Reducing the dimensions of an array in a list in Python

Question

I am reshaping the array in a list Test from (1, 3, 3) to (3, 3). How do I reshape for a more general form, say for a very numpy array from (1, n, n) to (n, n)?

import numpy as np

Test = [np.array([[[1, 2, 3], [4, 5, 6], [7, 8, 9]]])]
Test = Test[0].reshape(3, 3)

"I am reshaping the list Test from (1,3,3) to (3,3)." No, you are not. You are reshaping the Numpy Array inside the list. — Klaus D.
– Klaus D., Commented Aug 29, 2022 at 10:57
In your case Test = Test[0][0] would give you (n,n) for any (1,n,n) — Yevhen Kuzmovych
– Yevhen Kuzmovych, Commented Aug 29, 2022 at 10:58

mmdanziger · Accepted Answer · 2022-08-29 11:06:02Z

2

The list is not relevant. The simplest way to reshape to the smallest valid shape is squeeze:

Test = np.array([[[1, 2, 3], [4, 5, 6], [7, 8, 9]]])
assert Test.shape == (1, 3, 3)
Test = Test.squeeze()
assert Test.shape == (3, 3)

By smallest valid size, I mean to eliminate all dimensions that have length 1. You can customize it to only pick specific axes to zero out, but in practice, I find the default behavior is most useful. A super-useful feature of squeeze is that it's idempotent. You can keep "squeezing" an array as many times as you want.

Bonus: The same function exists in pandas pd.DataFrame.squeeze where it gives you a pd.Series from a single column pd.DataFrame.

edited Aug 29, 2022 at 11:06

answered Aug 29, 2022 at 10:59

mmdanziger

4,6982 gold badges36 silver badges53 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Reducing the dimensions of an array in a list in Python

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related