Creating KML file from R?

Question

I have a dataframe in R with three columns; latitude, longitude and cluster. I would like to export this data into a KML file for visualizing these clusters in Google Earth. Here's a snippet of my data:

    structure(list(longitude = c(-73.1576, 109.5244, 109.5235, 109.5237, 
109.5244, 19.9022, -73.1575, -73.1574, 109.5235, 19.9069, 109.5242, 
19.9023, -73.1579, -73.1578, -73.1594, 19.9051, 109.5242, 19.9009, 
-73.1575, -73.1578, -73.1592, -73.1577, 109.5235, 109.5246, 109.5243, 
-73.1578, 19.9052, -73.1569, 19.9022, 19.9061, -73.1595, 109.5238, 
-73.157, -73.1576, 19.9044, 109.5257, -73.1571, 109.5249, -73.1577, 
109.5245, -73.1573, 19.9021, 19.9022, 19.9029, -73.1575, -73.1573, 
-73.157, 109.5243, 19.907, 19.9031, -73.1574, -73.1569, 109.5245, 
-73.1593, 19.9057, 109.5242, 109.5236, 19.9054, 109.5235, 109.5241, 
19.9058, 19.9055, 109.5242, -73.1593, 109.5239, 19.9009, 109.5241, 
-73.1572, -73.1572, -73.1571, 109.5235, -73.1576, 109.5243, 19.9031, 
19.9021, 19.9021, 19.9008, 109.5245, 19.9022, 109.5239, 19.903, 
-73.1583, 109.5242, -73.1575, 19.9045, 19.9008, -73.1574, 19.9008, 
-73.1573, 109.5242, -73.1576, 19.9053, 109.5237, -73.1572, -73.157, 
19.9061, -73.158, 19.9007, 19.9028, 19.9058, 109.5245, 19.905, 
19.9031, -73.1574, -73.1584, 109.5236, 19.9052, 19.9055, 19.9022, 
19.9051, 109.5241, 109.5244, -73.1571, -73.157, -73.1583, 109.524, 
19.9061, -73.1572, 109.5241, 19.9058, 109.5243, 19.906, 109.5244, 
-73.1577, -73.1577, -73.1576, -73.1579, 109.5238, -73.1569, 109.5243, 
19.9045, 109.5238, 19.9054, -73.1578, 109.5237, 109.5236, -73.1571, 
19.9057, 109.5243, -73.1577, -73.1575, 109.5236, 19.9023, 109.5235, 
109.5237, 109.5236, 109.5234, -73.1573, 19.9044, -73.1578, 19.9022
), latitude = c(-37.0423, 21.4841, 21.482, 21.4827, 21.4839, 
39.6318, -37.0423, -37.0425, 21.4821, 39.6329, 21.4838, 39.6331, 
-37.0423, -37.0423, -37.0423, 39.6329, 21.4834, 39.6315, -37.0426, 
-37.0422, -37.0423, -37.0424, 21.4824, 21.4842, 21.4836, -37.0424, 
39.6329, -37.0424, 39.6331, 39.6277, -37.0428, 21.4827, -37.0424, 
-37.0425, 39.6326, 21.4826, -37.0425, 21.4836, -37.0425, 21.4835, 
-37.0423, 39.6304, 39.6332, 39.6325, -37.0422, -37.0426, -37.0423, 
21.4841, 39.6329, 39.6334, -37.0424, -37.0423, 21.4842, -37.0422, 
39.6325, 21.4837, 21.4825, 39.6329, 21.4822, 21.4837, 39.6326, 
39.6329, 21.4835, -37.0423, 21.4829, 39.6316, 21.4835, -37.0423, 
-37.0426, -37.0424, 21.4823, -37.0426, 21.4839, 39.6335, 39.6315, 
39.6316, 39.6317, 21.484, 39.6316, 21.4828, 39.6325, -37.0426, 
21.4833, -37.0425, 39.6325, 39.6315, -37.0426, 39.6316, -37.0425, 
21.4836, -37.0424, 39.6331, 21.4824, -37.0425, -37.0425, 39.6276, 
-37.0424, 39.6316, 39.6322, 39.6324, 21.4841, 39.6317, 39.6325, 
-37.0423, -37.0429, 21.4824, 39.633, 39.633, 39.6317, 39.633, 
21.4834, 21.484, -37.0423, -37.0426, -37.0425, 21.4833, 39.6324, 
-37.0424, 21.4836, 39.6325, 21.4835, 39.6277, 21.4842, -37.0422, 
-37.0423, -37.0422, -37.0424, 21.4828, -37.0425, 21.4837, 39.6326, 
21.4829, 39.633, -37.0421, 21.4828, 21.4823, -37.0426, 39.6326, 
21.4838, -37.0426, -37.0424, 21.4821, 39.6316, 21.4825, 21.4823, 
21.4822, 21.4823, -37.0424, 39.6325, -37.0425, 39.6315), cluster = c(2579, 
3050, 3050, 3050, 3050, 5222, 2579, 2579, 3050, 5222, 3050, 5222, 
2579, 2579, 2579, 5222, 3050, 5222, 2579, 2579, 2579, 2579, 3050, 
3050, 3050, 2579, 5222, 2579, 5222, 5222, 2579, 3050, 2579, 2579, 
5222, 3050, 2579, 3050, 2579, 3050, 2579, 5222, 5222, 5222, 2579, 
2579, 2579, 3050, 5222, 5222, 2579, 2579, 3050, 2579, 5222, 3050, 
3050, 5222, 3050, 3050, 5222, 5222, 3050, 2579, 3050, 5222, 3050, 
2579, 2579, 2579, 3050, 2579, 3050, 5222, 5222, 5222, 5222, 3050, 
5222, 3050, 5222, 2579, 3050, 2579, 5222, 5222, 2579, 5222, 2579, 
3050, 2579, 5222, 3050, 2579, 2579, 5222, 2579, 5222, 5222, 5222, 
3050, 5222, 5222, 2579, 2579, 3050, 5222, 5222, 5222, 5222, 3050, 
3050, 2579, 2579, 2579, 3050, 5222, 2579, 3050, 5222, 3050, 5222, 
3050, 2579, 2579, 2579, 2579, 3050, 2579, 3050, 5222, 3050, 5222, 
2579, 3050, 3050, 2579, 5222, 3050, 2579, 2579, 3050, 5222, 3050, 
3050, 3050, 3050, 2579, 5222, 2579, 5222)), row.names = c(NA, 
-151L), class = c("data.table", "data.frame"), .internal.selfref = <pointer: 0x10200e8e0>)

I created a list of spatialPolygondataframe's containing concave hulls as follows:

points_df <- split(points_df, points_df$cluster)
clusters <- names(points_df)
points_df <- lapply(points_df, function(x) { x$cluster <- NULL; x })
points_df <- lapply(points_df, st_as_sf, coords = c("longitude", "latitude"), crs = "+proj=longlat +datum=WGS84")
points_df <- lapply(points_df, concaveman)
points_df <- lapply(points_df, as_Spatial)

So I ended up with a list of SpatialPolygonsDataFrame's. Now I want to use the kmlPolygon function in maptools to generate a KML file for visualization in Google Earth as follows:

mycolors <- colorRampPalette(brewer.pal(8, "Set2"))(length(clusters))
kmlPolygon(obj=points_df, kmlfile="myfile.kml", col=mycolors)

This throws the following error:

Error in kmlPolygon(obj = points_df, kmlfile = "myfile.kml", col = mycolors) : 
  obj must be of class 'Polygons' or 'SpatialPolygonsDataFrame' [package 'sp']

Now if I use one element from the list (correctly):

kmlPolygon(obj=points_df[[1]]], kmlfile="myfile.kml", col=mycolors)

everything works and I am able to generate the kml file and visualize it in Google Earth. So my question is, how do I use the whole list of SpatialPolygonsDataFrame's - is there some sort of operation I need to do to combine them or ....?

Answer 1

I finally figured out the solution to this problem:

Using the same input from the question, and using the sp package and a workflow of Polygon >> Spatial Polygons >> Spatial Polygons dataframe (thanks R help; ?SpatialPolygons).

Prior to that I needed to find the concave hull for each cluster (using the concaveman package) and that required a matrix of coordinates which I created using the coordinates function in the sp package. (thanks to the question from @Spacedman - i realized this is needed)

points_df <- split(points_df, points_df$cluster) # Split points by cluster
clusters <- names(points_df)
points_df <- lapply(points_df, function(x) { x$cluster <- NULL; x })
points_df <- lapply(points_df, coordinates) # create coordinates
points_df <- lapply(points_df, concaveman) # calculate concave hull
points_df <- lapply(points_df, Polygon) # create Polygon
points_df <- lapply(seq_along(points_df), function(i) Polygons(list(points_df[[i]]), ID = clusters[i]  )) # Assign ID's
points_df <- SpatialPolygons(points_df, proj4string = CRS("+proj=longlat +datum=WGS84")) # Create spatial polygon
points_df <- SpatialPolygonsDataFrame(points_df, data.frame(i=1:5586,id = clusters, row.names = clusters)) # Create spatial polygons dataframe 
class(points_df)

mycolors <- colorRampPalette(col2rgb(brewer.pal(8, "Set2")))(length(clusters)) # set colors - but it doesn't work for some reason, am still investigating
kmlPolygons(obj=points_df, kmlfile="myfile.kml", col=mycolors) # write kml file

note that this uses the kmlPolygons function form maptools as opposed to the kmlPolygon - the distinction should be obvious from the name, though I think they could be merged into one function?

https://rdrr.io/cran/maptools/src/R/kmlPolygon.R

https://rdrr.io/cran/maptools/src/R/kmlPolygons.R