Increment some columns of multiple rows in python

Question

I am trying to update multiple columns of a set of rows simultaneously. (in my case I have the same number of columns and the number of rows -- I'm making a similarity matrix, but generally speaking, this shouldn't be important). Is there a way to do this more efficiently than my examples below? -- I'm open to using python lists, pandas or numpy; doesn't really matter; as long as it's the fastest.

Example 1 - Two nested for loops

adj_mat = np.array([[1,1,1,0,0,0,0,0,0],
                [1,1,1,0,0,0,0,0,0],
                [1,1,1,1,0,0,0,0,0],
                [0,0,1,1,1,1,0,0,0],
                [0,0,0,1,1,1,0,0,0],
                [0,0,0,1,1,1,1,0,0],
                [0,0,0,0,0,1,1,1,1],
                [0,0,0,0,0,0,1,1,1],
                [0,0,0,0,0,0,1,1,1]])
inlier_mask = np.array([False, False, False, True, True, True, False, 
False, False])
inlier_idx = np.array(np.nonzero(inlier_mask))[0].tolist()

for i in inlier_idx:
    for j in inlier_idx:
        adj_mat[i,j] += 1

print(adj_mat)

Output:

[[1 1 1 0 0 0 0 0 0]
 [1 1 1 0 0 0 0 0 0]
 [1 1 1 1 0 0 0 0 0]
 [0 0 1 2 2 2 0 0 0]
 [0 0 0 2 2 2 0 0 0]
 [0 0 0 2 2 2 1 0 0]
 [0 0 0 0 0 1 1 1 1]
 [0 0 0 0 0 0 1 1 1]
 [0 0 0 0 0 0 1 1 1]]

Example 2 - Only one for loop

adj_mat = np.array([[1,1,1,0,0,0,0,0,0],
                [1,1,1,0,0,0,0,0,0],
                [1,1,1,1,0,0,0,0,0],
                [0,0,1,1,1,1,0,0,0],
                [0,0,0,1,1,1,0,0,0],
                [0,0,0,1,1,1,1,0,0],
                [0,0,0,0,0,1,1,1,1],
                [0,0,0,0,0,0,1,1,1],
                [0,0,0,0,0,0,1,1,1]])
inlier_mask = np.array([False, False, False, True, True, True, False, 
False, False])
inlier_idx = np.array(np.nonzero(inlier_mask))[0].tolist()

for i in inlier_idx:
    adj_mat[i,inlier_idx] += 1

Output:

[[1 1 1 0 0 0 0 0 0]
 [1 1 1 0 0 0 0 0 0]
 [1 1 1 1 0 0 0 0 0]
 [0 0 1 2 2 2 0 0 0]
 [0 0 0 2 2 2 0 0 0]
 [0 0 0 2 2 2 1 0 0]
 [0 0 0 0 0 1 1 1 1]
 [0 0 0 0 0 0 1 1 1]
 [0 0 0 0 0 0 1 1 1]]

Is there a solution such as: adj_mat[inlier_idx,inlier_idx] += 1 that can achieve this with no loops?

Divakar · Accepted Answer · 2018-08-19 17:19:28Z

2

Use np.ix_ to create two broadcastable indexing arrays and then index and add -

adj_mat[np.ix_(inlier_idx, inlier_idx)] += 1
# or adj_mat[np.ix_(inlier_mask, inlier_mask)] += 1

Alternatively, we can manually create the broadcastable arrays with extending inlier_idx to 2D with None/np.newaxis -

inlier_idx = np.flatnonzero(inlier_mask)
adj_mat[inlier_idx[:,None], inlier_idx] += 1

edited Aug 19, 2018 at 17:19

answered Aug 19, 2018 at 17:12

Divakar

222k19 gold badges273 silver badges374 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Increment some columns of multiple rows in python

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related