Shapefile to raster conversion in R?

9
votes

I have a shapefile downloaded from the worldwildlife.org for the terrestrial ecoregions of the world. The file can be loaded here: http://worldwildlife.org/publications/terrestrial-ecoregions-of-the-world.

It comes as a standard shape file and I would like to do two things with it. First: take the shapefile from my local directory and clip it to an extent of eastern North America (ext= extent (-95, -50, 24, 63)) 

# Read shapefile using package "maptools"
eco_shp <- readShapeLines("F:/01_2013/Ecoregions/Global/wwf_terr_ecos.shp", 
                          proj4string=CRS("+proj=utm +zone=33 +datum=WGS84")) 


# Set the desired extent for the final raster using package "raster" 
ext <- extent(-95, -50, 24, 63)

I am sure I have to use the rasterize function in the package "raster" but I am still not able to get it work correctly. I would appreciate any suggestions on how to do this.

1
rrastershapefiledata-conversionrasterizing
Do you need it to be rasterised? Would cliping the shapefile by a polygon suffice? – mnel
Mnel: I do need it to be rasterized into an .asc file to match up with my other environmental layers. – I Del Toro
The first step of clipping by a shapfile would be helpful, then that new object could be rasterized and might be less intensive . – I Del Toro
why not just overlay the polygon onto the SpatialGridDataFrame you wish to match it to, (use overlay) – mnel
Mnel: My model requires that my input dataframes all be in an asc format and have the same spatial extent. If I only wanted to visualize it the "overlay" would work. :) – I Del Toro

1 Answers

13
votes

You are right to think that you should be using raster (rather than the sp raster Spatial classes) for spatial raster data. You should also use rgdal (rather than maptools) for reading, writing, and otherwise manipulating spatial vector data.

This should get you started:

library(rgdal)
library(raster)

## Read in the ecoregion shapefile (located in R's current working directory)
teow <- readOGR(dsn = "official_teow/official", layer = "wwf_terr_ecos")

## Set up a raster "template" to use in rasterize()
ext <-  extent (-95, -50, 24, 63)
xy <- abs(apply(as.matrix(bbox(ext)), 1, diff))
n <- 5
r <- raster(ext, ncol=xy[1]*n, nrow=xy[2]*n)

## Rasterize the shapefile
rr <-rasterize(teow, r)

## A couple of outputs
writeRaster(rr, "teow.asc")
plot(rr)

enter image description here