Exploring Alternatives to Mapview for Customizable Mapping in R

As a data analyst often relying on spatial data visualization, I frequently face challenges with the mapping tools available in R. Recently, I encountered a particular issue with the mapview package that many others might relate to. The problem arises when using mapview with the burst option, which, although great for interactive legends and auto-zoom, lacks control over the scale of the legend.

In the example I was working with, I had a dataset of sampled stations which included geographical coordinates and station names. The dataset was visualized to observe the locations and names on an OpenStreetMap background. The dataset looked something like this in the tibble format but included rows for each station:

“`

A tibble: x × 4

station_id station_names longitude latitude

1 101 Station A -120.5 40.7

2 102 Station B -120.6 40.8

…

“`

Using the following code, I attempted to create a map:

library(sf)
library(mapview)

sampled_stations %>%
    st_as_sf(coords = c("longitude", "latitude"), crs = 4326) %>%
    mapview(
        zcol = "station_names",
        col.regions = randomColors,  # Assuming randomColors is predefined
        map.types = "OpenStreetMap",
        burst = TRUE
    )

However, this left the legend cramped and partly unreadable, and adjusting the legend scale was not straightforward in mapview.

Finding a Solution with Leaflet

To achieve better control over the aesthetics and functionality of my map, particularly the scale and layout of the legend, I decided to switch to using the leaflet package. Leaflet is highly customizable and allows fine control over almost all aspects of a map, including legends.

Here’s how I approached creating a similar map using leaflet:

library(leaflet)
library(dplyr)

# Prepare the sampled_stations as an sf object
stations_sf <- sampled_stations %>%
    st_as_sf(coords = c("longitude", "latitude"), crs = 4326)

# Create a color palette
colorPalette <- colorFactor(rainbow(length(unique(stations_sf$station_names))), 
                            stations_sf$station_names)

# Build the map
leaflet(stations_sf) %>%
    addTiles() %>%
    addCircleMarkers(~longitude, ~latitude, radius = 8, 
                     color = ~colorPalette(station_names),
                     fillOpacity = 0.8, stroke = FALSE,
                     popup = ~station_names) %>%
    addLegend("bottomright", pal = colorPalette, values = ~station_names,
              title = 'Station Names',
              opacity = 1)

In this version with leaflet, we define a custom color palette and apply it to the map according to station names. Moreover, the legend is explicitly added with addLegend, where placement, opacity, and other visual aspects can be personalized. This flexibility allows the map to be constructed in a way that ensures all elements are visible and appropriately scaled, irrespective of the screen dimensions or the map’s initial zoom level.

By switching to leaflet, I found that I could solve the issue of adjusting the scale of the legend that I experienced with mapview. Plus, leaflet offered a deeper level of control, making it an ideal solution for my needs and perhaps yours too, especially when handling complex interactive maps and requiring specific visual configurations.