Merge two numpy arrays

Question

I am trying to merge two arrays with the same number of arguments.

Input:

first = [[650001.88, 300442.2,   18.73,  0.575,  650002.094, 300441.668, 18.775],
         [650001.96, 300443.4,   18.7,   0.65,   650002.571, 300443.182, 18.745],
         [650002.95, 300442.54,  18.82,  0.473,  650003.056, 300442.085, 18.745]]

second = [[1],
          [2],
          [3]]

My expected output:

final = [[650001.88, 300442.2,   18.73,  0.575,  650002.094, 300441.668, 18.775, 1],
         [650001.96, 300443.4,   18.7,   0.65,   650002.571, 300443.182, 18.745, 2],
         [650002.95, 300442.54,  18.82,  0.473,  650003.056, 300442.085, 18.745, 3]]

To do that I create simple loop:

for i in first:
        for j in second:
            final += np.append(j, i)

I got i filling that i missing something. First of all my loop i extremely slow. Secondly my data is quite have i got more than 2 mlns rows to loop. So I tried to find faster way for example with this code:

final = [np.append(i, second[0]) for i in first]

It working far more faster than previous loop but its appending only first value of second array. Can you help me?

mforpe · Accepted Answer · 2017-04-22 16:35:45Z

38

Use np.array and then np.concatenate,

import numpy as np

first = np.array([[650001.88, 300442.2,   18.73,  0.575,  
                   650002.094, 300441.668, 18.775],
                  [650001.96, 300443.4,   18.7,   0.65,   
                   650002.571, 300443.182, 18.745],
                  [650002.95, 300442.54,  18.82,  0.473,  
                   650003.056, 300442.085, 18.745]])

second = np.array([[1],
                   [2],
                   [3]])

np.concatenate((first, second), axis=1)

Where axis=1 means that we want to concatenate horizontally.

That works for me

answered Apr 22, 2017 at 16:35

mforpe

1,5991 gold badge13 silver badges23 bronze badges

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

Comments

user9389968 · Accepted Answer · 2018-02-21 09:19:32Z

15

Use np.column_stack:

import numpy as np

first = [[650001.88, 300442.2,   18.73,  0.575,  650002.094, 300441.668, 18.775],
         [650001.96, 300443.4,   18.7,   0.65,   650002.571, 300443.182, 18.745],
         [650002.95, 300442.54,  18.82,  0.473,  650003.056, 300442.085, 18.745]]

second = [[1],
          [2],
          [3]]

np.column_stack([first, second])

If you need it as a list, use the method tolist:

np.column_stack([first, second]).tolist()

answered Feb 21, 2018 at 9:19

user9389968

Comments

user7864386 · Accepted Answer · 2022-02-13 01:19:49Z

2

For this case, hstack (because second is already 2D) and c_ (because it concatenates along the second axis) would also work. In fact c_ would work even if second is shape (3,), as long as its length matches the length of first.

Assuming first and second are already numpy array objects:

out = np.c_[first, second]

or

out1 = np.hstack((first, second))

Output:

assert (out == np.array(final)).all() & (out == out1).all()

That being said, all are just different ways of using np.concatenate.

answered Feb 13, 2022 at 1:19

user7864386

Comments

Vincent Taing · Accepted Answer · 2020-11-18 04:19:27Z

0

Multiply a 5x3 matrix by a 3x2 matrix (real matrix product)

import numpy as np
np_1= np.arange(15).reshape(5,3)
np_1

edited Nov 18, 2020 at 4:19

Vincent Taing

3,3342 gold badges21 silver badges25 bronze badges

answered Jun 9, 2020 at 19:26

Pranjali Khandelwal

11 bronze badge

Collectives™ on Stack Overflow

Merge two numpy arrays

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related