`Vec::rdedup(_by(_key))`?

libs
1

Vec::dedup related methods "removes all but the first of consecutive elements in the vector". Is there a need for a method removes all but the last equal items?

2

What is the use case?

1 Like
3

I have a sorted Vec like [(1, 3), (1, 4), (2, 3)], each item is a (price, value) pair. I want to know for all things of the same price, how much value can I get.

A reasonable thought is v.dedup_by_key(|current, previous| current.0 == previous.0 && current.1 >= previous.1), but it will always remove the previous one according to the doc.

If rdedup_by_key is provided, the code above can be used, with dedup changed to rdedup.

I can call reverse -> dedup -> reverse to achieve the same result, but it would be so costly.

4

One alternative solution would be sorting them in reverse order.

5

It's probably worth while to check out some kind of DataFrame. I don't think it's worthwhile to add this specific table operation to Vec on its own, and if multiple such operations were to be added they might be better grouped by a different abstraction altogether.

6

Vec::dedup_by gives the closure &mut-access to both items, so you're free to modify the item that will be kept.

v.dedup_by(|next, prev| {
    next.0 == prev.0 && {
        std::mem::swap(next, prev);
        true
    }
})
6 Likes
7

The process for such proposals is to submit an ACP.

8

Oh I really like this idea