape


    

    
Out[2]:





(2, 3, 1)

In [3]:

    

>>> x[1:2]


    

    
Out[3]:





array([[[4],
        [5],
        [6]]])

In [4]:

    

x = np.array([[[1],[2],[3]], [[4],[5],[6]]])


    

In [5]:

    

x.ndim


    

    
Out[5]:





3

In [6]:

    

x = np.array([1, 2, 3])


    

In [7]:

    

x.ndim


    

    
Out[7]:





1

In [8]:

    

x.shape


    

    
Out[8]:





(3,)

In [9]:

    

x[1]


    

    
Out[9]:





2

In [10]:

    

x = np.array([[1, 2, 3], [4, 5, 6]])
x.ndim, x.shape


    

    
Out[10]:





(2, (2, 3))

In [11]:

    

x[0]


    

    
Out[11]:





array([1, 2, 3])

In [12]:

    

x[1]


    

    
Out[12]:





array([4, 5, 6])

In [13]:

    

%timeit x[0, 1]
x[0, 1]


    

    




The slowest run took 21.43 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 694 ns per loop






    
Out[13]:





2

In [14]:

    

%timeit x[(0, 1)]
x[(0, 1)]


    

    




The slowest run took 14.98 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 727 ns per loop






    
Out[14]:





2

In [15]:

    

%timeit x[0][1]
x[0][1]


    

    




The slowest run took 10.93 times longer than the fastest. This could mean that an intermediate result is being cached.
100000 loops, best of 3: 1.37 µs per loop






    
Out[15]:





2