Iterate along the two vectors of different length using map()

Question

I have two vectors of unequal length. One vector is a list of dataframes and the other vector is made up of unique values. How can I use map() to iterate the vectors over a custom function? Dummy data is as thus:

x2 <- letters
y2 <- list(df1, df2, df3, .....)

# where:
df1 <- data.frame(
                  alphabet = c(letters[1:2]),
                  latitude = c(7.302888, 7.302717),
                  longitude = c(5.143116, 5.143184)
                  )

df2 <- data.frame(
                  alphabet = c(letters[3:4]),
                  latitude = c(4.948605, 4.948610),
                  longitude = c(8.352012, 8.352157)
)

df3 <- data.frame(
                  alphabet = c(letters[5:6]),
                  latitude = c(7.681907, 7.681867),
                  longitude = c(6.414152, 6.414239 )
)

dist_func <- function(i, j) {
  
  mat = j[j$alphabet==i, c("latitude","longitude")]
  
  mat

}

map(x2, ~ map(y2, function(j) dist_func(.i, j)))

throws up error:

Expected output should be list of dataframes, where each dataframe is as thus:

  alphabet latitude longitude
1   a      7.302888  5.143116

any help will be appreciated.

Answer 1

Here's a tidyverse take with the new dplyr::group_split:

library(tidyverse)

my_output <- list_flatten(map(y2, \(y) group_split(y, alphabet)))

Output:

> my_output
[[1]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 a            7.30      5.14

[[2]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 b            7.30      5.14

[[3]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 c            4.95      8.35

[[4]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 d            4.95      8.35

[[5]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 e            7.68      6.41

[[6]]
# A tibble: 1 × 3
  alphabet latitude longitude
  <chr>       <dbl>     <dbl>
1 f            7.68      6.41

That's it!

PS. About map and your attempted solution:

If x and y had the same length, you could use map2_. Since that's not the case, you can use one map_ to iterate over x and another map_ to iterate over y.

Beware: the result will be a list of the same length as x with elements of the same length as y, even if they are just empty dataframes (and it could be messy). This way, you need to control how the output should be aggregated (rows/columns with map_dfc, map_dfr or with bind_cols, bind_rows after) and what should be kept (inside your custom function or after the iteration).

The code below is close to what you tried to do:

library(tidyverse) # at least `dplyr` and `purrr`

my_function <- function(i, j) j[j$alphabet == i, ] # c("latitude", "longitude") removed
my_output <- map(x2,\(x) map(y2,\(y) my_function(x, y)))

The unpleasant output:

> my_output[6:7] # `6` is the last non-empty and `7` the first empty
[[1]]
[[1]][[1]] # No "f" in `df1`
[1] alphabet  latitude  longitude
<0 linhas> (ou row.names de comprimento 0)

[[1]][[2]] # No "f" in `df2`
[1] alphabet  latitude  longitude
<0 linhas> (ou row.names de comprimento 0)

[[1]][[3]] # There's "f" in `df3`
  alphabet latitude longitude
2        f 7.681867  6.414239


[[2]]      # No "g" in `df1` to `df3`
[[2]][[1]]
[1] alphabet  latitude  longitude
<0 linhas> (ou row.names de comprimento 0)

[[2]][[2]]
[1] alphabet  latitude  longitude
<0 linhas> (ou row.names de comprimento 0)

[[2]][[3]]
[1] alphabet  latitude  longitude
<0 linhas> (ou row.names de comprimento 0)

my_output is a list with length 26 (a to z) and each element length 3 (df1 to df3).

Answer 2

The idiomatic way to do this in R is to bind the rows into a data frame and then split():

do.call(rbind, y2)  |>
    split(~alphabet)

# $a
#   alphabet latitude longitude
# 1        a 7.302888  5.143116

# $b
#   alphabet latitude longitude
# 2        b 7.302717  5.143184

# $c
#   alphabet latitude longitude
# 3        c 4.948605  8.352012

# $d
#   alphabet latitude longitude
# 4        d  4.94861  8.352157

# $e
#   alphabet latitude longitude
# 5        e 7.681907  6.414152

# $f
#   alphabet latitude longitude
# 6        f 7.681867  6.414239