Assigning the name in a nested list to it's subcomponents

user_13821482 0
Viewed 52

I have a nested list current_output that I would like to unlist and obtain the expected output. I would like to assign the name of the list in current_output to it's sub-components. For example, I would like to assign the name A to the three sub-components in current_output, and similarly for B and C to match those in the `expected.

A1 <- c(1:3)
A2 <- c(4:6)
A3 <- c(7:9)
B1 <- c(1:2)
B2 <- c(3:4)
B3 <- 1
C1 <- c(1:3)
C2 <- 1
C3 <- 1

l1 <- list(list(A1), list(A2), list(A3))
l2 <- list(list(B1),list(B2), list(1))
l3 <- list(list(C1), list(1), list(1))

current_output <- list(l1, l2, l3)
names(current_output) <- c("A", "B", "C")


expected <- list(A1, A2, A3, B1, B2, B3, C1, C2, C3)
names(expected) <- c(rep("A", 3), rep("B", 3), rep("C", 3))
names r list
3 Answers

It is better to have unique key i.e. names for the list elements - loop over the nested list, flatten, then invoke by concatenating again, and change the names by removing the suffix digits from the names

library(purrr)
library(dplyr)
library(stringr)
out <- current_output %>% 
   map(flatten) %>%   
   invoke(c, .) %>%  
   setNames(str_remove(names(.), "\\d+"))

-testing

> all.equal(out, expected)
[1] TRUE

Or another option is rrapply

rrapply::rrapply(current_output, how = 'flatten') %>% 
    setNames(rep(names(current_output), lengths(current_output)))

Or using base R, unlist the list elements, convert each element to list element with as.list and then remove the digits from the names with sub and assign back the names

out <- as.list(unlist(current_output))
names(out) <- sub("\\d+$", "", names(out))
user_37322710

Another base R approach that uses multiple unlist() steps.

setNames(
  unlist(current_output, recursive = F) %>% unlist(recursive = F), 
  rep(names(current_output), each = 3)
  )

Output

$A
[1] 1 2 3

$A
[1] 4 5 6

$A
[1] 7 8 9

$B
[1] 1 2

$B
[1] 3 4

$B
[1] 1

$C
[1] 1 2 3

$C
[1] 1

$C
[1] 1
user_166474960

Another possible solution:

library(tidyverse)

current_output %>% imap(set_names) %>% flatten %>% flatten

#> $A
#> [1] 1 2 3
#> 
#> $A
#> [1] 4 5 6
#> 
#> $A
#> [1] 7 8 9
#> 
#> $B
#> [1] 1 2
#> 
#> $B
#> [1] 3 4
#> 
#> $B
#> [1] 1
#> 
#> $C
#> [1] 1 2 3
#> 
#> $C
#> [1] 1
#> 
#> $C
#> [1] 1

Or simply using a nested for loop in base R:

expected <- list()

for (i in seq_along(current_output))
  for (j in seq_along(current_output[[i]]))
  { 
    expected <- c(expected, current_output[[i]][[j]])
    names(expected)[length(expected)] <- names(current_output)[i]
  }

expected
user_115644870
Related