Numpy Roll method Implementation in Python : Only 3 Steps

Numpy roll allows you to shift the elements of the NumPy array. It rolls the elements along the given axis. In this post, you will learn how to use the NumPy roll to shift NumPy array elements to the left or right for both 1D and 2D array.

Syntax of Numpy Roll Function

numpy.roll(a, shift, axis=None)

Explanation of the Parameter

a: It is the input array, can accept a 1D or 2D array.

shift: It is a positive or negative integer. The number of places you want to shift.

axis: Allows you to do shifting along rows or columns.

Steps to Implement Numpy Roll

Step 1: Import all the necessary libraries

In this article, I am using only NumPy modules. So let’s import them.

import numpy as np

Step 2: Create a Sample Numpy Array

In this step, I am creating both 1D and 2D NumPy arrays. I will implement the NumPy roll function on both of them.

1D Numpy array

array_1d = np.array([1,2,3,4,5,6,7])

2D Numpy array

array_2d = np.array([[1,2,3],[4,5,6],[7,8,9]])

Step 3: Implement the Numpy roll on the array

After the creation of the NumPy array let’s implement numpy.roll() method on them.

Applying Numpy Roll on a 1D array

Let’s shift the NumPy elements to 1 place. Then I will shift to 2 places and 3 places.

Shifting to 1 place

np.roll(array_1d,1)

Output

Here the last element will come to the first and remaining shift rightward.

Shifting to 2 place

np.roll(array_1d,2)

Output

Numpy roll on 1D array with shift =2 — roll on a 1D array with shift =2

The above code will shift the last two elements to the first and the remaining will shift right.

Shift to 3 Place

To shift the last three elements to the right use shift =3.

np.roll(array_1d,3)

Output

roll on a 1D array with shift =2 — roll on a 1D array with shift =3

In the same, if you want to shift the elements leftward then pass the shift value as negative. For example, If I will use shift =-1, then the first element will become the last element and the remaining will shift left.

Applying Numpy Roll on a 2D array

Implementing numpy.roll on 2D NumPy array is different from 1D. Here We have three cases.

Case 1: Apply roll() without axis

If you apply roll on a 2D array without axis. Then all the elements will be shifted right or left.

np.roll(array_2d,1)

Output

Applying roll() without axis on 2D Array

Case 2: Apply roll() without axis =0

Applying roll() with axis=0, will shift the elements of each column downward or upwards.

np.roll(array_2d,1,axis=0)

Output

Applying roll() with axis =0 on 2D Array

Case 3: Apply roll() without axis =1

In case 3, if you set the value of axis to 1, then the shifting of the elements will take place row-wise or horizontally.

np.roll(array_2d,1,axis=1)

Output

Applying roll() with axis =1 on 2D Array

Conclusion

Numpy roll allows reducing the time of indexing for a certain element in the array. You just shift the elements and then index it. Thus increase efficiency. These are the steps to implement the roll. Hope you have liked this article. If you have any queries then you can contact us.

Source:

Roll Documentation