Combine/merge lists by elements names

I have two lists, whose elements have partially overlapping names, which I need to merge/combine together into a single list, element by element:

> lst1 <- list(integers=c(1:7), letters=letters[1:5],
                words=c("two", "strings"))
> lst2 <- list(letters=letters[1:10], booleans=c(TRUE, TRUE, FALSE, TRUE),
                words=c("another", "two"), floats=c(1.2, 2.4, 3.8, 5.6))

> lst1
$integers
[1] 1 2 3 4 5 6 7

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

$words
[1] "two"     "strings"

> lst2
$letters
 [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

$booleans
[1]  TRUE  TRUE FALSE  TRUE

$words
[1] "another" "two"    

$floats
[1] 1.2 2.4 3.8 5.6

I tried using mapply, which basically combines the two lists by index (i.e.: "[["), while I need to combine them by name (i.e.: "$"). In addition, since the lists have different lengths, the recycling rule is applied (with rather unpredictable results).

> mapply(c, lst1, lst2)
$integers
 [1] "1" "2" "3" "4" "5" "6" "7" "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

$letters
[1] "a"     "b"     "c"     "d"     "e"     "TRUE"  "TRUE"  "FALSE" "TRUE" 

$words
[1] "two"     "strings" "another" "two"    

$<NA>
 [1] 1.0 2.0 3.0 4.0 5.0 6.0 7.0 1.2 2.4 3.8 5.6

Warning message:
In mapply(c, lst1, lst2) :
  longer argument not a multiple of length of shorter

As you might imagine, what I'm looking for is:

$integers
[1] 1 2 3 4 5 6 7

$letters
[1] "a" "b" "c" "d" "e" "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

$words
[1] "two"     "strings"   "another" "two"

$booleans
[1]  TRUE  TRUE FALSE  TRUE

$floats
[1] 1.2 2.4 3.8 5.6

Is there any way to achieve that? Thank you!


Solution 1:

You can do:

keys <- unique(c(names(lst1), names(lst2)))
setNames(mapply(c, lst1[keys], lst2[keys]), keys)

Generalization to any number of lists would require a mix of do.call and lapply:

l <- list(lst1, lst2, lst1)
keys <- unique(unlist(lapply(l, names)))
setNames(do.call(mapply, c(FUN=c, lapply(l, `[`, keys))), keys)

Solution 2:

An update of flodel's answer for tidyverse users:

list1 <- list(integers=c(1:7), letters=letters[1:5],
               words=c("two", "strings"))
list2 <- list(letters=letters[1:10], booleans=c(TRUE, TRUE, FALSE, TRUE),
               words=c("another", "two"), floats=c(1.2, 2.4, 3.8, 5.6))

input_list <- list(list1, list2, list1, list2)

We want to replicate the original desired output exactly twice for each element in the output list. Using map2 and reduce, we can achieve that with a little bit more clarity than the base R solution involving do.call, mapply, and lapply. First, we declare a function that combines two lists by their named elements using c(), then we call our function on the input list via reduce:

library(purrr)

cat_lists <- function(list1, list2) {  

  keys <- unique(c(names(list1), names(list2)))
  map2(list1[keys], list2[keys], c) %>% 
    set_names(keys)  

}

combined_output <- reduce(input_list, cat_lists)

Which gives us what we want:

> combined_output

#> $integers
#>  [1] 1 2 3 4 5 6 7 1 2 3 4 5 6 7
#> 
#> $letters
#>  [1] "a" "b" "c" "d" "e" "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "a" "b"
#> [18] "c" "d" "e" "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"
#> 
#> $words
#> [1] "two"     "strings" "another" "two"     "two"     "strings" "another"
#> [8] "two"    
#> 
#> $booleans
#> [1]  TRUE  TRUE FALSE  TRUE  TRUE  TRUE FALSE  TRUE
#> 
#> $floats
#> [1] 1.2 2.4 3.8 5.6 1.2 2.4 3.8 5.6