Adding an array to the end of another Python

Question

I'm very new to python and I have been faced with the task of taking several arrays into another array, this is inside of a loop. So if you had

a = np.array([2,3,4,3,4,4,5,3,2,3,4])

and

b = np.array([1,1,1,1,1,2,23,2,3,3,3])

and

c = np.array([])

and wanted the result

c = [[2,3,4,3,4,4,5,3,2,3,4],
     [1,1,1,1,1,2,23,2,3,3,3]]

so if I did c[0,:] I would get [2,3,4,3,4,4,5,3,2,3,4] I tried using c = [c, np.array(a)] then next iteration you get c = [c, np.array(b)] but I i do c[0,:] i get the error message list indices must be integers not tuples

EDIT:

When I get it to print out c it gives [array([2,3,4,3,4,4,5,3,2,3,4],dtype = unit8)]

Do you have any ideas?

I would highly recommend not doing this in a loop, especially for large arrays! Every time you concatenate two numpy arrays together, the result is a new copy of the array. Generating the copy becomes slower and slower as the array gets larger and larger on each loop iteration. If you must do your appending inside a loop then I suggest you use a regular Python list, and convert it to a numpy array outside the loop (unlike numpy arrays, you can append to Python lists in place without generating a copy). — ali_m
– ali_m, Commented Nov 30, 2015 at 18:08

Randy · Accepted Answer · 2015-11-30 16:53:21Z

5

In [10]: np.vstack((a,b))
Out[10]:
array([[ 2,  3,  4,  3,  4,  4,  5,  3,  2,  3,  4],
       [ 1,  1,  1,  1,  1,  2, 23,  2,  3,  3,  3]])

EDIT: Here's an example of using it in a loop to gradually build a matrix:

In [14]: c = np.random.randint(0, 10, 10)

In [15]: c
Out[15]: array([9, 5, 9, 7, 3, 0, 1, 9, 2, 0])

In [16]: for _ in xrange(10):
   ....:     c = np.vstack((c, np.random.randint(0, 10, 10)))
   ....:

In [17]: c
Out[17]:
array([[9, 5, 9, 7, 3, 0, 1, 9, 2, 0],
       [0, 8, 1, 9, 7, 5, 4, 2, 1, 2],
       [2, 1, 4, 2, 9, 6, 7, 1, 3, 2],
       [6, 0, 7, 9, 1, 9, 8, 5, 9, 8],
       [8, 1, 0, 9, 6, 6, 6, 4, 8, 5],
       [0, 0, 5, 0, 6, 9, 9, 4, 6, 9],
       [4, 0, 9, 8, 6, 0, 2, 2, 7, 0],
       [1, 3, 4, 8, 2, 2, 8, 7, 7, 7],
       [0, 0, 4, 8, 3, 6, 5, 6, 5, 7],
       [7, 1, 3, 8, 6, 0, 0, 3, 9, 0],
       [8, 5, 7, 4, 7, 2, 4, 8, 6, 7]])

edited Nov 30, 2015 at 16:53

answered Nov 30, 2015 at 15:53

Randy

14.9k3 gold badges40 silver badges44 bronze badges

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

6 Comments

Neo Streets Over a year ago

Hi when I do this getting array([[2],[1]]) instead of all of the elements

Randy Over a year ago

I'd recommend checking the code around it. np.vstack just vertically stacks the arrays you pass to it, and so something else in your code may be cutting off the rest of the results inadvertently.

Neo Streets Over a year ago

what i'm trying to do add them to a 3rd array c

Randy Over a year ago

Added an edit to show an example of doing it in a loop. Hope that helps.

Neo Streets Over a year ago

Thanks this is brilliant, a lot of help, and easy to follow

|

Kasravnd · Accepted Answer · 2015-11-30 15:54:09Z

3

Most numpythonic way is using np.array:

>>> c = np.array((a,b))
>>> 
>>> c
array([[ 2,  3,  4,  3,  4,  4,  5,  3,  2,  3,  4],
       [ 1,  1,  1,  1,  1,  2, 23,  2,  3,  3,  3]])

answered Nov 30, 2015 at 15:54

Kasravnd

108k19 gold badges167 silver badges195 bronze badges

Comments

ZdaR · Accepted Answer · 2015-11-30 15:54:36Z

0

You may try this:

>>> c = [list(a), list(b)]
>>> c
[[2, 3, 4, 3, 4, 4, 5, 3, 2, 3, 4], [1, 1, 1, 1, 1, 2, 23, 2, 3, 3, 3]]

answered Nov 30, 2015 at 15:54

ZdaR

23.1k7 gold badges70 silver badges90 bronze badges

Comments

thomas · Accepted Answer · 2015-11-30 15:55:15Z

You can concatenate arrays in numpy. For this to work, they must have the same size in all dimensions except the concatenation direction.

If you just say

>>> c = np.concatenate([a,b])

you will get

>>> c
array([ 2,  3,  4,  3,  4,  4,  5,  3,  2,  3,  4,  1,  1,  1,  1,  1,  2,
   23,  2,  3,  3,  3])

So in order to achieve what you want you first have to add another dimension to your vectors a and b like so

>>> a[None,:] 
array([[2, 3, 4, 3, 4, 4, 5, 3, 2, 3, 4]])

or equivalently

>>> a[np.newaxis,:]
array([[2, 3, 4, 3, 4, 4, 5, 3, 2, 3, 4]])

So you could do the following:

>>> c = np.concatenate([a[None,:],b[None,:]],axis = 0)
>>> c
array([[ 2,  3,  4,  3,  4,  4,  5,  3,  2,  3,  4],
   [ 1,  1,  1,  1,  1,  2, 23,  2,  3,  3,  3]])

Collectives™ on Stack Overflow

Adding an array to the end of another Python

4 Answers 4

6 Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

6 Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related