An Interactive Mapping R Package

Background Information

Released in 2015, the Leaflet package “revolutionized interactive web map creation from within R” (Lovelace, 2020). Leaflet sought to make creating pretty, interactive mobile maps simple and easy to understand.

Wildly popular, Leaflet is now used by a multitude of famous websites and GIS specialists, like the New York Times and Mapbox, for its open-source Javascript library for interactive maps (Agafonkin, 2016).

Version History

Version Number	Release Date
1.0.0	June 19, 2015
1.0.1	Feb. 25, 2016
1.1.0	Feb. 17, 2017
2.0.0	Apr. 20, 2018
2.0.1	June 4, 2018
2.0.2	Aug. 27, 2018
2.0.3	Nov. 14, 2019

(Cheng, 2019)

Usage of and Dependency to Other Packages

Since it was developed on top of the htmlwidgets package, Leaflet can render maps in R markdowns, Rstudio IDE, and shiny apps (Xie, 2015).

Imports

Leaflet imports the following packages:

base64enc
crosstalk
htmltools
htmlwidgets
leaflet.providers (>= 1.8.0)
magrittr
markdown
methods
png
raster
RColorBrewer
scales (>= 1.0.0)
sp
stats
viridis (>= 0.5.1)

(Cheng, 2019)

Suggests

Leaflet suggests the following packages:

knitr
maps
purrr
R6
rgdal
rgeos
RJSONIO
sf
shiny
testthat

(Cheng, 2019)

Usage

The main capabilities of this package all serve to add components to the map to communicate information or make the map easier to understand. The principal functionalities are:

to initialize the window
to create map
to add shapes
to add colors
to add markers
to add popups & labels
to add legend
to add layers

Example

To illustrate the functions of each of the above capabilities, we created an example by mapping some restaurant locations over census tract populations in Charlottesville, VA. The sections below will go through the main functionalities individually.

Initialize

Initializing creates the window for other functions to add to.

It uses these functions:

leaflet()
leafletOptions(): sets minZoom and maxZoom
setView(): sets the center of the map view and the zoom level
fitBounds(): fits the view into the rectangle
clearBounds(): clears the bound, so that the view will be automatically determined by the range of latitude/longitude data in the map layers if provided

Code:

# Initializing the window
leaflet()

Create

Creating the map requires “tiles” that tell Leaflet what type of map to output.

It uses these functions:

addTiles(): adding OpenStreetMap base map
addProviderTiles(): adding third-party provided base maps (via plugin)
addWMSTiles(): adding Web Map Services tiles (via the internet)

Code:

# Adding tiles will make an actual map appear in the window; default Leaflet map 
# (Set on Charlottesville)
leaflet() %>% addTiles() %>% setView(lng = -78.4766781, lat = 38.0293059, zoom = 12)

# Satellite image map of Charlottesville
leaflet() %>% setView(lng = -78.4766781, lat = 38.0293059, zoom = 12) %>%
  addProviderTiles(providers$Esri.WorldImagery)

# National Geographic style map of Charlottesville
leaflet() %>% setView(lng = -78.4766781, lat = 38.0293059, zoom = 12) %>%
  addProviderTiles(providers$Esri.NatGeoWorldMap)

Shapes

Adding shapes allows users to put in country, state, etc. shapes in the map to connect to data.

It uses these functions:

addPolygons(): adds map shapes (countries, states, counties, etc.)
addCircles(): adds circle with given radius around a point
addRectangles(): adds a rectangle with 4 given points

Code:

# Importing census data from Charlottesville
tracts <- readOGR('US_Census_Tract_Area_2010.shp', GDAL1_integer64_policy = TRUE)

# Initializing the window with our desired area data points and 
# adding polygons to make visible the census tracts we instantiated in leaflet() 
leaflet(tracts) %>% addTiles() %>% addPolygons(weight = 1, 
                                               smoothFactor = 0.5,
                                               opacity = 1.0, 
                                               fillOpacity = 0.2,
                                               highlightOptions = 
                                                 highlightOptions(color = "white", 
                                                                weight = 2,
                                                                bringToFront = TRUE))

Colors

Adding colors to a map can help convey more information within the visual.

It uses these functions:

colorNumeric(): continuous input, continuous colors
colorBin(), colorQuantile(): continuous input, discrete colors
colorFactor(): categorical input

Code:

# Instantiating a color palette
pal <- colorNumeric(palette = "Blues",
                    domain = as.numeric(tracts$Population))

# Using that palette for our map
leaflet(tracts) %>% addTiles() %>% 
  addPolygons(color = ~pal(tracts$Population),
              weight = 1, 
              smoothFactor = 0.5,
              opacity = 1.0,
              fillOpacity = 0.5,
              highlightOptions = highlightOptions(color = "white", 
                                                  weight = 2, 
                                                  bringToFront = TRUE))

Markers

Markers are used to highlight points on the map to add more information to the display.

It uses these functions:

addMarker(), addAwesomeMarker(): adding icon marker (default is dropped pin)
makeIcon(), makeAwesomeIcon(): customizing icon markers
iconList(), awesomeIconList(): set of icons with varying parameters
clusterMarkers(), clusterMarkerOptions(): clustering large quantities of markers (when you zoom out they combine into one icon)
addCircleMarkers(): adds circle markers with constant radius

Code:

# Importing data about cville restaurants
food <- read.csv('cville_food.csv', header = TRUE)

# Adding default markers
leaflet(data = food) %>% addTiles() %>% addMarkers(~long, ~lat)

# Marker icons can be personalized as well
restIcons <- icons(
  iconUrl = "https://image.flaticon.com/icons/svg/433/433087.svg",
  iconWidth = 38, iconHeight = 95
)

# Adding the icons to the map
leaflet(data = food) %>% addTiles() %>% addMarkers(~long, ~lat, icon = restIcons)

Labels

Popups and labels are good ways to explain what a marker is or represents, without overcrowding the visual. Labels appear when a marker is moused over. Popups appear when a marker is clicked on.

They use these functions:

addPopups(): add popups to map
Popups also work as an argument of addMarkers() when you want information to appear when you click a marker
Labels work as an argument of addMarkers() when you want information to appear when you hover over a marker
labelOptions()
Labels without markers: addLabelOnlyMarkers()

Code:

leaflet() %>% 
  addTiles() %>% 
  addTiles(group = "Street View") %>%
# Adding labels and popups to the map via parameters within addMarkers()
  addMarkers(data = food %>% filter(food$cost == '$'), group = '$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$'), group = '$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$$'), group = '$$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address))

Legends

Legends explain what some of the markers or colors represent on the map. This enables more information to be on visual without overcrowding the page.

Its main function is:

addLegend()

Code:

leaflet(tracts) %>% addTiles() %>% 
  addPolygons(color = ~pal(as.numeric(tracts$Population)),
              weight = 1, 
              smoothFactor = 0.5,
              opacity = 1.0,
              fillOpacity = 0.5,
              highlightOptions = highlightOptions(color = "white", 
                                                weight = 2, 
                                                bringToFront = TRUE)) %>% 
# Adding a legend to the top left of the window to illustrate which 
# shading levels correspond to populations
  addLegend("topleft",
            pal = pal, 
            values = as.numeric(tracts$Population), 
            title = "Est. Population (2010)", 
            labFormat = labelFormat(suffix = " citizens"), 
            opacity = 1)

Layers

Layers are another way to fit more information onto the map without overcrowding. In our example, it looks like adding a filter on restaurant markers– allowing the users to decide which type of restaurant (layer) they want to see.

Its main function is:

addLayersControl(): adds box that allows you to control which layers you see

Code:

leaflet() %>% 
  addTiles() %>% 
  addTiles(group = "Street View") %>%
  addProviderTiles(providers$Esri.WorldImagery, group = "Satellite") %>%
  addMarkers(data = food %>% filter(food$cost == '$'), group = '$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$'), group = '$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$$'), group = '$$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(name), popup = ~htmlEscape(address)) %>%
  # Adding Layers to the map of different map views or different price ranges of restaurants
  addLayersControl(baseGroups = c("Street View", "Satellite"), 
                   overlayGroups = c('$','$$', '$$$'),
                   options = layersControlOptions(collapsed = FALSE))

Final product

Wrapping all of the functionalities into one, the final product will look like:

leaflet() %>% 
  addTiles() %>% 
  addTiles(group = "Street View") %>%
  addProviderTiles(providers$Esri.WorldImagery, group = "Satellite") %>%
  addMarkers(data = food %>% filter(food$cost == '$'), group = '$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(food$name), popup = ~htmlEscape(food$address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$'), group = '$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(food$name), popup = ~htmlEscape(food$address)) %>%
  addMarkers(data = food %>% filter(food$cost == '$$$'), group = '$$$', ~long, ~lat, icon = restIcons, 
             label = ~htmlEscape(food$name), popup = ~htmlEscape(food$address)) %>%
  addLayersControl(baseGroups = c("Street View", "Satellite"),
                   overlayGroups = c('$','$$', '$$$'),
                   options = layersControlOptions(collapsed = FALSE)) %>%
  addPolygons(data = tracts,
              color = ~pal(tracts$Population),
              weight = 1, 
              smoothFactor = 0.5,
              opacity = 1.0,
              fillOpacity = 0.5,
              highlightOptions = highlightOptions(color = "white", weight = 2, 
                                                  bringToFront = TRUE)) %>% 
  addLegend("topleft",
            pal = pal, 
            values = as.numeric(tracts$Population), 
            title = "Est. Population (2010)", 
            labFormat = labelFormat(suffix = " citizens"), 
            opacity = 1)

Similar Packages

There are multiple packages in R that also create maps. While Leaflet creates interactive maps designed for mobile use, other packages address various other types of mapping. The function plot() is in basic R and allows for low level control for static maps. The grid package is another basic R package to make maps, but has limited capabilities. Focusing on interactive mapping packages allows for closer analysis between relevant ones.

Other Interactive Map-Making Packages

There are three other popular mapping packages: ggplot2, tmap, and mapview.

Ggplot2

The ggplot2 package is widely used and accepted, primarily because its “grammar of graphics” syntax is easy to use and understand. The ggplot2 package has many of the same capabilities to manipulate aesthetics as Leaflet that allow the user to customize their graph. The ggplot2 package can create static maps in addition to using wrapper packages such as ggiraph(), ggplotly() and gganimate() to create interactive and animated maps.

(Lovelace, 2020)

Tmap

The tmap package can create static, animated and interactive maps with minimal code. tmap uses the “grammar of graphics” syntax, similar to ggplot2.

(Lovelace, 2020)

Mapview

The mapview package can create static, animated and interactive maps. The mapview package is considered the “data-driven leaflet” package because of it’s advanced control of layering geographic objects.

(Appelhans, n.d.), (Lovelace, 2020)

Advantages and Disadvantages

The table below shows advantages and disadvantages to each package.

Package	Advantages	Disadvantages
Leaflet	Creates exceptional dynamic maps Has many aesthetic features that make maps popular in R community Works with `shiny()` to post interactive maps to websites Accepts many types of map objects Can create mobile maps Highly flexible	Has own syntax Static maps are not as good as ggplot2’s static maps
ggplot2	Uses “grammar of graphics” syntax Large user-community Many add-on packages to add additional features Highly flexible	Requires a `data.frame` object so data needs to converted to a dataframe using `tidy()` Interactive maps are not as good as leaflet’s Have to implement wrapper packages to make graphs interactive or animated (for example: `ggiraph()` or `plotly()`)
Tmap	Uses “grammar of graphics” syntax Can set options globally for all maps in code using `tmap_options()` Can export tmap maps to leaflet to add leaflet functionality and use shiny Interactive and static maps can be made with same code Accepts many types of map objects	Not as widely used Syntax is not widely known
Mapview	Powerful Accepts many types of map objects Many interactive features Easy way to create an interactive map with simple code	Not as widely used Syntax is not widely known

(Lovelace, 2020)

Reflection

Overall, Leaflet is an incredibly powerful tool for geographic data visualization.This package offers many impressive tools that are easy to implement in a professional and aesthetically pleasing way.

Pro/Cons

Pros	Cons
Used by professionals; it is scalable and reliable	Has its own syntax; slight learning curve
Can use a multitude of map views	Redundant functionalities
Many customizable features

Suggestions for Improvement

Simplifying functionalities– too many ways to do the same thing

References

Agafonkin, V. (2016). Leaflet for R. Retrieved from http://rstudio.github.io/leaflet/

Appelhans, T., Detsch, F., Reudenbach, C., & Woellauer, S. (n.d.). Interactive Viewing of Spatial Data in R. Retrieved March 3, 2020, from https://r-spatial.github.io/mapview/

Basille, M., & Moreno, M. (2018). Drawing beautiful maps programmatically with R, sf and ggplot2—Part 1: Basics. R-Spatial. https://www.r-spatial.org/r/2018/10/25/ggplot2-sf.html

Cheng, J. (2019). leaflet v2.0.3. Retrieved from https://www.rdocumentation.org/packages/leaflet/versions/2.0.3

Lovelace, R., Nowosad, J., & Muenchow, J. (2020, February 16). Geocomputation with R. Retrieved from https://geocompr.robinlovelace.net/adv-map.html

Xie, Y. (2015). Leaflet: Interactive web maps with R. Retrieved from https://blog.rstudio.com/2015/06/24/leaflet-interactive-web-maps-with-r/

Data Retrieved From:

https://opendata.charlottesville.org/datasets/us-census-tract-area-2010

and Google

Leaflet Package Overview

Ally Renehan, Lauren Perry, and Zach Martin

3/4/2020

An Interactive Mapping R Package

Background Information

Version History

Usage of and Dependency to Other Packages

Imports

Suggests

Usage

Example

Initialize

Create

Shapes

Colors

Markers

Labels

Legends

Layers

Final product

Similar Packages

Other Interactive Map-Making Packages

Ggplot2

Tmap

Mapview

Advantages and Disadvantages

Reflection

Pro/Cons

Suggestions for Improvement

References

Data Retrieved From: