I have an array called A:
A = np.array([[-1, 2, -3],
[4, -5, 6],
[-7, 8, -9]])
Now I want to extract positive and negative parts and save them in two new arrays like:
B = np.array([[0, 2, 0],
[4, 0, 6],
[0, 8, 0]])
for positive parts and
C = np.array([[-1, 0, -3],
[0, -5, 0],
[-7, 0, -9]])
for negative parts.
Could you please guide me how to get arrays B and C from array A in Python 3.6?
Use two np.wheres to replace the values below zero and above zero:
>>> import numpy as np
>>> A = np.array([[-1, 2, -3], [4, -5, 6], [-7, 8, -9]])
>>> B = np.where(A > 0, A, 0)
>>> C = np.where(A < 0, A, 0)
>>> B
array([[0, 2, 0],
[4, 0, 6],
[0, 8, 0]])
>>> C
array([[-1, 0, -3],
[ 0, -5, 0],
[-7, 0, -9]])
The three argument form of np.where broadcasts the arguments, the first and second already have the same shape (3, 3) but the 0 will be broadcasted to:
>>> Ag0_bc, B_bc, zero_bc = np.broadcast_arrays(A > 0, A, 0)
>>> Ag0_bc
array([[False, True, False],
[ True, False, True],
[False, True, False]], dtype=bool)
>>> zero_bc
array([[0, 0, 0],
[0, 0, 0],
[0, 0, 0]])
Then np.where will create a new array and fill each element in the new array with the corresponding element from the second argument (A) if the element in the first argument (A > 0) is True and take the element from the third argument 0 in case it's False.