Issue
Let's say I have three 2D numpy array and a weighted 1D numpy array that I want to weight this to some specific element in the array
import numpy as np
A= np.array([[2,0,2,2],[0,0,1,1],[1,2,1,0]])
B = np.array([[2,1,1,0],[2,2,0,2],[0,0,0,2]])
C = np.array([[0,0,3,2],[1,2,2,1],[2,3,1,5]])
Weight = np.array([0. , 0.5, 2. ])
Weight.reshape(-1,1)
1st condition: when we sum these arrays like below
sum = np.sum(A,axis=0) +np.sum(B,axis=0) + np.sum(C,axis=0)
for anything that is greater than 10, we select the array like below.
#selecting last 2 columns
A_1 = A[:,np.sum(A,axis=0) +np.sum(B,axis=0) + np.sum(C,axis=0) > 10]
B_1 = B[:,np.sum(A,axis=0) +np.sum(B,axis=0) + np.sum(C,axis=0) > 10]
C_1 = C[:,np.sum(A,axis=0) +np.sum(B,axis=0) + np.sum(C,axis=0) > 10]
(array([[2, 2],
[1, 1],
[1, 0]]),
array([[1, 0],
[0, 2],
[0, 2]]),
array([[3, 2],
[2, 1],
[1, 5]]))
2nd condition: when A_1 + B_1 - C_1 <0, we would like to modify the element to (A+B)*weight
diff = A_1 + B_1 - C_1 <0
(array([[False, False],
[ True, False],
[False, True]]),
array([[ 0, 0],
[-1, 2],
[ 0, -3]]))
For where we have True, we would like to change C_1 to
[[3, 2],
[(1+0)*0.5, 1],
[1, (0+2)*2.0]]
(1+0)*0.5 because 0.5 is at 2nd row and (0+2)*2.0 because 2.0 is at 3rd row
hence at the end C becomes..
array([[0, 0, 3, 2],
[1, 2, (1+0)*0.5, 1],
[2, 3, 1, (0+2)*2.0]]))
I would like to achieve this resultant C .. without using a loop Any help would be appreciated.
Solution
You can use masks:
AB = A+B
m1 = (AB+C).sum(0)>10
m2 = (AB-C) < 0
m = m1&m2
C = C.astype(float)
C[m] = (AB*Weight)[m]
Updated C
:
array([[0. , 0. , 3. , 2. ],
[1. , 2. , 0.5, 1. ],
[2. , 3. , 1. , 4. ]])
Answered By - mozway
0 comments:
Post a Comment
Note: Only a member of this blog may post a comment.