Downloading the data
WFS
For the example we will look at the beautiful town of Penmarch in France and we are going to get these borders.
The get_wfs() function connects to WFS service of the IGN that contain borders of the communes. It is also necessary to take a point inside penmarch so that the function detects all the shape that intersect this point.
penmarch_point <- st_sfc(st_point(c(-4.370, 47.800)), crs = 4326)
penmarch_borders <- get_wfs(shape = penmarch_point,
apikey = "administratif",
layer_name = "LIMITES_ADMINISTRATIVES_EXPRESS.LATEST:commune")
#> Request 1/1 downloading...
# Checking result
tm_shape(penmarch_borders) + # Borders of penmarch
tm_polygons(alpha = 0, lwd = 2) +
tm_shape(penmarch_point) + # Point use to retrieve data
tm_dots(col = "red", size = 2) +
tm_add_legend(type = "symbol", label = "lat : -4.370, long : 47.800",
col = "red", size = 1) +
tm_layout(main.title = "Penmarch borders from IGN",
main.title.position = "center",
legend.position = c("right", "bottom"),
frame = FALSE)
It’s as simple as that! Now you have to rely on your curiosity to explore the multiple possibilities that IGN offers. For example, who has never wondered how many road junctions there are in Penmarch?
Spoiler : there are 220 of them
dikes <- get_wfs(shape = penmarch_borders,
apikey = get_apikeys()[6],
layer_name = "BDCARTO_BDD_WLD_WGS84G:noeud_routier")
#> Request 1/1 downloading...
# Checking result
tm_shape(penmarch_borders) + # Borders of penmarch
tm_polygons(alpha = 0, lwd = 2) +
tm_shape(dikes) + # Point use to retrieve data
tm_dots(col = "red") +
tm_add_legend(type = "symbol", label = "Road junction", col = "red") +
tm_layout(main.title = "Road nodes recorded by the IGN in Penmarch",
main.title.position = "center",
legend.position = c("right", "bottom"),
frame = FALSE)
WMS raster
For raster, the process is the same but with the function get_wms_raster(). There’s plenty of elevation resources inside “altimetrie” category. The basic one is the Digital Elevation Model (DEM or MNT in French).
Rq :
- The raster is not loaded into R’s memory. It is an on-the-fly sequential reading of files from network, without downloading the whole file first. In other words, it is quite possible to access the DEM of an entire region. The counterpart is the connection time which is about 30s;
- Raster from
get_wms_raster()are stars object from thestarspackage. To learn more about conversion between other raster type in R go check this out.
apikey <- get_apikeys()[4]
layers_metadata <- get_layers_metadata("altimetrie", "wms")
dem_layer_name <- layers_metadata[2, "name"]
mnt <- get_wms_raster(shape = penmarch_borders,
apikey = apikey,
layer_name = dem_layer_name,
resolution = 25)
#> x cell_size : 24.873 [m]
#> y cell_size : 24.937 [m]
mnt[mnt < 0] <- NA # remove negative values in case of singularity
names(mnt) <- "Elevation [m]" # Rename raster ie the title legend
tm_shape(mnt) +
tm_raster(colorNA = NULL) +
tm_layout(title = "DEM of Penmarch", frame = FALSE)