Unpack lists that are elements of a np.array in Python

Question

Anybody know how I can "unpack" lists which are elements/cells of a np.array? This is what the output looks like now:

[[0 301227 0.863 0.456 -3.551 0.532 135.96200000000002 4
  list([0.49, 0.33, 0.33, 0.28, 0.26, 0.47, 0.28, 0.29, 0.41, 0.42, 0.85, 0.48])
  list([53.09, 119.1, 36.67, -1.91, 11.38, -20.86, -0.56, 4.01, -3.51, 0.9, -8.72, 1.04])
  66.66666666666667 1.0]]

I want to unpack those lists so that each value is just another item/column in my array. I can also make the lists into np.arrays if that helps. Should be an easy fix, I just haven't found it yet.

Cheers

Welcome to StackOverflow! Please take the time to read this post on how to [How do I ask a good question?](https://stackoverflow.com/help/how-to-ask) as well as how to provide a [minimal, Complete, and Verifiable example](https://stackoverflow.com/help/mcve) and revise your question accordingly — yatu, Sep 09 '20 at 14:17

score 1 · Answer 1 · answered Sep 09 '20 at 15:04

I think I understand your issue a little better now. I created something to try and match the situation you're facing, and it returns a list where every item inside every list becomes just another item as you wanted:

import numpy as np

input_list = [0, 2, 4, 1, 5,np.array([1, 2, 3]), 6, 1, 6, 1, 2, np.array([1, 2, 3])] 

def flatten(l):
    for item in l:
        try:
            yield from flatten(item)
        except TypeError:
            yield item

output_list = list(flatten(input_list)))

Output is: [0, 2, 4, 1, 5, 1, 2, 3, 6, 1, 6, 1, 2, 1, 2, 3]

Tested it with np arrays and list as well, should have no problems. Also, I didn't create the output list as a np.array, but you can easily change it to do so if you wish

Quite elegant little hack you have there! It works really well. I'm now feeding individual rows of my original array to it, then re-assembling the rows later on in a NumPy array. — Alex, Sep 09 '20 at 15:16
Thank you! I just found a question that has a problem similar to yours, code similar to mine and other hacks that might have better performance! https://stackoverflow.com/questions/2158395/flatten-an-irregular-list-of-lists — Guilherme Freitas, Sep 09 '20 at 15:19

score 0 · Accepted Answer · edited Sep 09 '20 at 14:30

0

If I understood your question correctly, you can use numpy.concatenate() for that:

list = np.concatenate(input_list).ravel()

If you want the output to be in list form, you can append .tolist()

edited Sep 09 '20 at 14:30

Dharman

30,962
25
85
135

answered Sep 09 '20 at 14:24

Guilherme Freitas

26
4

1

Ah yes! It works when appending ```.tolist()``` . Thanks! – Alex Sep 09 '20 at 14:40
However, it doesn't unpack the lists it only makes the np.array structure a list. So it doesn't actually solve my problem. There are still lists within the list. – Alex Sep 09 '20 at 14:52

Unpack lists that are elements of a np.array in Python

2 Answers2