Return list of lists from foreach loop in R - r

Question

I have a function which returns a list of two objects (a list l and a number n). I want to loop over this function in a foreach loop.
create_lists <- function(){
l = sample(100, 5)
n = sample(100, 1)
return(list(l=l, n=n))}
Because create_lists has a list as ouput, this post told me to use a combine function which looks like this:
combine_custom <- function(list1, list2){
ls = c(list1$l, list2$l)
ns = c(list1$n, list2$n)
return(list(l = ls, n = ns))
}
So now my foreach loop looks like this:
m = foreach(i=1:5, .combine = combine_custom)%do%{
create_lists()}
My desired output would be:
m$l
[[1]]
[1] 100 25 86 21 28
[[2]]
[1] 78 37 79 41 61
[[3]]
[1] 73 22 78 94 13
[[4]]
[1] 15 28 76 78 52
[[5]]
[1] 32 93 92 2 1
m$n
[1] 52 56 3 79 82
But what I get is something like this:
$l
[1] 84 28 75 59 68 84 28 75 59 68
$n
[1] 31 91 18 98 39
So I have two problems:
1) Why is everything but two of the l lists dropped?
2) How can I make m$l to be a list of lists?
EDIT:
I tried another approach I got from here which does not use c:
combine_custom <- function(list1, list2){
ls = list1$l[[length(list1$l)+1]] = list(list2$l)
ns = c(list1$n, list2$n)
return(list(l = ls, n = ns))
}
But this gave the same result as described above, to be exact:
$l
$l[[1]]
[1] 65 84 48 81 82
$n
[1] 88 79 92 36 71

Answer 1

I have found another way which avoids the problem mentioned above, namely that combine has to create a new list first and later only append lists.
Also, the real function I am using actually returns a list of lists, so the following proved useful:
combine_custom <- function(list1, list2) {
if (plotrix::listDepth(list1$l) > plotrix::listDepth(list2$l)) {
ls <- c(list1$l, list(list2$l))
} else {
ls <- c(list(list1$l), list(list2$l))
}
ns <- c(list1$n, list2$n)
return(list(l = ls, n = ns))
}
This is not perfect if the function can return lists of varying nesting depths, but it works in my case.

Answer 2

The combine part is giving a lot of trouble, because on the first iteration, it needs to make a list out of two lists , but on the second iteration, it needs to append one list as an element to a list of lists.
Another approach (may or may not work depending on the size of your actual data/problem) is to use the purrr package for working with lists:
> m <- foreach(i=1:3)%do%{create_lists()}
> m
[[1]]
[[1]]$l
[1] 21 33 12 50 36
[[1]]$n
[1] 74
[[2]]
[[2]]$l
[1] 12 80 39 78 6
[[2]]$n
[1] 74
[[3]]
[[3]]$l
[1] 9 61 75 63 94
[[3]]$n
[1] 2
> purrr::transpose(m)
$l
$l[[1]]
[1] 21 33 12 50 36
$l[[2]]
[1] 12 80 39 78 6
$l[[3]]
[1] 9 61 75 63 94
$n
$n[[1]]
[1] 74
$n[[2]]
[1] 74
$n[[3]]
[1] 2
Hope that helps!

Answer 3

Thank you #Maria H., you solved my problem! The 'plotrix' package didn't work for me, but I used 'collapse' and it worked fine:
combine_custom1 <- function(a, b) {
if (collapse::ldepth(a) > collapse::ldepth(b)) {
ls <- c(a, list(b))
} else {
ls <- c(list(a), list(b))
}
return(ls)
}