raster_example.r

library(raster)

#1 open a raster file as memory-mapped on-disk object
bio1 <- raster("bio1.bil") # the 1st bioclimatic variable from the Chelsa dataset

#2 get the projection
proj4string(bio1)

#3 plot it
plot(bio1)

#4 plot part of it
plot(bio1, xlim = c(-80,-35), ylim = c(-60, 15))

#5 extract part of the raster based on a rectangular window
samerica <- crop(bio1, extent(-80, -35, -60, 15))
plot(samerica)

#6 extract part of it based on a shapefile polygon
library(rgdal)
cnt <- readOGR("~/Documents/Courses/R data analysis/Data/Countries", "CNTRY92") # get all countries as polygons
braz <- subset(cnt, NAME == "Brazil") # extract brazil
plot(braz, add = T, border = "red")
brarast <- mask(samerica, braz)

#7 plot that newly extracted raster with some spatial points
plot(brarast)
ex <- extent(braz)
sp <- SpatialPoints(cbind(runif(50, ex[1], ex[2]), runif(50, ex[3], ex[4])))
plot(sp, add = T, col = "red")

#8 extract the raster's value under these points
values <- extract(brarast, sp)

#9 aggregate the raster to a coarser resolution
bra_coarse <- aggregate(brarast, 9, mean)

#10 use the raster as a standard matrix
brarast[100:120, 150:180] * 2 # doesn't quite work in R as this is now a Vector

#11 apply image windowing functions to the raster
window_mean <- focal(bra_coarse, matrix(rep(1,9), nrow = 3), mean)