rbind the list returned from a function without loop in R

user_7453767 0
Viewed 65

In R, for a function whose input is one variable and output is a list, how can I apply the function to a list of inputs and combine the output lists into one data frame or matrix?

For example:

fun = function(x){
    list = c(x, x+1, x+2)
    return(list)
}

Now I'd like to apply fun() to a list, say

list = c(1,11,21)

The result should be 

1,2,3
11,12,13
21,22,23

I know do.call and loop will do it. But is there a better way of doing it, perhaps without looping? The actual function is too big and too slow, I would rather not looping it.

rbind dataframe r loops
3 Answers

We can use lapply : 

do.call(rbind, lapply(list, fun))

#     [,1] [,2] [,3]
#[1,]    1    2    3
#[2,]   11   12   13
#[3,]   21   22   23

Instead of numeric vector if you return a named list, you can use map.* functions from purrr.

fun = function(x){
  list = list(a = x, b = x+1, c = x+2)
  return(list)
}

purrr::map_df(list, fun)
#OR
#purrr::map_dfr(list, fun)

#     a     b     c
#  <dbl> <dbl> <dbl>
#1     1     2     3
#2    11    12    13
#3    21    22    23
user_39629140

If you want a matrix, sapply is probably sufficient.

lst <- c(1,11,21)  ## don't use `list` since it is a function name!

t(sapply(lst, fun))
#      [,1] [,2] [,3]
# [1,]    1    2    3
# [2,]   11   12   13
# [3,]   21   22   23
user_65740380

Here are some base R solutions:

  • An easy option is using Vectorize + t, i.e.,
t(Vectorize(fun)(list))

such that

> t(Vectorize(fun)(list))
     [,1] [,2] [,3]
[1,]    1    2    3
[2,]   11   12   13
[3,]   21   22   23
  • Another way is using replicate (considering the additive property of your function fun), i.e.,
t(replicate(length(list),fun(0)))+list

such that

> t(replicate(length(list),fun(0)))+list
     [,1] [,2] [,3]
[1,]    1    2    3
[2,]   11   12   13
[3,]   21   22   23
user_121587570
Related