Reshape

We can create powerful arrays with the reshape method.

The syntax is: np.reshape(ary, newshape)

the reshape changes the shape of the array. the new shape is a tuple or a list that must be the same multiplication result as the old shape

For example: old shape: (2, 5), multiplication result: 2*5=10, the new shape can be any combination of numbers as long as the multiplication is 10, meaning the new shape can be either one of: (1, 10) , (10, 1) , (5, 2)

Example 1

ary = np.arange(10) # --> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
ary = np.reshape(ary, (5, 2))
>>> [[0, 1],
     [2, 3],
     [4, 5],
     [6, 7],
     [8, 9]]

Example 2

ary = np.zeros(6) # --> [0., 0., 0., 0., 0., 0.]
ary = np.reshape(ary, (2, 3))
>>> [[0., 0., 0.],
     [0., 0., 0.]]

Exmaple 3

Reshape to higher dimensions

ary = np.arange(12)
ary = np.reshape(ary, (2, 3, 2))
>>> [[[ 0,  1],
      [ 2,  3],
      [ 4,  5]],

     [[ 6,  7],
      [ 8,  9],
      [10, 11]]]