How to get a shape of a searched map in tmap

Question

I am using

library(tmap)
qtm("London")

to get a map of London but can't find a way to get it in a shape?

tm_shape

Otherwise, I can only have it in view mode, not in plot mode.

Answer 1

Yes, that is correct. You have to do it another way to get a map layer in tmap_mode("plot").

    

In tmap_mode("view") you can just do it like that:

tm_shape(lnd) + tm_polygons(alpha = 0.7, col = "Pop_2001", legend.show = FALSE) + 
  tm_basemap(server = c('OpenStreetMap'))

With tm_basemap(server = c('OpenStreetMap')) not being necessary, but I've put it anyway, because here you can change the basemap, for example to stamen.     

For tmap_mode("plot") you have to download the tiles manually. There are several ways to do that.

1. The tmaptools-package offers a function to do that read_osm(). To get the Tiles you do lonmap <- bb("London") %>% read_osm(type = "osm"), which returns a raster object.

    

2. Another alternative is openmap() from the OpenStreetMap-package, which would look like that:

london <- bb("London")
lonmap <- openmap(upperLeft = london[3:4], 
                  lowerRight = london[1:2], 
                  type = "osm",
                  mergeTiles = TRUE)

    

3. And of course there is get_map() from the ggmap-package for which you now need an API-key. This can't be put directly into tm_shape() however.

    

But to be honest read_osm() and openmap() both give me the error

java.lang.NullPointerException

and I haven't been able to find a solution to it.

    

4. So I use even another one: The getTiles() function from the cartography-package. getTiles only takes sf or sp objects and returns a RasterBrick-object

library(tmap)
library(cartography)

lonmap <- getTiles(x = lnd, type = "osm")

tmap_mode("plot")

tm_shape(lonmap) + tm_rgb() + tm_shape(lnd) + 
  tm_polygons(alpha = 0.5,col = "Pop_2001", legend.show = FALSE) 

Which gives you this:

A workaround for being able to use a bounding box would be:

lonmap <-  bb("London") %>% 
            matrix(ncol = 2, byrow = TRUE) %>%
                SpatialPoints(proj4string = CRS("+init=epsg:4326")) %>%
                    getTiles(type = "osm")

The lnd, is a SpatialPolygonsDataFrame. Never mind the content. You can get it from here: https://github.com/Robinlovelace/Creating-maps-in-R/blob/master/data/london_sport.shp