Map insets are small zoom-in or zoom-out maps that focus on particular
territories within a general map. mapinsetr provides a set of
functions that helps to create such insets.
require(devtools)
devtools::install_github("riatelab/mapinsetr")The previous
version of
the package was based on sp, rgeos and maptools.
This version (0.2.0) is based on the sf
package only.
The aim of this example is to create a base map of Algeria’s
municipalities with an inset on the Algiers wilaya.
Here are the few steps of the base map creation process:
- download data on Algerian administrative boundaries,
- create a mask that will help to extract the municipalities,
- create an inset of Alger municipalities within the previously created mask,
- merge the original basemap and inset
The raster package offers an access to the GADM
database through the getData() function. The
ISO3 code of Algeria is DZA and we will need 2 levels of
administrative data (0 for the state and 2 for the
municipalities).
Basemaps (SpatialPolygonsDataFrame) are downloaded unprojected in
WGS84. We have to transform them into simple feature collections (sf
objects) and project them in a valid projection (the EPSG code 30791
stands for “Nord Sahara 1959 / Voirol Unifie Nord”).
library(raster)
library(sf)
# download data from gadm.org
adm0 <-st_as_sf( getData(name = "GADM", country = "DZA", level = 0, path = tempdir() ))
adm0 <- st_transform(adm0, 30791)
adm2 <- st_as_sf(getData(name = "GADM", country = "DZA", level = 2, path = tempdir()))
adm2 <- st_transform(adm2, 30791)
# plot
plot(st_geometry(adm0), lwd = 2)
plot(st_geometry(adm2), add = T, col = "grey80", lwd = 0.2)
Our goal is to extract Alger municipalities. To extract those
munuicipalities we will use a mask.
This mask has to be an sf object. The create_mask function create a
mask based on a bbox, a sf or sfc extent or an interactively defined
rectangle.
In this example we use a mask defined by Alger willaya extent + a 2 km
buffer.
library(mapinsetr)
alger_extent <- st_buffer(adm2[adm2$NAME_1=="Alger", ],dist = 2000)
box_alger <- create_mask(bb = alger_extent)
plot(st_geometry(adm0), lwd = 2)
plot(st_geometry(adm2), add=T, col = "grey80", lwd = 0.2)
plot(st_geometry(box_alger), border = "red", add=T, lwd = 2)
Alternatively, one can define the mask interactively on the plot :
library(mapinsetr)
plot(st_geometry(adm0), lwd = 2)
plot(st_geometry(adm2), add=T, col = "grey80", lwd = 0.2)
box_alger_interactive <- create_mask(interactive = TRUE, add=TRUE,
prj = st_crs(adm0))
The move_and_resize function extracts polygons or multipolygons (x)
within a mask (mask), moves them to coordinates (xy, bottom left
corner of the inset) and resizes them according to a factor (k).
zoom_alger <- move_and_resize(x = adm2, mask = box_alger, xy = c(-1000000, -100000), k = 15)
box <- move_and_resize(x = box_alger, mask = box_alger, xy = c(-1000000, -100000), k = 15)
plot(st_geometry(adm0), lwd = 2)
plot(st_geometry(adm2), add=T, col = "grey80", lwd = 0.2)
plot(st_geometry(box_alger), border = "red", add=T, lwd = 2)
plot(st_geometry(zoom_alger), add=T, col = "grey80", lwd = 0.2)
plot(st_geometry(box), border = "red", lwd = 1, add=T)
We currently have 2 sf objects, the next step is to merge them in a
single one inset_rbinder merge the initial basemap and the inset.
adm2final <- inset_rbinder(list(adm2, zoom_alger))
plot(st_geometry(adm2final), col = "grey80", lwd = 0.2)
title("Final basemap with inset!")
The following example build a map of Municipalities types.
library(cartography)
adm2final[adm2final$CCA_2==2223, "TYPE_2"] <- "Commune"
plot(st_geometry(adm0), lwd = 2)
typoLayer(adm2final, var = "TYPE_2",
legend.title.txt = "Commune Type",
border = "white", lwd = 0.1,
col = carto.pal("multi.pal", 3),
legend.pos = "topright", add = T)
plot(st_geometry(box_alger), border = "red", add=T, lwd = 2)
plot(st_geometry(box), border = "red", lwd = 2, add=T)
layoutLayer(title = "Commune Types in Algeria with a Zoom on Alger Region",
sources = "GADM 2.8, 2015", author = "@rCarto", theme = "green.pal")