Insert columns from array into matrices from other array

Question

Suppose I have a numpy array array of matrices

A=numpy.array([[[1, 0, 0],[1, 1, 0],[0, 1, 0],[1, 1, 1]],[[1, 0, 0],[1, 1, 1],[0, 0, 1],[1, 0, 1]]])

and another array

B=np.array([[1,2,3,4],[5,6,7,8]])

which we interpret as an array of columns to insert into A.

How can I now insert the columns from B into A, i.e. to obtain

[[[1, 1, 0, 0],[2, 1, 1, 0],[3, 0, 1, 0],[4, 1, 1, 1]],[[5, 1, 0, 0],[6, 1, 1, 1],[7, 0, 0, 1],[8, 1, 0, 1]]]

?

I tried to use numpy.insert, but unfortunately I did not find to solve this problem with this method.

Bhargav · Accepted Answer · 2024-09-27 13:30:02Z

1

You can use np.concatenate to join A along the last axis (axis=2) - which represents the columns in each matrix

import numpy as np


A = np.array([[[1, 0, 0],[1, 1, 0],[0, 1, 0],[1, 1, 1]],
              [[1, 0, 0],[1, 1, 1],[0, 0, 1],[1, 0, 1]]])
B = np.array([[1,2,3,4],[5,6,7,8]])


B_reshaped = B.reshape(A.shape[0], A.shape[1], 1)


result = np.concatenate([B_reshaped, A], axis=2)

print(result)

Result

[[[1, 1, 0, 0],[2, 1, 1, 0],[3, 0, 1, 0],[4, 1, 1, 1]],
 [[5, 1, 0, 0],[6, 1, 1, 1],[7, 0, 0, 1],[8, 1, 0, 1]]]

edited Sep 27, 2024 at 13:30

answered Sep 27, 2024 at 13:23

Bhargav

4,8912 gold badges12 silver badges29 bronze badges

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

2 Comments

mrconcerned Over a year ago

It might be helpful to include what your code does.

Bhargav Over a year ago

iam typing.....

mozway · Accepted Answer · 2024-09-27 13:30:23Z

1

You could add a dimension to B and concatenate:

out = np.concatenate([B[..., None], A], axis=2)

Generic approach for an arbitrary number of dimensions (A having n dimensions and B having n-1):

out = np.concatenate([B[..., None], A], axis=-1)

Using insert:

out = np.insert(A, 0, B, axis=2)

Output:

array([[[1, 1, 0, 0],
        [2, 1, 1, 0],
        [3, 0, 1, 0],
        [4, 1, 1, 1]],

       [[5, 1, 0, 0],
        [6, 1, 1, 1],
        [7, 0, 0, 1],
        [8, 1, 0, 1]]])

edited Sep 27, 2024 at 13:30

answered Sep 27, 2024 at 13:23

mozway

267k13 gold badges56 silver badges106 bronze badges

Comments

Soudipta Dutta · Accepted Answer · 2024-09-30 11:08:39Z

This may save a bit of memory and improve performance :

import numpy as np

# Arrays A and B
A = np.array([[[1, 0, 0], [1, 1, 0], [0, 1, 0], [1, 1, 1]], 
              [[1, 0, 0], [1, 1, 1], [0, 0, 1], [1, 0, 1]]])

B = np.array([[1, 2, 3, 4], 
              [5, 6, 7, 8]])

res = np.concatenate([B[:,:,None],A],axis=2)
print(res)
'''
[[[1 1 0 0]
  [2 1 1 0]
  [3 0 1 0]
  [4 1 1 1]]

 [[5 1 0 0]
  [6 1 1 1]
  [7 0 0 1]
  [8 1 0 1]]]

'''

If the dataset doesn't fit into memory, we can use memory-mapped arrays. This avoids loading the entire array into RAM at once.

import numpy as np

# Arrays A and B
A = np.array([[[1, 0, 0], [1, 1, 0], [0, 1, 0], [1, 1, 1]], 
              [[1, 0, 0], [1, 1, 1], [0, 0, 1], [1, 0, 1]]])

B = np.array([[1, 2, 3, 4], 
              [5, 6, 7, 8]])

n_rows, n_cols, n_depth = A.shape

res = np.memmap('res.dat', dtype='int', mode='w+', shape=(n_rows, n_cols, n_depth + 1))
res[:, :, :1] = B[:, :, None]# Insert B as the first column
res[:, :, 1:] = A # Add the rest of A's columns
print(res)

'''
[[[1 1 0 0]
  [2 1 1 0]
  [3 0 1 0]
  [4 1 1 1]]

 [[5 1 0 0]
  [6 1 1 1]
  [7 0 0 1]
  [8 1 0 1]]]
'''

Collectives™ on Stack Overflow

Insert columns from array into matrices from other array

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

2 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related