提取二维数组中最小点的索引

Question

I'm testing values of A and B in a function and need to find the mimum value of that function and what values of A and B produce this minimum. I'm struggling to find a way to get these values out of the array I have made.

I start by defining some arrays

ones = np.ones((3456,5184))
array = np.zeros([5,5])
real = threshold
test = testsunthree

Where real and test are arrays defined early in the code.

for A in range(-2,2):
  for B in range(-2,2):
    array[A,B]=((np.mean((real-(test*(A*0.1)+((B*0.1)*ones))**2))))

Here I am testing values of A and B over the range between -2 and 2 and for each value plugging them into the function. The value of that function is then stored in the empty array defined above.

array_min = array[array != 0].min()
print (array_min)
print zip(*np.where(array == array.min()))

I am then trying to identify the minimum value of that array and want to find for what values of A and B is this function at its minimum. This is the part I am struggling with as the last two lines are giving incorrect values.

Answer 1

This is giving you the minimum value of the array called array . I put random values for threshold and testsunthree to make the code work and change the utilisation of the min() function

import numpy as np

threshold = 4
testsunthree = 2
ones = np.ones((3456,5184))
array = np.zeros([5,5])
real = threshold
test = testsunthree

for A in range(-2,2):
  for B in range(-2,2):
     array[A,B]=((np.mean((real-(test*(A*0.1)+((B*0.1)*ones))**2))))

array_min = np.min(array)
print (array_min)

I quickly made a dirty solution to obtain the different A,B values for which you have the min value of your array, it can be improved easily as it's only a quick fix to show you the way

index = np.where(array == array_min)
A_B_value = []

for i in range(0, len(index[0])):
    min_index = []
    min_index.append(index[0][i])
    min_index.append(index[1][i])
    A_B_value.append(min_index)

print(A_B_value)