In Chrisjb/basemapR: Base Maps For ggplot2

knitr::opts_chunk$set(echo = TRUE,
                      fig.path = "man/figures/README-")

Installing basemapR

Install from github using devtools:

library(devtools)
install_github('Chrisjb/basemapR')

Credits

A lot of credit has to go to Hiroakki Yutani for his blog post from which I borrowed a lot of code for the base_map function. I hope they don't mind!

Adding a basemap to ggplot2

the base_map function can be added to a ggplot2 call as follows:

library(ggplot2)
library(sf)
library(basemapR)

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2) +
  geom_sf(data = localauth_data, fill = NA)

bbox

A bounding box created using st_bbox or the basemapR function expand_bbox (see below). The bounding box defines the extents over which we want the base map to be returned.

This will normally be the largest layer on our map. Sometimes we will have two distinct layers neither of which covers the full extent of the map on their own. For example:

library(dplyr)
library(basemapR)
library(sf)


camden <- localauth_data %>% 
  filter(Name  == 'Camden')

wandsworth <- localauth_data %>% 
  filter(Name  == 'Wandsworth')


ggplot() +
  geom_sf(data = camden) +
  geom_sf(data = wandsworth)

We need to create a bbox that combines both layers for the base map to cover the full extents of the canvas:

# create bbox polygons for each and union and then convert back to bbox object
my_bbox <- st_bbox(camden) %>%
  st_as_sfc() %>%
  st_union(st_as_sfc(st_bbox(wandsworth))) %>%
  st_bbox()

ggplot() +
  base_map(bbox = my_bbox, basemap = 'hydda', increase_zoom = 2) +
  geom_sf(data = camden, fill = NA) +
  geom_sf(data = wandsworth, fill = NA) +
  ggthemes::theme_map()

increase_zoom

The zoom parameter is calculated automatically. It can be increased by setting increase_zoom to an integer value. In the first example we set increase_zoom = 2. We can play around with this value as per our desired aesthetic.

basemap

Various options for base maps. The attribution for the base layer should be included on the maps and is returned as a message from the function.

dark

attribution: © OpenStreetMap contributors © CARTO

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'dark') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

hydda

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'hydda') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

positron attribution: Tiles courtesy of http://openstreetmap.se/ OpenStreetMap Sweden; Map data © OpenStreetMap contributors

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'positron') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

voyager attribution: © OpenStreetMap contributors © CARTO

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'voyager') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

wikimedia please see attribution details: https://wikimediafoundation.org/wiki/Maps_Terms_of_Use

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'wikimedia') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

mapnik attribution: © OpenStreetMap contributors

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'mapnik') +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

nolabels

If nolabels=TRUE the function will fetch the basemap without street or place labels. This option is only available for some basemaps and will return a message if unavailable.

ggplot() +
  base_map(st_bbox(localauth_data), increase_zoom = 2, basemap = 'hydda', nolabels = TRUE) +
  geom_sf(data = localauth_data, fill = NA) +
  ggthemes::theme_map()

expand_bbox

This function takes in a bbox object and expands it by a set number of meters in the X and Y directions. If we want an asymmetric expansion (100m to the east but 1km to the west) we can also specify the parameters X2 and Y2.

# define a single point
point <- data.frame(x = -0.086543 ,
                    y= 51.504567) %>%
  st_as_sf(coords= c('x','y'), crs = 4326)

# standard bbox for our point
my_bbox_1 <- st_bbox(point)

# expand the bbox by 1000m
my_bbox_2 <- expand_bbox(my_bbox_1, X = 1000, Y = 1000)

We will have to also set the x and y limits ourselves in coord_sf to clip the canvas to our basemap:

ggplot() + 
  base_map(my_bbox_2, increase_zoom = 1, basemap = 'positron')+
  geom_sf(data = point) +
  coord_sf(xlim = c(my_bbox_2['xmin'], my_bbox_2['xmax']),
           ylim = c(my_bbox_2['ymin'],my_bbox_2['ymax']),crs = 4326)