Change the dimension of a NumPy array

Last Updated : 22 Mar, 2023

Let’s discuss how to change the dimensions of an array. In NumPy, this can be achieved in many ways. Let’s discuss each of them.

Method #1: Using Shape()

Syntax :

array_name.shape()

Python3

# importing numpy
import numpy as np
 
 
def main():
 
    # initialising array
    print('Initialised array')
    gfg = np.array([1, 2, 3, 4])
    print(gfg)
 
    # checking current shape
    print('current shape of the array')
    print(gfg.shape)
     
    # modifying array according to new dimensions
    print('changing shape to 2,2')
    gfg.shape = (2, 2)
    print(gfg)
 
if __name__ == "__main__":
    main()

Output:

Initialised array
[1 2 3 4]
current shape of the array
(4,)
changing shape to 2,2
[[1 2]
 [3 4]]

Method #2: Using reshape()

The order parameter of reshape() function is advanced and optional. The output differs when we use C and F because of the difference in the way in which NumPy changes the index of the resulting array. Order A makes NumPy choose the best possible order from C or F according to available size in a memory block.

Difference between Order C and F

Syntax :

numpy.reshape(array_name, newshape, order= 'C' or 'F' or 'A')

Python3

# importing numpy
import numpy as np
 
 
def main():
 
    # initialising array
    gfg = np.arange(1, 10)
    print('initialised array')
    print(gfg)
     
    # reshaping array into a 3x3 with order C
    print('3x3 order C array')
    print(np.reshape(gfg, (3, 3), order='C'))
 
    # reshaping array into a 3x3 with order F
    print('3x3 order F array')
    print(np.reshape(gfg, (3, 3), order='F'))
 
    # reshaping array into a 3x3 with order A
    print('3x3 order A array')
    print(np.reshape(gfg, (3, 3), order='A'))
 
 
if __name__ == "__main__":
    main()

Output :

initialised array
[1 2 3 4 5 6 7 8 9]
3x3 order C array
[[1 2 3]
 [4 5 6]
 [7 8 9]]
3x3 order F array
[[1 4 7]
 [2 5 8]
 [3 6 9]]
3x3 order A array
[[1 2 3]
 [4 5 6]
 [7 8 9]]

Method #3 : Using resize()

The shape of the array can also be changed using the resize() method. If the specified dimension is larger than the actual array, The extra spaces in the new array will be filled with repeated copies of the original array.

Syntax :

numpy.resize(a, new_shape)

Python3

# importing numpy
import numpy as np
 
 
def main():
 
    # initialise array
    gfg = np.arange(1, 10)
    print('initialised array')
    print(gfg)
     
    # resized array with dimensions in
    # range of original array
    gfg1=np.resize(gfg, (3, 3))
    print('3x3 array')
    print(gfg1)
     
    # resized array with dimensions larger than
    # original array
    gfg2=np.resize(gfg, (4, 4))
     
    # extra spaces will be filled with repeated
    # copies of original array
    print('4x4 array')
    print(gfg2)
     
    # resize array with dimensions larger than 
    # original array
    gfg.resize(5, 5)
     
    # extra spaces will be filled with zeros
    print('5x5 array')
    print(gfg)
 
 
if __name__ == "__main__":
    main()

Output :

initialised array
[1 2 3 4 5 6 7 8 9]
3x3 array
[[1 2 3]
 [4 5 6]
 [7 8 9]]
4x4 array
[[1 2 3 4]
 [5 6 7 8]
 [9 1 2 3]
 [4 5 6 7]]
5x5 array
[[1 2 3 4 5]
 [6 7 8 9 0]
 [0 0 0 0 0]
 [0 0 0 0 0]
 [0 0 0 0 0]]