如何合并两个numpy数组的数字

Question

Strange question, I can concat two digits by int(str(2)+str(3)) but how to this for two numpy arrays? Like

x = np.array([[1,2,3],[4,5,6]])
y = np.array([[4,5,6],[1,2,3]])
z = np.xx(x,y)
print(z)

array([[14, 25, 36],
       [41, 52, 63]])

Answer 1

Here is a pure numpy solution, which does not involve any mappings to other datastructures. It should be far faster than the list comprehension, especially for large matrices.

import numpy as np

x = np.array([[10,2,3],[4,5,6]])
y = np.array([[4,5,6],[1,2,3]])

digits = np.log10(y).astype(np.int)+1

z = x*(10**digits)+y
print z

I changed the 1 to a 10 to show it works for multiple digits; but if your numbers are always in the range 0-9, as in your example, you can of course completely do away with the digit logic.

Answer 2

You can use zip and a list comprehension:

In [7]: np.array([[str(c) + str(d) for c, d in zip(a, b)] for a, b in zip(x, y)],
                                                                          dtype=int)     
Out[7]: 
array([[14, 25, 36],                                                                             
       [41, 52, 63]])    

or:

In [20]: np.array([str(a)+str(b) for a, b in zip(*np.hstack((x, y)))],
                                                        dtype=int).reshape(x.shape)                                                                                                
Out[20]: 
array([[14, 25, 36],                                                                             
       [41, 52, 63]])