Move an element in the array row-wise and shifting the others?

Question

I am working with numpy arrays and and I would like to move a certain value located at certain index by 2 positions on the right while shifting the others element involved on the left by one position each. To be more precise, here below is what I would like to achieve:

from:

     start = [0., 1., 2.],
             [0., 0., 0.],
             [5., 0., 1.]

to:

       end = [1., 2., 0.],
             [0., 0., 0.],
             [5., 0., 1.]

As you can see from the first row, 0 has been moved by two position on the right and 1 and 2 by one position on the left. So far, I succeded by moving an element by one position and moving the other by defining:

def right_move(arr, index, n_steps:int):


   row = index[0]
   col = index[1]

   try:
       arr[row,col], arr[row, col + n_steps] = arr[row, col + n_steps], arr[row, col]

   except:
       pass

   return arr

where n_steps=1. If I input n_steps=2, it won't work since the element in the middle of the swapping remain unchanged.

Can someone help me? Other easier solution are more then welcome!

Thanks

It seems like you need numpy.roll check it out docs.scipy.org/doc/numpy/reference/generated/numpy.roll.html — Francesco
– Francesco, Commented Feb 28, 2020 at 17:46

jfaccioni · Accepted Answer · 2020-02-28 17:47:36Z

1

You can use numpy.roll:

import numpy as np

arr = np.array([[0., 1., 2.],
                [0., 0., 0.],
                [5., 0., 1.]])
arr[0] = np.roll(arr[0], 2)
arr
# output:
# [[1., 2., 0.],
#  [0., 0., 0.],
#  [5., 0., 1.]]

answered Feb 28, 2020 at 17:47

jfaccioni

7,5591 gold badge11 silver badges27 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Andy L. · Accepted Answer · 2020-02-28 23:28:35Z

It is supposed the functionality of np.roll. However, np.roll doesn't allow roll on individual row. It piqued my interest to implement a function using as_strided to extend this functionality to np.roll

from numpy.lib.stride_tricks import as_strided

def custom_roll(arr, r_tup):
    m = np.asarray(r_tup)
    arr_roll = arr[:, [*range(arr.shape[1]),*range(arr.shape[1]-1)]].copy() #need `copy`
    strd_0, strd_1 = arr_roll.strides
    n = arr.shape[1]
    result = as_strided(arr_roll, (*arr.shape, n), (strd_0 ,strd_1, strd_1))

    return result[np.arange(arr.shape[0]), (n-m)%n]

Just passing a tuple indicate number of steps for each row. In your case, it rolls only row 0 either 2 positions forward or -1 position backward. so the tuple is (2,0,0) or (-1,0,0)

start = np.array([[0., 1., 2.],
                  [0., 0., 0.],
                  [5., 0., 1.])

out = custom_roll(start, (2,0,0))

Out[780]:
array([[1., 2., 0.],
       [0., 0., 0.],
       [5., 0., 1.]])

Or

out = custom_roll(start, (-1,0,0))

Out[782]:
array([[1., 2., 0.],
       [0., 0., 0.],
       [5., 0., 1.]])

Roll first and last row: (2,0,2)

out =  custom_roll(start, (2,0,2))

Out[784]:
array([[1., 2., 0.],
       [0., 0., 0.],
       [0., 1., 5.]])

Collectives™ on Stack Overflow

Move an element in the array row-wise and shifting the others?

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related