reshape the array with specific form-CodePudding

I have an specific array which each rows has to array. I want to reshape it. But, I don't know how to reshape it to a 2d array. Here is my array:

x = np.array([[[array([0.23006942, 0.29322573, 0.37292298]),
     array([0.23006942])],
    [array([0.29322573, 0.37292298, 0.23006942]),
     array([0.32480389])],
    [array([0.37292298, 0.23006942, 0.32480389]),
     array([0.31427784])]],

   [[array([0.09971349, 0.08827682, 0.0900638 ]),
     array([0.02251595])],
    [array([0.08827682, 0.0900638 , 0.02251595]),
     array([0.0986413])],
    [array([0.0900638 , 0.02251595, 0.0986413 ]),
     array([0.02144376])]]], dtype=object)

Here is the desired output:

 out = np.array(

[
[array([0.23006942, 0.29322573, 0.37292298]),array([0.23006942])],

[array([0.29322573, 0.37292298, 0.23006942]),array([0.32480389])],

[array([0.37292298, 0.23006942, 0.32480389]),array([0.31427784])],

[array([0.09971349, 0.08827682, 0.0900638 ]),array([0.02251595])],

[array([0.08827682, 0.0900638 , 0.02251595]),array([0.0986413])],

[array([0.0900638 , 0.02251595, 0.0986413 ]),array([0.02144376])]

])

Any help appreciated.

Thanks

I've tried to use the reshape. But, it does not solve.

CodePudding user response：

I think you can use reshape (in a correct way):

# Your code
import numpy as np
x = np.array([[[np.array([0.23006942, 0.29322573, 0.37292298]),
         np.array([0.23006942])],
        [np.array([0.29322573, 0.37292298, 0.23006942]),
         np.array([0.32480389])],
        [np.array([0.37292298, 0.23006942, 0.32480389]),
         np.array([0.31427784])]],
    
       [[np.array([0.09971349, 0.08827682, 0.0900638 ]),
         np.array([0.02251595])],
        [np.array([0.08827682, 0.0900638 , 0.02251595]),
         np.array([0.0986413])],
        [np.array([0.0900638 , 0.02251595, 0.0986413 ]),
         np.array([0.02144376])]]], dtype=object)
# What you need
out = x.reshape(6,2)
out

Output

array([
       [array([0.23006942, 0.29322573, 0.37292298]), array([0.23006942])],
       [array([0.29322573, 0.37292298, 0.23006942]), array([0.32480389])],
       [array([0.37292298, 0.23006942, 0.32480389]), array([0.31427784])],
       [array([0.09971349, 0.08827682, 0.0900638 ]), array([0.02251595])],
       [array([0.08827682, 0.0900638 , 0.02251595]), array([0.0986413])],
       [array([0.0900638 , 0.02251595, 0.0986413 ]), array([0.02144376])]
      ],
      dtype=object)

CodePudding user response：

It's ragged, it's not concatenated. Is it a question of flattening the array to a list of arrays?

out = [x.tolist() for arr in x for x in arr]

Output:

[[array([0.23006942, 0.29322573, 0.37292298]), array([0.23006942])],
 [array([0.29322573, 0.37292298, 0.23006942]), array([0.32480389])],
 [array([0.37292298, 0.23006942, 0.32480389]), array([0.31427784])],
 [array([0.09971349, 0.08827682, 0.0900638 ]), array([0.02251595])],
 [array([0.08827682, 0.0900638 , 0.02251595]), array([0.0986413])],
 [array([0.0900638 , 0.02251595, 0.0986413 ]), array([0.02144376])]]