Davis Vaughan

This has come up one other time, in tidyverse/dplyr#5525 I am pretty sure I'm of the camp that this should continue to not be supported, since Date, POSIXct, and factor...

Sounds good tidyverse/dplyr#5525, at least, was manual setting of `dim()` > I do not even remember why I had added that dimension statement in the code that made me stumble...

tidyverse/dplyr#5779 uses data from this repo https://github.com/NOAA-EDAB/neus-atlantis/, which also seems to use {ncdf4} a lot, so I bet it is the same underlying issue. We should figure out why that...

Okay @lionel- we now have: ``` vec_duplicate_any(x) vec_duplicate_all(x) vec_duplicate_flg(x, first = FALSE) vec_duplicate_loc(x, first = FALSE) ``` With a nice relationship of: ``` which(vec_duplicate_flg(x)) == vec_duplicate_loc(x) which(vec_duplicate_flg(x, first = TRUE))...

Postponing until 0.3.0 until we can figure out suffix naming for `flg`

It has come up elsewhere that we could also use the naming scheme: - `vec_duplicate_detect` -> `vec_detect_duplicate()` - `vec_duplicate_id()` -> `vec_locate_first()` - `vec_duplicate_loc()` -> `vec_locate_duplicate()` (This nicely avoids `vec_duplicate_flg()`) I...

Tracking in https://docs.google.com/spreadsheets/d/1lSrVbq6CQICpX1IQcamyPEhkcw232Um9TLWEjV00Kfw/edit#gid=0 We will come back to these as we have an explicit need for them, or for the 1.0 release

I now think it would be useful to have: ```r # error if size 0, requires size >0 to be able to return a single location back vec_arg_min(x) vec_arg_max(x) #...

I also like that Python's `min()` function has a `default` argument https://docs.python.org/3/library/functions.html#min So maybe its something like: ```r vec_min # A tibble: 2 × 2 #> g date #> #>...

`vec_min(x, empty = 1L)` would probably be useful for https://github.com/tidyverse/dplyr/issues/6167#issuecomment-1024533145