Leaflet on R: How to create layers and colors for each factor level in dataframe

Question

(Btw, is there another package that would do what I want?)

Here's my dataset:

> head(df)
    id groupID       lat       lon
511  1  277706 -10.89834 -37.05893
377  2  277706 -10.88870 -37.07695
98   3  277705 -10.89050 -37.09406
392  4  277697 -10.92131 -37.10525
6    5  277705 -10.89050 -37.09406
364  6  277697 -10.93730 -37.11600

I want to use leaflet to plot every row on a map based on their latitude and longitude. Additionaly, each marker (or popup or CircleMarker) is supposed to have a different color and layer based on the groupID variable.

The problem is the data changes every day and I don't know how many different unique levels I will have for the groupID variable. In this sample dataset there are 12 levels, but it may range from 5 to around 30. The examples in the documentation work with a pre-defined number of levels.

Here's what I've tried:

colorsmap = colors()[1:length(unique(df3$groupID))]
groupColors = colorFactor(palette = "RdYlBu", domain = df3$groupID)

leaflet(data = df3) %>%
  addTiles() %>%
  addCircleMarkers(lng = ~lon, lat = ~lat, color = ~groupColors(groupID),
                   group = ~groupID) %>%
  #addLegend(position = "topright", pal = groupColors, values = ~groupID) %>%
  addLayersControl(~groupID)

It does provide a plot, but when I select only one level, the other ones do not disapper as they should:

image:

The dataset itself:

> dput(df3)
structure(list(id = 1:20, groupID = c(277698L, 277715L, 277704L, 
277706L, 277705L, 277705L, 277719L, 277705L, 277705L, 277709L, 
277706L, 277705L, 277704L, 277706L, 277715L, 277702L, 277719L, 
277706L, 277715L, 277706L), lat = c(-10.8172615660655, -10.8904055150991, 
-10.8887597563482, -10.90203509, -10.9001514, -10.8997748900025, 
-10.8960177351343, -10.8896179908615, -10.8991450456802, -10.9224848475651, 
-10.9000373151094, -10.8905013650562, -10.8889438100208, -10.9001234797436, 
-10.8861897462579, -10.9326053452642, -10.8916601751623, -10.902075281944, 
-10.8822231928033, -10.9079483812524), lon = c(-36.9248145687343, 
-37.0665064455395, -37.0921721937304, -37.05829295, -37.0969278, 
-37.0976847916125, -37.0840372102666, -37.0963566353117, -37.0945971936751, 
-37.0549293249471, -37.066113628594, -37.0940632483155, -37.095505683692, 
-37.0590422449149, -37.0782556623101, -37.0698746017798, -37.0841003949028, 
-37.0593363285999, -37.0724709841895, -37.0817244836096)), .Names = c("id", 
"groupID", "lat", "lon"), row.names = c(20L, 23L, 8L, 36L, 14L, 
13L, 16L, 2L, 11L, 1L, 26L, 6L, 5L, 31L, 22L, 50L, 17L, 34L, 
25L, 42L), class = "data.frame")

Answer 1

library(leaflet)

groups = as.character(unique(df3$groupID))

map = leaflet(df3) %>% addTiles(group = "OpenStreetMap")
for(g in groups){
  d = df3[df3$groupID == g, ]
  map = map %>% addCircleMarkers(data = d, lng = ~lon, lat = ~lat, 
                                 color = ~groupColors(groupID),
                                 group = g)

}
map %>% addLayersControl(overlayGroups = groups)

Answer 2

For completeness, this is now easily doable with mapview using both zcol and burst arguments:

library(mapview)
library(sp)

df3 <- structure(list(id = 1:20, 
                      groupID = c(277698L, 277715L, 277704L, 
                                  277706L, 277705L, 277705L, 
                                  277719L, 277705L, 277705L, 277709L, 
                                  277706L, 277705L, 277704L, 277706L, 
                                  277715L, 277702L, 277719L, 
                                  277706L, 277715L, 277706L), 
                      lat = c(-10.8172615660655, -10.8904055150991, 
                              -10.8887597563482, -10.90203509, -10.9001514, 
                              -10.8997748900025, -10.8960177351343, 
                              -10.8896179908615, -10.8991450456802, 
                              -10.9224848475651, -10.9000373151094, 
                              -10.8905013650562, -10.8889438100208, 
                              -10.9001234797436, -10.8861897462579, 
                              -10.9326053452642, -10.8916601751623, 
                              -10.902075281944, -10.8822231928033, 
                              -10.9079483812524), 
                      lon = c(-36.9248145687343, -37.0665064455395, 
                              -37.0921721937304, -37.05829295, -37.0969278, 
                              -37.0976847916125, -37.0840372102666, 
                              -37.0963566353117, -37.0945971936751, 
                              -37.0549293249471, -37.066113628594, 
                              -37.0940632483155, -37.095505683692, 
                              -37.0590422449149, -37.0782556623101, 
                              -37.0698746017798, -37.0841003949028, 
                              -37.0593363285999, -37.0724709841895, 
                              -37.0817244836096)), 
                 .Names = c("id", "groupID", "lat", "lon"), 
                 row.names = c(20L, 23L, 8L, 36L, 14L, 13L, 16L, 2L, 11L, 
                               1L, 26L, 6L, 5L, 31L, 22L, 50L, 17L, 34L, 25L, 42L), 
                 class = "data.frame")

## convert df3 to spatialPointsDataFrame for use with mapview
coordinates(df3) <- ~ lon + lat
proj4string(df3) <- "+init=epsg:4326"

## now burst column "groupID"
mapview(df3, zcol = "groupID", burst = TRUE)

NOTE: This is currently only available in the development version of mapview which can be installed with

devtools::install_github("environmentalinformatics-marburg/mapview", ref = "develop")

Answer 3

I know that the issue is a bit old and solution have already been given. But I came across the same issue recently and found a quick dirty hack. If you effectively use a layer control only with levels from one factor you will have radio buttons displaying only one level at a time:

addLayersControl(overlayGroups = data$factor)

But if you simply add another layer control from a whole layer you will have all groups activated at a time (comprising every levels from the factor):

addLayersControl(overlayGroups = c(data$factor, "layer name"))