opposite of `bind_cols()` in R [duplicate]

Question

It is trivial to bind vectors into a data.frame with data.frame(), tibble(), or bind_cols():

library(tidyverse)
x <- 1:5
y <- letters[1:5]
d <- tibble(x, y)

What is the most efficient way to reverse this operation - i.e., take a dataframe and break it into a list with length equal to the ncol(d)? In the example above, the desired output would look like this:

> list(x = x, y = y)

$x
[1] 1 2 3 4 5

$y
[1] "a" "b" "c" "d" "e"

A single function that accomplishes this would be useful, for example, when piping a dataframe to pmap():

d %>%
  break_into_list() %>% # missing function here
  pmap(function(x, y) paste(x, y)) # scales to dataframe with ncol > 2

Answer 1

We could use unclass which also gives an attribute for 'row.names'

unclass(d)

---

Or with as.list

as.list(d)

If the names should be NULL, use setNames

setNames(as.list(d), NULL)

-output

[[1]]
[1] 1 2 3 4 5

[[2]]
[1] "a" "b" "c" "d" "e"

---

Regarding the use of pmap, we don't need to convert to list. Assuming that we need to loop over the rows (which the pmap does)

library(dplyr)
library(purrr)
d %>% 
    pmap(~ f(..1, ..2))

where ..1 and ..2 are the 1st and 2nd columns that are used as input to the function where the f will be applied on each row. If there are more columns and the function requires more inputs, use ..3, ..4 etc.