Unpacking numpy array inside

Question

I have the following numpy array:

a=[['Sb' array([4.24035696, 2.44817292, 7.41858935])]
 ['I' array([2.08076032, 3.69501392, 5.37518666])]
 ['I' array([6.35173963, 1.22916823, 8.9947238 ])]
 ['I' array([ 4.24036048, -0.04551256,  5.37518684])]
 ['I' array([4.24035383, 4.88618543, 8.99472472])]
 ['I' array([6.3999514 , 3.69501663, 5.37518685])]
 ['I' array([2.12897688, 1.22916548, 8.99472467])]]

I'd like to unpack the inner array to have:

a=[['Sb' 4.24035696 2.44817292 7.41858935]
 ...
 ['I' 2.12897688 1.22916548 8.99472467]]

I can't figure this one out in a simple way.

What have you tried? It isn't a simple structure; so don't expect a simple fix. It's a (7,2) object dtype array. a[:,0] is the characters column. a[:,1] the arrays. np.stack(a[:,1]) will give you a (7,3) float array. That could be concatenated with the character one, but getting the right dtype could be tricky. String and floats don't coexist in one array easily. — hpaulj
– hpaulj, Commented Jun 18, 2022 at 16:08

hpaulj · Accepted Answer · 2022-06-18 16:14:45Z

See my comment for explanation:

In [142]: array=np.array
     ...: a=np.array([['Sb', array([4.24035696, 2.44817292, 7.41858935])],
     ...:  ['I', array([2.08076032, 3.69501392, 5.37518666])],
     ...:  ['I', array([6.35173963, 1.22916823, 8.9947238 ])],
     ...:  ['I', array([ 4.24036048, -0.04551256,  5.37518684])],
     ...:  ['I', array([4.24035383, 4.88618543, 8.99472472])],
     ...:  ['I', array([6.3999514 , 3.69501663, 5.37518685])],
     ...:  ['I', array([2.12897688, 1.22916548, 8.99472467])]], dtype=object)

In [143]: a
Out[143]: 
array([['Sb', array([4.24035696, 2.44817292, 7.41858935])],
       ['I', array([2.08076032, 3.69501392, 5.37518666])],
       ['I', array([6.35173963, 1.22916823, 8.9947238 ])],
       ['I', array([ 4.24036048, -0.04551256,  5.37518684])],
       ['I', array([4.24035383, 4.88618543, 8.99472472])],
       ['I', array([6.3999514 , 3.69501663, 5.37518685])],
       ['I', array([2.12897688, 1.22916548, 8.99472467])]], dtype=object)

In [144]: a[:,0]
Out[144]: array(['Sb', 'I', 'I', 'I', 'I', 'I', 'I'], dtype=object)

In [145]: np.stack(a[:,1])
Out[145]: 
array([[ 4.24035696,  2.44817292,  7.41858935],
       [ 2.08076032,  3.69501392,  5.37518666],
       [ 6.35173963,  1.22916823,  8.9947238 ],
       [ 4.24036048, -0.04551256,  5.37518684],
       [ 4.24035383,  4.88618543,  8.99472472],
       [ 6.3999514 ,  3.69501663,  5.37518685],
       [ 2.12897688,  1.22916548,  8.99472467]])

Combine them into an object dtype array:

In [148]: res = np.concatenate((a[:,[0]],np.stack(a[:,1])),axis=1,dtype=object)

In [149]: res
Out[149]: 
array([['Sb', 4.24035696, 2.44817292, 7.41858935],
       ['I', 2.08076032, 3.69501392, 5.37518666],
       ['I', 6.35173963, 1.22916823, 8.9947238],
       ['I', 4.24036048, -0.04551256, 5.37518684],
       ['I', 4.24035383, 4.88618543, 8.99472472],
       ['I', 6.3999514, 3.69501663, 5.37518685],
       ['I', 2.12897688, 1.22916548, 8.99472467]], dtype=object)

Yes not straightforward, I tried a few functions to combine the separate arrays but the np.stack(a[:,1]) was the important trick missing. Thank you very much.

Collectives™ on Stack Overflow

Unpacking numpy array inside

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related