Getting started with happign

Before starting

We can load the happign package, and some additional packages we will need (sf to manipulate spatial data and tmap to create maps)

library(happign)
library(sf)
library(tmap);tmap_mode("plot")
#> tmap mode set to plotting

WFS, WMS and WMTS service

happign use three web service from IGN :

WMS raster : data in raster format e.g. images (.jpg, .png, .tif, …)
WMTS : same as WMS raster but images are precalculated
WFS : data in vector format (.shp, …).

More detailed information are available here for WMS, here for WMTS and here for WFS.

To download data from IGN web services at least two elements are needed :

A layer name ;
An input shape read by sf package.

Layer name

It is possible to find the names of available layers from the IGN website. For example, the first layer name in WFS format for “Administratif” category is “ADMINEXPRESS-COG-CARTO.LATEST:arrondissement”

All layer’s name can be accessed from R with the get_layers_metadata() function. This one connects directly to the IGN site which allows to have the last updated resources. It can be used for WMS and WFS :

administratif_wfs <- get_layers_metadata(data_type = "wfs")
administratif_wms <- get_layers_metadata(data_type = "wms-r")
administratif_wms <- get_layers_metadata(data_type = "wmts")

head(administratif_wfs)
#>                                                       Name
#> 1              OCS-GERS_BDD_LAMB93_2016:oscge_gers_32_2016
#> 2              OCS-GERS_BDD_LAMB93_2019:oscge_gers_32_2019
#> 3                   ADMINEXPRESS-COG.LATEST:arrondissement
#> 4         ADMINEXPRESS-COG.LATEST:arrondissement_municipal
#> 5                           ADMINEXPRESS-COG.LATEST:canton
#> 6 ADMINEXPRESS-COG.LATEST:chflieu_arrondissement_municipal
#>                   Title          Abstract
#> 1       OCSGE Gers 2016   OCSGE Gers 2016
#> 2      OCSGE Gers 2019   OCSGE Gers 2019 
#> 3 ADMINEXPRESS COG 2023      édition 2023
#> 4 ADMINEXPRESS COG 2023      édition 2023
#> 5 ADMINEXPRESS COG 2023      édition 2023
#> 6 ADMINEXPRESS COG 2023      édition 2023

You can specify an apikey to focus on specific category. API keys can be directly retrieved on the IGN website from the expert web services or with get_apikeys() function.

get_apikeys()
#>  [1] "administratif" "adresse"       "agriculture"  
#>  [4] "altimetrie"    "cartes"        "cartovecto"   
#>  [7] "clc"           "economie"      "enr"          
#> [10] "environnement" "geodesie"      "lambert93"    
#> [13] "ocsge"         "ortho"         "orthohisto"   
#> [16] "parcellaire"   "satellite"     "sol"          
#> [19] "topographie"   "transports"

administratif_wmts <- get_layers_metadata("wmts", "administratif")

head(administratif_wmts)
#>                                              Title
#> 1                           ADMINEXPRESS COG CARTO
#> 2                                 ADMINEXPRESS COG
#> 3 Limites administratives mises à jour en continu.
#>                                                                 Abstract
#> 1    Limites administratives Express COG code officiel géographique 2023
#> 2   Limites administratives Express COG code officiel géographique. 2023
#> 3 Limites administratives mises à jour en continu ; Edition : 2024-03-25
#>                               Identifier
#> 1          ADMINEXPRESS-COG-CARTO.LATEST
#> 2                ADMINEXPRESS-COG.LATEST
#> 3 LIMITES_ADMINISTRATIVES_EXPRESS.LATEST

Downloading the data

Now that we know how to get a layer name, it only takes a few lines to get plethora of resources. For the example we will look at the beautiful town of Penmarch in France. A part of this town is stored as a shape in happign.

penmarch <- read_sf(system.file("extdata/penmarch.shp", package = "happign"))

WFS

get_wfs can be used to download borders :


penmarch_borders <- get_wfs(x = penmarch,
                            layer = "LIMITES_ADMINISTRATIVES_EXPRESS.LATEST:commune")
#> Features downloaded : 1
# Checking result
tm_shape(penmarch_borders)+
   tm_polygons(alpha = 0, lwd = 2)+
tm_shape(penmarch)+
   tm_polygons(col = "red")+
tm_add_legend(type = "fill", border.col = "black", border.lwd =2,
              col = NA, labels = "border from get_wfs")+
tm_add_legend(type = "fill", col = "red", labels = "penmarch shape from happign package")+
tm_layout(main.title = "Penmarch borders from IGN",
          main.title.position = "center",
          legend.position = c(0.7, -0.1),
          outer.margins = c(0.1, 0,0,0),
          frame = FALSE)
#> Legend labels were too wide. The labels have been resized to 0.61. Increase legend.width (argument of tm_layout) to make the legend wider and therefore the labels larger.

plot of chunk unnamed-chunk-5

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 hedges for biodiversity there are in Penmarch?

Spoiler : there are 436 of them !

hedges <- get_wfs(x = penmarch_borders,
                 layer = "BDTOPO_V3:haie",
                 spatial_filter = "intersects")
#> Features downloaded : 436

# Checking result
tm_shape(penmarch_borders) + # Borders of penmarch
   tm_borders(lwd = 2) +
tm_shape(hedges) + # Point use to retrieve data
   tm_lines(col = "red", size = 0.3) +
   tm_add_legend(type = "line", label = "Hedges", col = "red") +
   tm_layout(main.title = "Hedges recorded by the IGN in Penmarch",
             main.title.position = "center",
             legend.position = c("right", "bottom"),
             frame = FALSE)

plot of chunk unnamed-chunk-6

WMS raster

For raster, the process is the same, but with the function get_wms_raster(), but you need to specify the resolution (note that it must be in the same coordinate system as the crs parameter). There’s plenty of elevation resources inside “altimetrie” category. A basic one is the Digital Elevation Model (DEM or MNT in French). Borders of Penmarch are used to download the DEM. Note that for DEM, we don’t want an RGB image but values of each pixels. That why rgb=FALSE is used below.

layers_metadata <- get_layers_metadata("wms-r", "altimetrie")
dem_layer <- layers_metadata[2, 1] #LEVATION.ELEVATIONGRIDCOVERAGE

mnt <- get_wms_raster(x = penmarch_borders,
                      layer = dem_layer,
                      res = 25,
                      crs = 2154,
                      rgb = FALSE)
#> 0...10...20...30...40...50...60...70...80...90...100 - done.
#> Raster is saved at : C:\Users\PaulCarteron\AppData\Local\Temp\Rtmp63I4Tv\filea98465250a.tif

mnt[mnt < 0] <- NA # remove negative values in case of singularity

tm_shape(mnt) +
   tm_raster(title = "Elevation [m]") +
tm_shape(penmarch_borders)+
   tm_borders(lwd = 2)+
tm_layout(main.title = "DEM of Penmarch",
          main.title.position = "center",
          legend.position = c("right", "bottom"),
          legend.bg.color = "white", legend.bg.alpha = 0.7)

plot of chunk unnamed-chunk-7

Rq :

Raster from get_wms_raster() are SpatRaster object from the terra package. To learn more about conversion between other raster type in R go check this out.

WMTS

For WMTS, no resolution is needed beacause images are precalculated but a zoom level is needed. The higher the zoom level is, the more precis image is. If you only need visualisation, i recommend to use WMTS instead of WMS.

layers_metadata <- get_layers_metadata("wmts", "ortho")
ortho_layer <- layers_metadata[1, 3] #HR.ORTHOIMAGERY.ORTHOPHOTOS

hr_ortho <- get_wmts(x = penmarch_borders,
                     layer = ortho_layer,
                     zoom = 14)
#> 0...10...20...30...40...50...60...70...80...90...100 - done.

tm_shape(hr_ortho) +
   tm_rgb(title = "Orthophoto Hight Resolution") +
tm_shape(penmarch_borders)+
   tm_borders(lwd = 2)+
tm_layout(main.title = "Orthophoto Hight Resolution",
          main.title.position = "center",
          legend.position = c("right", "bottom"),
          legend.bg.color = "white", legend.bg.alpha = 0.7)

plot of chunk unnamed-chunk-8

Paul Carteron