match arg like in pryr ? or `slice_first_by_key()` ?

[moodymudskipper]

pryr::join has a match = "all" arg that can also be "first".

it should be similar to %>% group_by(keys) %>% slice(1)

it could also be a pre-processor so we can apply it on any side.

slice_first_by_key() ?

pryr's doc mentions it's faster so there might be some optimisation to do but let's not focus on performance unless there's explicit demand

[moodymudskipper]

https://stackoverflow.com/questions/70661273/how-do-i-add-an-x-and-y-suffix-to-all-duplicate-variables-after-the-join

[moodymudskipper]

This is %>% summarise_by_keys(across()[1,]) equivalent to %>% summarise_by_keys(slice(across(), 1))

We also find the use case %>% summarise_by_keys(across()[which.max(some_col),]), equivalent to %>% summarise_by_keys(slice_max(across()))

And I am not sure we need slice_max_by_key(), slice_min_by_key(), slice_median_by_key()...

Maybe just add some examples using across() to introduce the idiom.