Difference between revisions of "Plotting maps"

From SourceWiki
Jump to navigation Jump to search
Line 3: Line 3:
 
==Introduction==
 
==Introduction==
  
I decided to give up on plotting nice maps with [http://grass.itc.it grass]. The ps.map function is very powerful, but hard to remember for my not-so-frequent use. R has great plotting capabilities, and of course it does maps. It has an easy [http://cran.r-project.org/src/contrib/Descriptions/spgrass6.html grass interface]. What else do you want (importing other GIS formats? Try [http://cran.r-project.org/src/contrib/Descriptions/rgdal.html rgdal]). Plotting GIS data is very nicely implemented in the [http://cran.r-project.org/src/contrib/Descriptions/sp.html sp] package, part of the [http://r-spatial.sourceforge.net/ spatial data in R] project, but getting your plot the way you want is quite challenging.  
+
R can plot very nice maps. I decided to give up plotting maps in [http://grass.itc.it grass]. The ps.map function is very powerful, but hard to remember for my not-so-frequent use. I use the [http://cran.r-project.org/src/contrib/Descriptions/spgrass6.html grass interface] but R imports very many [http://www.gdal.org/formats_list.html raster] and [http://www.gdal.org/ogr/ogr_formats.html vector] formats via the [http://cran.r-project.org/src/contrib/Descriptions/rgdal.html rgdal] package.
 +
 
 +
Plotting GIS data is very nicely implemented in the [http://cran.r-project.org/src/contrib/Descriptions/sp.html sp] package, part of the [http://r-spatial.sourceforge.net/ spatial data in R] project, but getting your plot the way you want is quite challenging. Here is an example.
  
 
==Required packages==
 
==Required packages==
  
spgrass6 reads data from the GRASS database. Sp implements spplot and lattice is used to define the colours (trells.par.set).
+
spgrass6 reads data from the GRASS database, use rgdal for other formats. sp implements spplot, and lattice is used to define the colours (trells.par.set).
  
 
   library(sp)
 
   library(sp)
Line 13: Line 15:
 
   library(lattice)
 
   library(lattice)
  
==Procedure==
+
==Plotting a map==
 
 
Here's what I came up with:
 
  
      paute <- rast.get6("paute_z")
+
  #direct import from GRASS, use readGDAL() for other formats
      colours <- topo.colors(1000) # if you want pretty topographic colours
+
 
      colours <- gray(c(rep(100,10),100:15)/100) # for greyscale colours
+
  paute <- rast.get6("paute_DEM")
      trellis.par.set(sp.theme(regions=list(col = colours)))
+
 
      scale = list("SpatialPolygonsRescale", layout.scale.bar(height=0.1), offset = c(690000,9735000), scale = 20000, fill=c("transparent","black"))
+
  # set some nice topographic colours, if you want dull greyscale colors, use:
      text1 = list("sp.text", c(690000,9740000), "0")
+
  # colours <- gray(c(rep(100,10),100:15)/100)
      text2 = list("sp.text", c(710000,9740000), "20 km")
+
 
      arrow = list("SpatialPolygonsRescale", layout.north.arrow(type=1), offset = c(698000,9720000), scale = 10000)
+
  colours <- topo.colors(1000)
      spplot(paute,draw = T, colorkey=list(space="right", height=0.5), cuts=127,
+
  trellis.par.set(sp.theme(regions=list(col = colours)))
              scales=list(draw=T), sp.layout=list(scale,text1,text2,arrow))
+
 
 +
  # define the objects necessary to draw a scalebar and north arrow:
 +
 
 +
  scale = list("SpatialPolygonsRescale", layout.scale.bar(height=0.1), offset = c(690000,9735000), scale = 20000, fill=c("transparent","black"))
 +
  text1 = list("sp.text", c(690000,9740000), "0")
 +
  text2 = list("sp.text", c(710000,9740000), "20 km")
 +
  arrow = list("SpatialPolygonsRescale", layout.north.arrow(type=1), offset = c(698000,9720000), scale = 10000)
 +
 
 +
  # and finally plot the map:
 +
 
 +
  spplot(paute_DEM,draw = T, colorkey=list(space="right", height=0.5), cuts=127,
 +
      scales=list(draw=T), sp.layout=list(scale,text1,text2,arrow))
  
 
Good to know:
 
Good to know:

Revision as of 18:12, 3 August 2008

Introduction

R can plot very nice maps. I decided to give up plotting maps in grass. The ps.map function is very powerful, but hard to remember for my not-so-frequent use. I use the grass interface but R imports very many raster and vector formats via the rgdal package.

Plotting GIS data is very nicely implemented in the sp package, part of the spatial data in R project, but getting your plot the way you want is quite challenging. Here is an example.

Required packages

spgrass6 reads data from the GRASS database, use rgdal for other formats. sp implements spplot, and lattice is used to define the colours (trells.par.set).

  library(sp)
  library(spgrass6)
  library(lattice)

Plotting a map

  #direct import from GRASS, use readGDAL() for other formats
  
  paute <- rast.get6("paute_DEM")
  
  # set some nice topographic colours, if you want dull greyscale colors, use:
  # colours <- gray(c(rep(100,10),100:15)/100)
  
  colours <- topo.colors(1000)
  trellis.par.set(sp.theme(regions=list(col = colours)))
  
  # define the objects necessary to draw a scalebar and north arrow:
  
  scale = list("SpatialPolygonsRescale", layout.scale.bar(height=0.1), offset = c(690000,9735000), scale = 20000, fill=c("transparent","black"))
  text1 = list("sp.text", c(690000,9740000), "0")
  text2 = list("sp.text", c(710000,9740000), "20 km")
  arrow = list("SpatialPolygonsRescale", layout.north.arrow(type=1), offset = c(698000,9720000), scale = 10000)
  
  # and finally plot the map:
  
  spplot(paute_DEM,draw = T, colorkey=list(space="right", height=0.5), cuts=127,
     scales=list(draw=T), sp.layout=list(scale,text1,text2,arrow))

Good to know:

  • cuts defines the number of colour intervals. The default is 15 and that's not really nice for continuous DEM images
  • the colorkey options are part of levelplot {lattice} and are quite customizable

Results

Map paute.png