Concatenate NumPy 2D array with column (1D array)

Question

Suppose I have a 2D NumPy array values. I want to add new column to it. New column should be values[:, 19] but lagged by one sample (first element equals to zero). It could be returned as np.append([0], values[0:-2:1, 19]). I tried: Numpy concatenate 2D arrays with 1D array

temp = np.append([0], [values[1:-2:1, 19]])
values = np.append(dataset.values, temp[:, None], axis=1)

but I get:

ValueError: all the input array dimensions except for the concatenation axis 
must match exactly

I tried using c_ too as:

temp = np.append([0], [values[1:-2:1, 19]])
values = np.c_[values, temp]

but effect is the same. How this concatenation could be made. I think problem is in temp orientation - it is treated as a row instead of column, so there is an issue with dimensions. In Octave ' (transpose operator) would do the trick. Maybe there is similiar solution in NumPy?

Anyway, thank you for you time.

Best regards,
Max

If temp is shape (n,) (1d), you can make a 2d 'column' with temp[:,None] or temp.reshape(n,1). That can then be concatenate with a 2d (n,m) array, on axis=1. The key difference with Octave is that numpy arrays can be true 1d. — hpaulj
– hpaulj, Commented Aug 19, 2019 at 21:27

hpaulj · Accepted Answer · 2019-08-20 00:37:29Z

In [76]: values = np.arange(16).reshape(4,4)                                                                 
In [77]: temp = np.concatenate(([0], values[1:,-1]))                                                         
In [78]: values                                                                                              
Out[78]: 
array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11],
       [12, 13, 14, 15]])
In [79]: temp                                                                                                
Out[79]: array([ 0,  7, 11, 15])

This use of concatenate to make temp is similar to your use of append (which actually uses concatenate).

Sounds like you want to join values and temp in this way:

In [80]: np.concatenate((values, temp[:,None]),axis=1)                                                       
Out[80]: 
array([[ 0,  1,  2,  3,  0],
       [ 4,  5,  6,  7,  7],
       [ 8,  9, 10, 11, 11],
       [12, 13, 14, 15, 15]])

Again I prefer using concatenate directly.

David Kabii · Accepted Answer · 2019-08-19 22:19:09Z

0

You need to convert the 1D array to 2D as shown. You can then use vstack or hstack with reshaping to get the final array you want as shown:

a = np.array([[1, 2, 3],[4, 5, 6]])
b = np.array([[7, 8, 9]])
c = np.vstack([ele for ele in [a, b]])
print(c)
c = np.hstack([a.reshape(1,-1) for a in [a,b]]).reshape(-1,3)
print(c)

Either way, the output is:

[[1 2 3] [4 5 6] [7 8 9]]

Hope I understood the question correctly

answered Aug 19, 2019 at 22:19

David Kabii

6601 gold badge6 silver badges17 bronze badges

Collectives™ on Stack Overflow

Concatenate NumPy 2D array with column (1D array)

2 Answers 2

Comments

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Linked

Related