Question

I have a datset, which is

dat<-matrix(seq(1:5290),c(115,46))

1. How can I convert dat to a raster file with GIS – related metadata such as map projection?

2. and then how to map the data to a us map with

   lon <- seq(-124.25,length=115,by=0.5)    
   lat <- seq(26.25,length=46,by=0.5)
   

3. what should I do if I want to calculate the mean of each state? or specific region?

thanks very much for your feedback.

D

Answer 1

You could try the following:

library(raster)
library(maps)

dat<-matrix(seq(1:5290),c(115,46))  

Create a raster with your desired extent

r <- raster(dat)
e <- extent(c(-124.5, -67.25, 26.25, 48.75))
extent(r) <- e

and plot it

plot(r)
map("usa", add=T)

take the mean over the whole raster

mean(getValues(r))

mean over a region

r.sub <- crop(r, c(-100, -54, 40, 45))
mean(getValues(r.sub))

---

Extract from polygons

You need a SpatialPolygonsDataFrame. Here I am creating a dummy one with readWKT() but you could also read a shapefile with readShapePoly() from the maptools package.

library(rgeos)
poly <- readWKT("POLYGON((-100 30, -100 31, -70 30, -88 29, -100 30))")
plot(r)
plot(poly, add=T)

And extract the mean:

dat <- extract(r, poly)
sapply(dat, mean)