Skip to content

Multiple legends with levelplot and spplot

2 messages · kirsty fulton, Oscar Perpiñan

Original
kirsty fulton
# 
I am trying to overlay two shapefiles on a raster grid. The raster is a digital elevation model. The first shapefile is a polygon, which shows the outline of a country, and the second shapefile is a polyline, which represents rivers. I am currently using levelplot to plot the raster and spplot to overlay the polyline and polygon. The polyline is coloured according to the attribute table using the STATUS column - this plots "perennial" status as blue and "ephemeral" status as orange. My code is shown below. This produces a plot with a colorkey for the raster, but I can't work out how to add a key for the polyline. Apologies if this is very obvious but I can't seem to find an answer on any existing threads.
Any help would be much appreciated. 

The code is as follows:
# Read rasterdata_location_DEM1 <- "E:/AGA_ProcessDatasets/DEM_Data/gt30e020n40.tif"iDEM <- raster(data_location_DEM1)
# Read shapefilesioutlineCountry <- readOGR(dsn = "E:/AGA_ProcessDatasets", layer = "UNAfricanCountries_Final_Buffer100m")proj4string(ioutlineCountry) <- CRS("+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0")
irivers <- readOGR(dsn = "E:/AGA_ProcessDatasets/Hydrology_Data", layer = "Rivers_Africa_37333_ForPlotting_Split")proj4string(irivers) <- CRS("+proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0")
# define colours for riversrivercolours <- c("orange","blue")
# define colours for DEMcols <- gray.colors(128,start=1,end=0)
# define extents for plottingplotmin <- cellStats(iDEM,stat='min')plotmax <- cellStats(iDEM,stat='max')breaks <- (plotmax - plotmin)/128legendbreaks <- round((plotmax - plotmin)/5,0)scale <- seq(plotmin, plotmax, by=breaks)xmin <- iDEM at extent@xminxmax <- iDEM at extent@xmaxymin <- iDEM at extent@yminymax <- iDEM at extent@ymax
# define output png filefilenamePre = paste(iCountryName,"_Hydrology2.png",sep="")png(filename="Hydrology.png",width=1000,height=486,units="px")
# plot dataa<-levelplot(iDEM,margin=FALSE,xlab="",ylab="",at=scale,col.regions=cols,             colorkey=list(space="right",width=0.75,height=0.75)) +   spplot(irivers,zcol="STATUS",col.regions=rivercolours) +   spplot(ioutlineCountry,zcol="NAME_0",fill="transparent",col="black") # add title to legend a$legend$right <-   list(fun = mergedTrellisLegendGrob(a$legend$right,list(fun = textGrob,args = list("maSL",rot = 0)),vertical = FALSE))
# printprint(a,split=c(1,1,2,1),newpage=FALSE)
dev.off()
3 days later
Oscar Perpiñan
# 
Hello,

You could try to define the legend for the Spatial* object inside the
levelplot call using the `key` argument. For example:

library(sp)
library(raster)
library(rasterVis)

f <- system.file("external/test.grd", package="raster")
r <- raster(f)

data(meuse)
coordinates(meuse) = c("x", "y")
meuse$class <- cut(meuse$zinc, 3)

ptsCols <- c('black', 'red', 'blue')

lvp <- levelplot(r, margin = FALSE,
                 ## Define the legend for the spplot graphic
                 key = list(space = 'top', text = list(levels(meuse$class)),
                     points = list(pch = 21, fill = ptsCols)),
                 ## Legend for the Raster object
                 colorkey=list(space="right",width=0.75,height=0.75))

spp <- spplot(meuse["class"], col.regions = ptsCols)

lvp + spp


Best,

Oscar.