Cartographie avec R : le package cartography

SEMIN-R
Museum National d’Histoire Naturelle | Vendredi 16 juin 2017

Timothée Giraud
UMS RIATE
http://rgeomatic.hypotheses.org/

Les fonctionnalités spatiales de R

Les indispensables

Import / Export

rgdal est une interface entre R et les librairies GDAL (Geospatial Data Abstraction Library) et PROJ4.

library("rgdal")
nuts3 <- readOGR(dsn = "data", layer = "nuts3")
OGR data source with driver: ESRI Shapefile 
Source: "data", layer: "nuts3"
with 1448 features
It has 7 fields

Manipulation et affichage

sp fournit des classes et des methodes pour les données spatiales dans R.

library("sp")
plot(nuts3)

plot(nuts3, col = "#DAE3E6", border = "#8A0641", lwd = 0.5)

Géotraitements

rgeos donne accès à la librairie d’opérations spatiales GEOS (Geometry Engine - Open Source) qui permet notamment d’effectuer les géotraitements suivants :

  • Area / Perimeter
  • Distances
  • Dissolve
  • Buffer
  • Overlap / intersect / difference
  • Contains / within
  • Union

Agrégation des polygones / dissolve

library("rgeos")
europe <- gUnaryUnion(spgeom = nuts3)
plot(nuts3, lwd = 0.5)
plot(europe, lwd = 2, border = "red", add=T)

Création de zones tampons / buffer

library("rgeos")
europeBuffer <- gBuffer(spgeom = europe, width = 50000)
plot(europe, col = "#92C5D6")
plot(europeBuffer, add = T, border = "red")

Le futur : le package sf

  • Première release : 31 octobre 2016
  • Auteur principal et maintainer : Edzer Pebesma
  • Financement :

Avec sp et rgdal :

library('sp')
library('rgdal')
nuts3 <- readOGR(dsn = "data", layer = "nuts3", verbose = FALSE)
str(nuts3[1:3,])
Formal class 'SpatialPolygonsDataFrame' [package "sp"] with 5 slots
  ..@ data       :'data.frame': 3 obs. of  7 variables:
  .. ..$ id        : Factor w/ 1448 levels "AT111","AT112",..: 1 2 3
  .. ..$ birth_2008: num [1:3] 272 1195 748
  .. ..$ death_2008: num [1:3] 445 1480 1142
  .. ..$ gdppps1999: num [1:3] 509 2262 1368
  .. ..$ gdppps2008: num [1:3] 641 3272 1783
  .. ..$ pop1999   : num [1:3] 39149 137469 100114
  .. ..$ pop2008   : num [1:3] 37452 146383 97350
  ..@ polygons   :List of 3
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 1
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] 4806396 2728404
  .. .. .. .. .. .. ..@ area   : num 8.58e+08
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:10, 1:2] 4819333 4824015 4808360 4802891 4797862 ...
  .. .. .. ..@ plotOrder: int 1
  .. .. .. ..@ labpt    : num [1:2] 4806396 2728404
  .. .. .. ..@ ID       : chr "0"
  .. .. .. ..@ area     : num 8.58e+08
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 1
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] 4827772 2765414
  .. .. .. .. .. .. ..@ area   : num 1.81e+09
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:9, 1:2] 4819333 4806425 4800667 4793126 4799329 ...
  .. .. .. ..@ plotOrder: int 1
  .. .. .. ..@ labpt    : num [1:2] 4827772 2765414
  .. .. .. ..@ ID       : chr "1"
  .. .. .. ..@ area     : num 1.81e+09
  .. ..$ :Formal class 'Polygons' [package "sp"] with 5 slots
  .. .. .. ..@ Polygons :List of 1
  .. .. .. .. ..$ :Formal class 'Polygon' [package "sp"] with 5 slots
  .. .. .. .. .. .. ..@ labpt  : num [1:2] 4795824 2692300
  .. .. .. .. .. .. ..@ area   : num 1.34e+09
  .. .. .. .. .. .. ..@ hole   : logi FALSE
  .. .. .. .. .. .. ..@ ringDir: int 1
  .. .. .. .. .. .. ..@ coords : num [1:9, 1:2] 4808360 4813350 4781472 4790339 4776667 ...
  .. .. .. ..@ plotOrder: int 1
  .. .. .. ..@ labpt    : num [1:2] 4795824 2692300
  .. .. .. ..@ ID       : chr "2"
  .. .. .. ..@ area     : num 1.34e+09
  ..@ plotOrder  : int [1:3] 2 3 1
  ..@ bbox       : num [1:2, 1:2] 4776667 2655076 4855527 2793852
  .. ..- attr(*, "dimnames")=List of 2
  .. .. ..$ : chr [1:2] "x" "y"
  .. .. ..$ : chr [1:2] "min" "max"
  ..@ proj4string:Formal class 'CRS' [package "sp"] with 1 slot
  .. .. ..@ projargs: chr "+proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs"

Avec sf :

library(sf)
nuts3 <- st_read(dsn = "data", layer = "nuts3", quiet = TRUE)
str(nuts3[1:3,])
Classes 'sf' and 'data.frame':  3 obs. of  8 variables:
 $ id        : Factor w/ 1448 levels "AT111","AT112",..: 1 2 3
 $ birth_2008: num  272 1195 748
 $ death_2008: num  445 1480 1142
 $ gdppps1999: num  509 2262 1368
 $ gdppps2008: num  641 3272 1783
 $ pop1999   : num  39149 137469 100114
 $ pop2008   : num  37452 146383 97350
 $ geometry  :sfc_MULTIPOLYGON of length 3; first list element: List of 1
  ..$ :List of 1
  .. ..$ : num [1:10, 1:2] 4819333 4824015 4808360 4802891 4797862 ...
  ..- attr(*, "class")= chr  "XY" "MULTIPOLYGON" "sfg"
 - attr(*, "sf_column")= chr "geometry"
 - attr(*, "agr")= Factor w/ 3 levels "constant","aggregate",..: NA NA NA NA NA NA NA
  ..- attr(*, "names")= chr  "id" "birth_2008" "death_2008" "gdppps1999" ...

Import / Export

library(sf)
nuts3 <- st_read(dsn = "data", layer = "nuts3")
Reading layer `nuts3' from data source `/home/tg/Documents/prz/semin-r_2017/data' using driver `ESRI Shapefile'
converted into: POLYGON
Simple feature collection with 1448 features and 7 fields
geometry type:  MULTIPOLYGON
dimension:      XY
bbox:           xmin: 2641758 ymin: 1427614 xmax: 7313157 ymax: 5411284
epsg (SRID):    NA
proj4string:    +proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs

Manipulation et affichage

plot(st_geometry(nuts3))

plot(st_geometry(nuts3), col = "#DAE3E6", border = "#8A0641", lwd = 0.5)

Géotraitements

Agrégation des polygones / dissolve

europe <- st_union(x = nuts3)
plot(st_geometry(nuts3), lwd = 0.5)
plot(europe, lwd = 2, border = "red", add=T)

Création de zones tampons / buffer

europeBuffer <- st_buffer(x = europe, dist = 50000)
plot(st_geometry(europe), col = "#92C5D6")
plot(europeBuffer, add = T, border = "red")

Le package cartography

Installation

  • Version stable (CRAN)
install.packages("cartography")
  • Version de développement (Github)

Cette version permet d’utiliser les objets sf.

devtools::install_github(repo = "Groupe-ElementR/cartography", ref = "devsf")

Utilisation

Cartes choroplèthes

library(cartography)

# chargement de données
data(nuts2006)

# Calcul du taux de croissance annuel moyen
nuts2.df$cagr <- 100 * (((nuts2.df$pop2008 / nuts2.df$pop1999)^(1/9)) - 1) 

# Cartographie
choroLayer(spdf = nuts2.spdf, df = nuts2.df, var = "cagr")
title("Taux de croissance en Europe")

Après ce premier jet, il est ensuite possible de paramétrer très finement la carte : palette de couleurs, discrétisation, légende, couches d’habillage…

# Construire une palette de couleurs
cols <- carto.pal(pal1 = "green.pal", n1 = 2, 
                  pal2 = "red.pal", n2 = 4) 

# Affichage de couches d'habillage
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
plot(world.spdf, col  = "#E3DEBF", border=NA, add=TRUE)

# Cartographie du taux de croissance annuel moyen
choroLayer(spdf = nuts2.spdf, df = nuts2.df, var = "cagr",
           breaks = c(-2.43,-1.0,0.0,0.5,1.0,2.0,3.1), 
           col = cols,
           border = "grey40",
           lwd = 0.5, 
           legend.pos = "right",
           legend.title.txt = "taux de croissance\nannuel moyen", 
           legend.values.rnd = 2, 
           add = TRUE) 

# Affichage de couches d'habillage
plot(nuts0.spdf,border = "grey20", lwd=0.75, add=TRUE)

# Ajout des titres, légende, sources, etc.
layoutLayer(title = "Taux de croissance en Europe", 
            author = "cartography", 
            sources = "Eurostat, 2008", frame = TRUE, col = NA, 
            scale = NULL,coltitle = "black",
            south = TRUE) 

Cartes en symboles proportionnels

Cartographie d’un stock (la population nationale) avec des figurés proportionnels.

# Affichage de couches d'habillage
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
plot(world.spdf, col  = "#E3DEBF", border=NA, add=TRUE)
plot(nuts0.spdf, col = "#D1914D",border = "grey80", add=TRUE)

# Cartographie de la population des pays en cercles proportionnels
propSymbolsLayer(spdf = nuts0.spdf, df = nuts0.df,
                 var = "pop2008", 
                 symbols = "circle", col =  "seagreen4",
                 legend.pos = "right", inches = 0.35,
                 legend.title.txt = "Total\npopulation (2008)",
                 legend.style = "c")

# Ajout de labels
dflab <- nuts0.df[order(nuts0.df$pop2008, decreasing = TRUE),][1:8,]
dflab$lab <- paste(dflab$id, "\n", round(dflab$pop2008/1000000,0), "M", sep ="")

# Label plot of the 8 most populated countries
labelLayer(spdf = nuts0.spdf, 
           df = dflab, 
           txt = "lab", 
           col = "#690409", 
           cex = 0.8, 
           font = 2) 

# Ajout des titres, légende, sources, etc.
layoutLayer(title = "Countries Population in Europe",
            theme = 'green.pal',
            frame = FALSE,
            author = "cartography",
            sources = "Eurostat, 2008",
            scale = NULL,
            south = TRUE)

Cartes en symboles proportionnels colorés

# Load data
data(nuts2006)

# Compute the compound annual growth rate
nuts2.df$cagr <- (((nuts2.df$pop2008 / nuts2.df$pop1999)^(1/9)) - 1) * 100

# Plot a layer with the extent of the EU28 countries with only a background color
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
# Plot non european space
plot(world.spdf, col  = "#E3DEBF", border = NA, add = TRUE)
# Plot Nuts2 regions
plot(nuts2.spdf, col = "grey60",border = "white", lwd = 0.4, add = TRUE)

# Set a custom color palette
cols <- carto.pal(pal1 = "blue.pal", n1 = 2, pal2 = "red.pal", n2 = 4)

# Plot symbols with choropleth coloration
propSymbolsChoroLayer(spdf = nuts2.spdf, 
                      df = nuts2.df, 
                      var = "pop2008", 
                      inches = 0.1, 
                      var2 = "cagr", 
                      col = cols, 
                      breaks = c(-2.43,-1,0,0.5,1,2,3.1), 
                      border = "grey50",  
                      lwd = 0.75, 
                      legend.var.pos = "topright", 
                      legend.var.values.rnd = -3,
                      legend.var.title.txt = "Total Population", 
                      legend.var.style = "e", 
                      legend.var2.pos = "right", 
                      legend.var2.title.txt = "Compound Annual\nGrowth Rate") 

# layout
layoutLayer(title = "Demographic trends, 1999-2008", coltitle = "black",
            sources = "Eurostat, 2011", scale = NULL,
            author = "cartography", frame ="", col = NA)

Cartes de flux

Il s’agit de représenter des données, agrégées à un niveau régional, sur les jumelages entre villes.

# Données sur les jumelages
head(twincities.df)
i j fij
DE14 AT11 1
DE21 AT11 1
DE23 AT11 1
DE26 AT11 2
DE91 AT11 1
DEB3 AT11 1
# Creation d'une couche de liens
twincities.sf <- getLinkLayer(x = nuts2.spdf, df = twincities.df[,1:2])

# Affichage des liens créés
plot(st_geometry(twincities.sf), lwd = 0.2)

# Affichage de couches d'habillage
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
plot(world.spdf, col  = "#E3DEBF", border=NA, add=TRUE)
plot(nuts2.spdf, col = "#D1914D",border = "grey80", add=TRUE)

# Cartographie des liens
gradLinkLayer(x = twincities.sf, df = twincities.df,   
              var = "fij", 
              breaks = c(2,5,15,20,30), 
              lwd = c(0.1,1,4,10), 
              col = "#92000090",
              legend.pos = "right", legend.frame = TRUE,
              legend.title.txt = "Number of Agreements\n(regional level)",
              add = TRUE)

# Ajout des titres, légende, sources, etc.
layoutLayer(title = "International Twinning Agreements Between Cities", 
            author = "cartography", 
            sources = "Sources: Adam Ploszaj & Wikipedia, 2011",
            scale = NULL, south = TRUE, frame = TRUE, col = NA, 
            coltitle = "black")

Discontinuités

# Load data
data(nuts2006)

# Get a SpatialLinesDataFrame of countries borders
nuts0.contig <- getBorders(spdf = nuts0.spdf)

plot(nuts0.spdf, col = "grey", border = NA)
plot(st_geometry(nuts0.contig), 
     col = 1:nrow(nuts0.contig), 
     add=T, lwd = 4)

# Get the GDP per capita
nuts0.df$gdpcap <- nuts0.df$gdppps2008/nuts0.df$pop2008*1000000

# Plot a layer with the extent of the EU28 countries with only a background color
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
# Plot non european space
plot(world.spdf, col  = "#E3DEBF", border=NA, add=TRUE)

# Plot GDP per capita with a choropleth layer
choroLayer(spdf = nuts0.spdf, df = nuts0.df, var = "gdpcap", border = "grey80",
           col = carto.pal(pal1 = "kaki.pal", n1 = 6), method = "quantile",
           nclass = 6, add=TRUE, legend.pos = "right", 
           legend.values.rnd = -2,
           legend.title.txt = "GDP per Capita\n(in euros)")

# Plot discontinuities
discLayer(x = nuts0.contig, # sf of borders
          df = nuts0.df, # data frame on countries
          var = "gdpcap", # variable used to compute discontinuties 
          type = "rel", # type of discontinuity measure 
          method="equal", # discretisation of discontinuities
          nclass=4, # number of discontinuities classes
          threshold = 0.5, # representation threshold of discontinuities  
          sizemin = 0.5, # minimum size of discontinuities lines
          sizemax = 6, # maximum size of discontinuities lines
          col="red", # color of the lines
          legend.values.rnd = 1,
          legend.title.txt = "Discontinuities in \nGDP per Capita\n(relative)",
          legend.pos = "topright", 
          add=TRUE)

# Layout
layoutLayer(title = "Wealth Disparities in Europe", coltitle = "black",
            sources = "Eurostat, 2011", scale = NULL,
            author = "cartography", frame ="", col = NA)

Carroyages

library(cartography)
# Load data
data(nuts2006)

# Create a grid layer
nuts2.spdf@data <- nuts2.df
mygrid <- getGridLayer(x = nuts2.spdf, 
                       cellsize = 200000 * 200000, 
                       var = "pop2008")
                       
# Plot dentsity of population
## conversion from square meter to square kilometers
mygrid$densitykm <- mygrid$pop2008 * 1000 * 1000 / mygrid$gridarea

# Plot a layer with the extent of the EU28 countries with only a background color
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")
# Plot non european space
plot(world.spdf, col  = "#E3DEBF", border=NA, add=TRUE)

# Set a custom color palette
cols <- carto.pal(pal1 = "wine.pal", n1 = 6)

# Plot the gridded population density
choroLayer(x = mygrid, var = "densitykm", 
           border = "grey80", col = cols, legend.pos = "topright",
           method = "q6", add = TRUE, legend.values.rnd = 1,
           legend.title.txt = "Population Density\n(inhabitant/km²)")

# Layout
layoutLayer(title = "Population Density", coltitle = "black",
            sources = "Eurostat, 2011", scale = NULL, 
            author = "cartography", frame ="", col = NA)

Cartes lissées

# set margins
opar <- par(mar = c(0, 0, 1.2, 0))
# Load data
data(nuts2006)

nuts3.spdf@data = nuts3.df
# Create a grid layer
mygrid <- getGridLayer(x = sf::st_as_sf(nuts3.spdf), 
                       cellsize = 50000 * 50000, 
                       type = "regular", 
                       var = c("pop2008", "gdppps2008"))


# Compute data for the grid layer
mygrid$gdp <- mygrid$gdppps2008*1000000

# list of breaks
v <- c(2920, 5000, 10000, 15000, 20000, 23500, 30000, 35000, 40000, 42720)
# Plot a layer with the extent of the EU28 countries with only a background
# color
plot(nuts0.spdf, border = NA, col = NA, bg = "#A6CAE0")

# Plot non european space
plot(world.spdf, col = "#E3DEBF", border = NA, add = TRUE)

# set a color palette
cols <- c(rev(carto.pal("green.pal", 5)), carto.pal("orange.pal", 4))

# compute & display the potential map
smoothLayer(x = mygrid, var = "gdp", var2 = "pop2008", breaks = v, 
            span = 1e+05, beta = 2, mask = nuts0.spdf, resolution = 49000, col = cols, 
            legend.title.txt = "Potential\nGDP per capita\n(in euros)", legend.values.rnd = -2, 
            border = "grey80", lwd = 0.5, add = T, legend.pos = "topright")

# plot Europe contour
plot(rgeos::gBuffer(nuts0.spdf, FALSE, 1), add = T, col = NA, border = "grey50")

# plot a layout
layoutLayer(title = "Wealth Inequalities in Europe, 2008", 
            author = "Package cartography v2.0.0", 
            sources = "Source: Eurostat, 2011", frame = TRUE, scale = 500, north = FALSE, 
            theme = "sand.pal")

# plot a text on the map
text(x = 6271272, y = 3743765, labels = "Distance function:\n- type = exponential\n- beta = 2\n- span = 100 km", 
     cex = 0.8, adj = 0, font = 3)

Ressources

La présentation est accessible à cette adresse :
https://rcarto.github.io/semin-r_2017

Son code source est accessible sur GitHub.

Package sf

La page GitHub du package sf

C’est dans ce dépôt GitHub que se déroule le développement du package et que se tiennent les discussions à son sujet.

Les ressources publiées par Edzer Pebesma

Le créateur et maintainer de sf a publié un certain de nombre de vignettes et de billets autour du package.

Billets de blogs / tutoriels

La cartographie

Béguin & Pumain (2003)

Béguin & Pumain (2003)

  • Michelle Béguin et Denise Pumain. “La représentation des données géographiques, Statistique et cartographie.” (2003). Paris, Armand Colin, Coll. Cursus, 192p.
Lambert & Zanin (2016)

Lambert & Zanin (2016)

  • Nicolas Lambert et Christine Zanin. “Manuel de cartographie: principes, méthodes, applications.” (2016). Paris, Armand Colin, Coll. Cursus, 224p.