Finding the indexes of an array-CodePudding

If i have an array like this

a = np.array([[False, False, False, False, False, False, False, False, False, False],[False, False, False, False, False, False, False, False, False, False],[False, False, False,  True,  True, False, False, False, False, False],

I tried np.random.choice but it doesnt work for 1_D arrays :(

CodePudding user response：

A possible solution is to loop through different indexes until you find a match

E.g.


import random
index_0 = 0
index_1 = 0
found = False
while not found:
   temp_i0 = random.randint(len(array))
   temp_i1 = random.randint(len(array[0]))
   if array[temp_i0][temp_i1]:
      index_0 = temp_i0
      index_1 = temp_i1
      found = True

CodePudding user response：

You can use np.random.choice with a little extra work:

You can use:

np.unravel_index(np.random.choice(np.flatnonzero(a)), a.shape)

In words:

Flatten a and find the indices where a is nonzero in a using np.flatnonzero. Then, pick a random index among nonzero indices using np.random.choice. This gives us a random index in the flattened array, so we convert it back into 2D row/column coordinates using np.unravel_index.