How to separate a dataframe by group and then write separate KML files for each group in R?

Question

I have a data frame that has latitudes and longitudes along with attributes. I want to split this data frame up by grouping one of the columns and write these groups into separate KML files in R. For example I want separate KML files based on color (one for Blue, one for Green, and one for Red). For reference, my actual data set I have 50+ colors.

    Name Latitude   Longitude Color
   Dawn 48.93553 -111.677777  Blue
   Lore 45.13333 -114.888888 Green
   Nice 40.73499 -114.999900  Blue
    Cat 41.89999 -114.999200   Red
   Lake 48.99678 -116.787876   Red
 Forest 42.88889 -112.999900   Red

You can use `split` to split your data frame into a list by color (ie, `split(df, df$Color)`, then loop through the list and write the KML file using the `writeOGR` function in the `rgdal` - see this question for more help: [How to create a KML file using R](https://stackoverflow.com/questions/7813141/how-to-create-a-kml-file-using-r) — jpsmith, Mar 23 '23 at 20:35
@jpsmith Thanks for your answer, split and the for loop gets me part of the way there. I am struggling to understand how to use the forloop to name each of the kml files based in the group. — ksweet, Mar 23 '23 at 21:08

margusl · Accepted Answer · 2023-03-23T21:26:44.020

You could first convert your data.frame to sf object, it can be split() to named list as a regular data.frame, and then save each part as KML through sf::st_write(). purrr::iwalk() cycles though the list and can also access list names, so the split factor ends up in file names too.

library(sf)
#> Linking to GEOS 3.9.3, GDAL 3.5.2, PROJ 8.2.1; sf_use_s2() is TRUE

df_ <- read.table(header = T, text ="
   Name Latitude   Longitude Color
   Dawn 48.93553 -111.677777  Blue
   Lore 45.13333 -114.888888 Green
   Nice 40.73499 -114.999900  Blue
    Cat 41.89999 -114.999200   Red
   Lake 48.99678 -116.787876   Red
 Forest 42.88889 -112.999900   Red")

st_as_sf(df_, coords = c("Longitude", "Latitude"), crs = "WGS84") |>
  split(~ Color) |>
  purrr::iwalk(\(features, color) st_write(features, paste0(color,".kml")))
#> Writing layer `Blue' to data source `Blue.kml' using driver `KML'
#> Writing 2 features with 2 fields and geometry type Point.
#> Writing layer `Green' to data source `Green.kml' using driver `KML'
#> Writing 1 features with 2 fields and geometry type Point.
#> Writing layer `Red' to data source `Red.kml' using driver `KML'
#> Writing 3 features with 2 fields and geometry type Point.

Resulting files:

fs::dir_info(glob = "*.kml")[1:5]
#> # A tibble: 3 × 5
#>   path       type         size permissions modification_time  
#>   <fs::path> <fct> <fs::bytes> <fs::perms> <dttm>             
#> 1 Blue.kml   file          559 rw-         2023-03-23 23:09:14
#> 2 Green.kml  file          415 rw-         2023-03-23 23:09:14
#> 3 Red.kml    file          702 rw-         2023-03-23 23:09:14

^{Created on 2023-03-23 with reprex v2.0.2}

This is perfect, thank you! Do you know if there is a way to write the KML files such that there is not an additional name and description column at the beginning? Or if you can specify which column is written first? Thank you, thank you — ksweet, Mar 23 '23 at 21:42

How to separate a dataframe by group and then write separate KML files for each group in R?

1 Answers1