Specific array filtering - code optimization

Question

I have a nested list where all sublists have a length 2, eg:

{{1,2}, {1,1}, {3,4}, {3,5}, ...}

I need to filter it for two requirements:

I get a huge construction:

DeleteDuplicates[
            DeleteCases[list, e_ /; First@e == Last@e],
        First@#1 == First@#2 &
        ]

Could it be golfed, for example, including union of duplicates in DeleteCases?

You put code optimization in your title – What does this mean? What is wrong with your current code? It's really not a huge construction. What you can do is to use DeleteDuplicatesBy[..., First] instead of DeleteDuplicates[..., First@#1 == First@#2 &]. — Domen, Commented Apr 30 at 12:24
@Domen, yes, it's much better with DeleteDuplicatesBy! But may be is it possible to do all filtering with one function? — lesobrod, Commented Apr 30 at 12:27
You mean like DeleteDuplicates[list, First@#1 == First@#2 || First@#1 == Last@#1 || First@#2 == Last@#2 &]? — Domen, Commented Apr 30 at 12:38
Does it absolutely have to be one function? If not I feel like something like: $$ $$ del = DeleteDuplicatesBy[list, First]; $$ $$ Select[del, ! Equal @@ # &] $$ $$ is short and readable, and also gives good performance for large lists. — ydd, Commented Apr 30 at 13:27

lericr · Accepted Answer · 2024-04-30 13:38:19Z

2

Not sure what you consider optimal, but in response to your comment

is it possible to do all filtering with one function?

you can certainly compose the two operations into one function and even use a point-free style:

DeleteDuplicatesBy[First]@*Select[Apply[Unequal]]@list

answered Apr 30 at 13:38

lericr

31.7k2 gold badges20 silver badges75 bronze badges

Add a comment |

1 Answer 1