as_tibble控制拼合级别

Question

In the example below I have a list where one element is a list with a sublist. It appears that as_tibble() is flattening down until the 3rd level of list before stopping. It would be nice to be able to control the flattening so I could preserve the entire list with sublist in element "c" of the example. I suspect some combination of map() , flatten() , etc. is needed, but cannot figure it out.

Issue Reproduced

as_tibble() returns 2 rows, but I really only want 1 row with a column corresponding to each list element ("a", "b", "c").

library(dplyr)
x <- list(a = 1, 
          b = 2, 
          c = list(x=list(y=3, 
                          z=4), 
                   xx=NULL))
parsed <- as_tibble(x)
parsed
#> # A tibble: 2 x 3
#>       a     b c         
#>   <dbl> <dbl> <list>    
#> 1    1.    2. <list [2]>
#> 2    1.    2. <NULL>

parsed[1,]$c
#> $x
#> $x$y
#> [1] 3
#> $x$z
#> [1] 4

parsed[2,]$c
#> $xx
#> NULL

Desired Behavior where I unnest only the first level

parsed <- as_tibble(x)
parsed
#> # A tibble: 1 x 3
#>       a     b c         
#>   <dbl> <dbl> <list>    
#> 1    1.    2. <list [2]>

parsed[1,]$c
#> $x
#> $x$y
#> [1] 3
#> $x$z
#> [1] 4
#> $xx
#> NULL

Answer 1

This is not elegant but here is one way to do it.

library(purrr)
x <- list(a = 1, 
          b = 2, 
          c = list(x=list(y=3, 
                          z=4), 
                   xx=NULL))

# define a function to preserve the list
unnest_one_level <- function(x){
  if(is.list(x)) list(x) else x
}

res <- x %>% map_dfr(unnest_one_level)
res
#> # A tibble: 1 x 3
#>       a     b c         
#>   <dbl> <dbl> <list>    
#> 1    1.    2. <list [2]>

res$c
#> [[1]]
#> [[1]]$x
#> [[1]]$x$y
#> [1] 3
#> 
#> [[1]]$x$z
#> [1] 4
#> 
#> [[1]]$xx
#> NULL

Answer 2

as.tibble will repeat each column recycle each column value till they all have a shared number of elements. Here you are a and b have length 1, but c has length 2. So it seems you really want to treat it as a length one list, you just need to nest it in a list. I think this will do what you want in this case at least.

parsed <- as_tibble(modify_if(x, ~length(.x)>1, list))