Python remove items from a list of lists i.e. remove items from sublists

Question

I have a list of lists and I am looking to remove items from the individual lists. The example list is as follows:

List = [[461, 'N', 'N', 547], [549, 'N', 'N', 623], [926, 'N', 'N', 'N', 'N', 1099], [1101, 'N', 'N', 'N', 'N', 1262], [1638, 'N', 'N', 'N', 'N', 1795], [1797, 'N', 'N', 'N', 'N', 1942], [2279, 'N', 'N', 2357], [2359, 'N', 'N', 2425], [2686, 'N', 'N', 2764], [2766, 'N', 'N', 2832], [3099, 'N', 'N', 3182], [3184, 'N', 'N', 3254], [3333, 'N', 'N', 3403], [3405, 'N', 'N', 3475], [3564, 'N', 'N', 3642], [3644, 'N', 'N', 3710]]

What is a way to remove all of the 'N' such that the resultant list would be:

[[461, 547], [549, 623], [926, 1099], [1101, 1262], [1638, 1795], [1797, 1942], [2279, 2357], [2359, 2425], [2686, 2764], [2766, 2832], [3099, 3182], [3184, 3254], [3333, 3403], [3405, 3475], [3564, 3642], [3644, 3710]]

For now the individual lists will always start and end with an integer. In between will be a combination of "Y"s and "N"s. I have done some logic filtering and now need to simplify the lists to only contain the numbers.

I have tried list comprehensions unsuccessfully. I also tried to implement a function to recursively replace the letters but I am trying to actually remove the items not replace, (see: https://stackoverflow.com/a/13782720/20786144).

https://stackoverflow.com/questions/18072759/comprehension-on-a-nested-iterables/45079294 — Rahul, Dec 20 '22 at 05:17
using `List` as variable name is not good Idea. 1. Its not convention to use PascalCase for variable names. Use snake_case. 2. It may conflict with `typing.List` — Rahul, Dec 20 '22 at 05:23

Rahul · Accepted Answer · 2022-12-20T05:26:22.050

3

You can easily achieve this using listcomp:

exclusion_list = {"Y", "N"} # Using set is performance+
[[i for i in subl if i not in exclusion_list] for subl in List]

Which is equivalent of:

nlist = []
for subl in List:
    nsubl = []
    for i in subl:
        if i not in exclusion_list:
            nsubl.append(i)
    nlist.append(nsubl)

edited Dec 20 '22 at 05:26

answered Dec 20 '22 at 05:13

Rahul

10,830
4
53
88

Thank you for the this explanation, I was working through a listcomp unsuccessfully but with your explanation I know see what it wasn't working. And thank you for the additional link above! – tmckinney Dec 20 '22 at 05:30
If your question got resolved, you can accept this answer (by clicking check mark below score). – Rahul Dec 21 '22 at 07:23

score 2 · Answer 2 · answered Dec 20 '22 at 05:46

It sounds like the rules may change with this list. Maybe what you want is to keep all integers. You could visit each sublist in a loop, filter with a list comprehension and assign back to the original list with slicing.

for l in List:
    l[:] = [v for v in l if isinstance(v, int)]

CrazyChucky · Answer 3 · 2022-12-20T13:35:18.093

1

The first item of a list is index 0. Negative indexes count from the end, and the last item of a list is index -1. So to get the first and last items of each list, you could do this (assuming the starting list is named lst):

output = []
for sublist in lst:
    output.append([sublist[0], sublist[-1]])

Which can be condensed to a list comprehension:

output = [[sublist[0], sublist[-1]] for sublist in lst]

Or, using tuple unpacking:

output = [[first, last] for first, *_, last in lst]

edited Dec 20 '22 at 13:35

answered Dec 20 '22 at 05:19

CrazyChucky

3,263
4
11
25

Thank you for this answer, it definitely is a solution for the given example and is very powerful should the sorting/indices hold! Thanks again – tmckinney Dec 20 '22 at 05:40

ScottC · Answer 4 · 2022-12-20T05:28:35.667

You can use remove:

my_list = [[461, 'N', 'N', 547], [549, 'N', 'N', 623], [926, 'N', 'N', 'N', 'N', 1099], [1101, 'N', 'N', 'N', 'N', 1262], [1638, 'N', 'N', 'N', 'N', 1795], [1797, 'N', 'N', 'N', 'N', 1942], [2279, 'N', 'N', 2357], [2359, 'N', 'N', 2425], [2686, 'N', 'N', 2764], [2766, 'N', 'N', 2832], [3099, 'N', 'N', 3182], [3184, 'N', 'N', 3254], [3333, 'N', 'N', 3403], [3405, 'N', 'N', 3475], [3564, 'N', 'N', 3642], [3644, 'N', 'N', 3710]]

for x in my_list:
    while 'N' in x:
        x.remove('N')
    while 'Y' in x:
        x.remove('Y')
        
print(my_list)

Output:

[[461, 547], [549, 623], [926, 1099], [1101, 1262], [1638, 1795], [1797, 1942], [2279, 2357], [2359, 2425], [2686, 2764], [2766, 2832], [3099, 3182], [3184, 3254], [3333, 3403], [3405, 3475], [3564, 3642], [3644, 3710]]

While I doubt this process is going to be a performance bottleneck, that's a lot of iterating. Every `in` and every `remove` is going to keep checking and rechecking every first integer to see if it's an `'N'`. (Also the asker mentioned there can be `'Y'`.) — CrazyChucky, Dec 20 '22 at 05:21

GaryMBloom · Answer 5 · 2022-12-20T15:39:00.053

-2

Many people like using list comprehensions as a way to make their code shorter. Assuming you're familiar enough with Python, list comprehensions are a good way to go.

list_2 = [ [elem for elem in sub_list if not isinstance(elem, str)] for sub_list in List]

This makes for compact code, but it's not always easy for newer programmers to follow. Expanding it out into steps may be a better idea.

edited Dec 20 '22 at 15:39

answered Dec 20 '22 at 06:37

GaryMBloom

5,350
1
24
32

Python remove items from a list of lists i.e. remove items from sublists

5 Answers5

Output: