finding the minimum value in an array with returned values

I have an array for example:

[1 2 3 4 
 2 3 4 0
 5 4 0 6]

And I want to find the indexes of all the values that are closer to to the value 3.9 (in my example,4)

I tried using :

import numpy as np
def find_nearest(array, value):
    idx = (np.abs(array - value)).argmin()

and

np.where(array== array.min())

but none of the options gives me the correct answer.

I do excpect to get:

(3,1),(1,2),(2,1)

In my original code, I iterate an array with the shape of 3648X5472 so "FOR" loops might be too heavy.

hope to get some help here, thank you

>Solution :

You can use:

a = np.array([[1, 2, 3, 4], 
              [2, 3, 4, 0],
              [5, 4, 0, 6]])

v = 3.9

b = abs(a-v)

xs, ys = np.where(b == np.min(b))

output:

>>> xs
array([0, 1, 2])

>>> ys
array([3, 2, 1])

Alternative output:

>>> np.c_[np.where(b == np.min(b))]
array([[0, 3],
       [1, 2],
       [2, 1]])
# or 
>>> np.argwhere(b==np.min(b))