I am looking to remove nested list columns from a tibble with lists. For example in the list below I want to remove "affected_rows" without specifically referencing the column name. How would i best approach this?
my_list <- list(
a =
tibble(
code = c("ax","yz"),
affected_rows =
c(list(1:10),list(200))
),
b =
tibble(
workid = c("123","456"),
sheet = c("sheet1", "sheet2")
)
)
We can use select after looping over the list
library(dplyr)
library(purrr)
my_list <- map(my_list, ~ .x %>%
select(-where(is.list)))
-output
my_list
$a
# A tibble: 2 × 1
code
<chr>
1 ax
2 yz
$b
# A tibble: 2 × 2
workid sheet
<chr> <chr>
1 123 sheet1
2 456 sheet2
Or negate the output of where
my_list <- map(my_list, ~.x %>%
select(negate(where(is.list))))
my_list
$a
# A tibble: 2 × 1
code
<chr>
1 ax
2 yz
$b
# A tibble: 2 × 2
workid sheet
<chr> <chr>
1 123 sheet1
2 456 sheet2
Or a more compact option with discard
map(my_list, discard, is.list)
$a
# A tibble: 2 × 1
code
<chr>
1 ax
2 yz
$b
# A tibble: 2 × 2
workid sheet
<chr> <chr>
1 123 sheet1
2 456 sheet2
Or may use atomic_elem from collapse
library(collapse)
map(my_list, atomic_elem, keep.class = TRUE)
$a
# A tibble: 2 × 1
code
<chr>
1 ax
2 yz
$b
# A tibble: 2 × 2
workid sheet
<chr> <chr>
1 123 sheet1
2 456 sheet2
Another possible solution:
library(tidyverse)
map(my_list, ~ .x %>% mutate(across(where(is.list), as.null)))
#> $a
#> # A tibble: 2 × 1
#> code
#> <chr>
#> 1 ax
#> 2 yz
#>
#> $b
#> # A tibble: 2 × 2
#> workid sheet
#> <chr> <chr>
#> 1 123 sheet1
#> 2 456 sheet2
in base R:
lapply(my_list, Filter, f = Negate(is.list))
$a
# A tibble: 2 x 1
code
<chr>
1 ax
2 yz
$b
# A tibble: 2 x 2
workid sheet
<chr> <chr>
1 123 sheet1
2 456 sheet2