Collectives™ on Stack Overflow

Find centralized, trusted content and collaborate around the technologies you use most.

Teams

Q&A for work

Connect and share knowledge within a single location that is structured and easy to search.

numpy array concatenate: "ValueError: all the input arrays must have same number of dimensions"

How to concatenate these numpy arrays?

first np.array with a shape (5,4)

[[  6487    400 489580      0]
 [  6488    401 492994      0]
 [  6491    408 489247      0]
 [  6491    408 489247      0]
 [  6492    402 499013      0]]
second np.array with a shape (5,)
[  16.   15.   12.  12.  17. ]
final result should be
[[  6487    400    489580    0   16]
 [  6488    401    492994    0   15]
 [  6491    408    489247    0   12]
 [  6491    408    489247    0   12]
 [  6492    402    499013    0   17]]
I tried np.concatenate([array1, array2])
but i get this error
ValueError: all the input arrays must have same number of dimensions
What am I doing wrong?
                How the heck is that second array supposed to have shape (1,)? Is there some sort of weird thing with object arrays going on here?
– user2357112
                Feb 1, 2017 at 21:26
                Then your array is seriously messed up in some way, and you need to figure out what's going on.
– user2357112
                Feb 1, 2017 at 21:29
To use np.concatenate, we need to extend the second array to 2D and then concatenate along axis=1 -
np.concatenate((a,b[:,None]),axis=1)
Alternatively, we can use np.column_stack that takes care of it -
np.column_stack((a,b))
Sample run -
In [84]: a
Out[84]: 
array([[54, 30, 55, 12],
       [64, 94, 50, 72],
       [67, 31, 56, 43],
       [26, 58, 35, 14],
       [97, 76, 84, 52]])
In [85]: b
Out[85]: array([56, 70, 43, 19, 16])
In [86]: np.concatenate((a,b[:,None]),axis=1)
Out[86]: 
array([[54, 30, 55, 12, 56],
       [64, 94, 50, 72, 70],
       [67, 31, 56, 43, 43],
       [26, 58, 35, 14, 19],
       [97, 76, 84, 52, 16]])
If b is such that its a 1D array of dtype=object with a shape of (1,), most probably all of the data is contained in the only element in it, we need to flatten it out before concatenating. For that purpose, we can use np.concatenate on it too. Here's a sample run to make the point clear -
In [118]: a
Out[118]: 
array([[54, 30, 55, 12],
       [64, 94, 50, 72],
       [67, 31, 56, 43],
       [26, 58, 35, 14],
       [97, 76, 84, 52]])
In [119]: b
Out[119]: array([array([30, 41, 76, 13, 69])], dtype=object)
In [120]: b.shape
Out[120]: (1,)
In [121]: np.concatenate((a,np.concatenate(b)[:,None]),axis=1)
Out[121]: 
array([[54, 30, 55, 12, 30],
       [64, 94, 50, 72, 41],
       [67, 31, 56, 43, 76],
       [26, 58, 35, 14, 13],
       [97, 76, 84, 52, 69]])
                the shape of the second was because of this array2 = np.array(np.round(data[:,0]/20)) i fixed with array2 = np.array(np.round(data[:,0]/20)).astype(int)
– RaduS
                Feb 1, 2017 at 21:40
                @hpaulj Looks like it to me, but I'm not that familiar with np.column_stack. It is basically a 2d concatenate that special cases 1d inputs, right?
– Paul Panzer
                Feb 2, 2017 at 2:20
It is not related to the original question.
But it can be useful when adding columns vertically in a loop.
a = np.empty([0,1], dtype=float) 
b = np.array([[6487, 400, 489.580],
              [6488, 401, 492.994],
              [6491, 408, 489.247],
              [6491, 408, 489.247],
              [6492, 402, 499.013]])
for c in range(0, 3):
    col = b[:, c]
    col = col.reshape(len(col),1)
    a = np.vstack((a, col))
print(a)
----------------------------------------------
Output:
[[6487.   ]
 [6488.   ]
 [6491.   ]
 [6491.   ]
 [6492.   ]
 [ 400.   ]
 [ 401.   ]
 [ 408.   ]
 [ 408.   ]
 [ 402.   ]
 [ 489.58 ]
 [ 492.994]
 [ 489.247]
 [ 489.247]
 [ 499.013]]
        
Thanks for contributing an answer to Stack Overflow!
Please be sure to answer the question. Provide details and share your research!
But avoid …
Asking for help, clarification, or responding to other answers.
Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.