Conditionally replace values on three 2d numpy arrays

I have three 2×2 numpy arrays like so:

import numpy as np

#create three arrays
one = np.array([[1, -1], 
                [-1, 1]])

two = np.array([[-1, 0], 
                [-2, 2]])

three = np.array([[1, 1], 
                [-4, 3]])

I want to make a fourth 2×2 array and replace the values based on one, two and three so that if in a particular location all three arrays have values less than one, to make the new value -1, and if all three arrays have values greater than one, to make the new value 1. If the condition is not met I want to turn the values to 0.

I am attempting this like so:

#make a copy of one 
final = one.copy()

#where all of the three initial arrays are > 0 make the value 1
final[(one > 0) & (two > 0) & (three >0)] = 1

#where all of the three initial arrays are < 0 make the value -1
final[(one < 0) & (two < 0) & (three <0)] = -1

which returns:

array([[ 1, -1],
       [-1,  1]])

So in this case the index at [0,0] and [0,1] I want to return zero while [1,0] should be -1 and [1,1] should be 1 or this:

[0, 0,
[-1, 1]

I think it is because I am not saying how to change the value to 0 when the original two conditions are not met, and I can’t seem to work out how to do that.

>Solution :

With np.select for multiple choices on conditions (and default value 0):

final = np.select([(one > 0) & (two > 0) & (three > 0), 
                   (one < 0) & (two < 0) & (three < 0)], [1, -1])
print(final)

[[ 0  0]
 [-1  1]]