CODEWITHSUNDEEP
pythonarraysnumpy
ben890
ben890

Reputation: 1133

turn list of python arrays into an array of python lists

I have a list of python arrays like the following:

[array([[0., 0., 0.]]),
 array([[0., 0., 0.]]),
 array([[0., 0., 0.]])]

My goal is to change them to an array of lists like the following:

array([[ 0, 0, 0],
       [ 0, 0, 0],
       [ 0, 0, 0]])

I tried the following however I'm getting I'm an error:

np.array([my_array[i].tolist() for i in my_array])

The following is the error:

TypeError: only integer scalar arrays can be converted to a scalar index

Could anyone help me understand what's going and what I'm doing wrong.

Upvotes: 0

Views: 81

Answers (2)

J.Z
J.Z

Reputation: 419

Try: np.array([i[0] for i in my_array])

Since in your case, i is not the index of a, it represents each element in my_array

And each of your element in my_array is a nested list, you may want to flat the nested list when doing the conversion

Upvotes: 0

modesitt
modesitt

Reputation: 7210

You can just concatenate row-wise with vstack since they are of shape (1, 3) in your example to get the result of shape (n, 3).

np.vstack(my_array)

Why your current code fails is because you are iterating over every sub-array in my_array and trying to index my_array with that. So you are indexing a list with array([[0., 0., 0.]]) which is not a scaler index because it is an array.

Alternative options:

np.array(my_array).reshape(-1, 3)
np.array([e.reshape(-1) for e in my_array])
...

Upvotes: 1

Related Questions