I have raster of the following features:
library(raster)
library(rgeos)
test <- raster(nrow=225, ncols=478, xmn=-15.8, xmx=32, ymn=-9.4, ymx=13.1)
I want to mask in this raster the cells that are within a given distance of a point.
I create the spatial points as followed:
p2=readWKT("POINT(31.55 -1.05)")
Then I create a spatial polygon object by adding a 0.5 buffer:
p2_Buffered <- gBuffer(p2, width = 0.5)
mask(test, mask=p2_Buffered,inverse=T)
When I mask my raster given this spatial object, I have the following error message:
Error in .polygonsToRaster(x, y, field = field, fun = fun, background
= background, : number of items to replace is not a multiple of replacement length
I do not understand because this is script I have been running many many times with different point and different buffer width without any problem.
What is strange is that when I change the width of the buffer, it works fine:
p2_Buffered <- gBuffer(p2, width = 0.4)
mask(test, mask=p2_Buffered,inverse=T)
This is also true for a different focal point:
p2=readWKT("POINT(32.55 -1)")
p2_Buffered <- gBuffer(p2, width = 0.5)
mask(test, mask=p2_Buffered,inverse=T)
I would like to identify the specific problem I have for that point because this is a script I should run in a routine (I have been doing it without any problem so far).
Thanks a lot
This is indeed a bug with polygons that go over the edge of a raster. It has been fixed in version 2.3-40 (now on CRAN), so it should go away if you update the raster package.
Here is a workaround (removing the part of the polygon that goes over the edge).
library(raster)
library(rgeos)
r <- raster(nrow=225, ncols=478, xmn=-15.8, xmx=32, ymn=-9.4, ymx=13.1)
e <- as(extent(r), 'SpatialPolygons')
p <- readWKT("POINT(31.55 -1.05)")
pb <- gBuffer(p, width = 0.5)
pbe <- intersect(pb, e)
values(r)
x <- mask(r, mask=pbe, inverse=TRUE)
You usually need to set some values to the raster layer. For a mask layer its always best to set values to 1.
library(raster)
library(rgeos)
# make sample raster
test <- raster(nrow=225, ncols=478, xmn=-15.8, xmx=32, ymn=-9.4, ymx=13.1)
# set values of raster for mask
test <- setValues(test, 1)
# make point buffer
p2=readWKT("POINT(15 5)")
p2_Buffered <- gBuffer(p2, width = 1.5)
# name projection of buffer (assume its the same as raster)
projection(p2_Buffered) <- projection(test)
# visual check
plot(test); plot(p2_Buffered, add=T)
If you want to trim down your raster layer to the just the single polygon then try this workflow.
step1 <- crop(test, p2_Buffered) # crop to same extent
step2 <- rasterize(p2_Buffered, step1) # rasterize polygon
final <- step1*step2 # make your final product
plot(final)
If you just want to poke a hole in your raster layer then use the mask function
# rasterize your polygon
p2_Buffered <- rasterize(p2_Buffered, test, fun='sum')
# now mask it
my_mask <- mask(test, mask=p2_Buffered,inverse=T) # try changing the inverse argument
plot(my_mask)
Related
I'm trying to crop a large multipolygon shapefile by a single, smaller polygon. It works using st_intersection, however this takes a very long time, so I'm instead trying to convert the multipolygon to a raster, and crop that raster by the smaller polygon.
## packages - sorry if I've missed any!
library(raster)
library(rgdal)
library(fasterize)
library(sf)
## load files
shp1 <- st_read("pathtoshp", crs = 27700) # a large multipolygon shapefile to crop
### image below created using ggplot- ignore the black boundaries!
shp2 <- st_read("pathtoshp", crs = 27700) # a single, smaller polygon shapefile, to crop shp1 by
plot(shp2)
## convert to raster (faster than st_intersection)
projection1 <- CRS('+init=EPSG:27700')
rst_template <- raster(ncols = 1000, nrows = 1000,
crs = projection1,
ext = extent(shp1))
rst_shp1 <- fasterize(shp1, rst_template)
plot(rst_shp1)
rst_shp2 <- crop(rst_shp1, shp2)
plot(rst_shp2)
When I plot shp2, the upper boundary is flat, rather than fitting the true boundary of the shp2 polygon.
Any help would be greatly appreciated!
Maybe try raster::mask() instead of crop(). crop() uses the second argument as an extent with which to crop a raster; i.e. it's taking the bounding box (extent) of your second argument and cropping that entire rectangle from your raster.
Something important to understand about raster objects is that they are all rectangular. The white space you see surrounding your shape are just NA values.
raster::mask() will take your original raster, and a spatial object (raster, sf, etc.) and replace all values in your raster which don't overlap with your spatial object to NA (by default, you can supply other replacement values). Though I will say, mask() will likely also take awhile to run, so you may be better off just sticking with sf objects.
I would suggest moving to the "terra" package (faster and easier to use than "raster").
Here is an example.
library(terra)
r <- rast(system.file("ex/elev.tif", package="terra"))
v <- vect(system.file("ex/lux.shp", package="terra"))[4]
x <- crop(r, v)
plot(x); lines(v)
As edixon1 points out, a raster is always rectangular. If you want to set cells outside of the polygon to NA, you can do
x <- crop(r, v, mask=TRUE)
plot(x); lines(v)
In this example it makes no sense, but you could first rasterize
x <- crop(r, v)
y <- rasterize(v, x)
m <- mask(x, y)
plot(m); lines(v)
I am not sure if this answers your question. But if it does not, then please edit your question to make it reproducible, for example using the example data above.
I am working with a very large raster. I want to increase values in each pixel randomly by 0.3-0.5 of its original value. What kind of loop should I apply to achieve it elegantly?
Example raster built below. My raster is a .tif, and I would prefer not convert it to matrix first, unless it is the best solution?
library(raster)
## Create a matrix with random data & use image()
xy <- matrix(rnorm(400),20,20)
image(xy)
# Turn the matrix into a raster
rast <- raster(xy)
# Give it lat/lon coords for 36-37°E, 3-2°S
extent(rast) <- c(36,37,-3,-2)
# ... and assign a projection
projection(rast) <- CRS("+proj=longlat +datum=WGS84")
plot(rast)
No loops are necessary. You can access the underlying pixel data directly and simply add a set of random numbers to it:
rast2 <- rast # a copy of the existing raster
random_nums <- runif(length(rast2), min = 0.3, max = 0.5) # a set of random numbers the size of the image
rast2#data#values <- rast2#data#values * random_nums # multiply the pixel data by the random values
What I´m looking for is a way to rasterize (or fasterize) geometries within each tile of a certain extent step by step and join the parts of the rasterized geometries to entire raster objects.
## create sample data
# create a frame
library(sf)
ob = st_sf(st_sfc(st_polygon(list(rbind(c(0,0), c(0,9), c(6,9), c(6,0), c(0,0)))))
# create tiles
library(GSIF)
tl <- getSpatialTiles(as(ob, 'Spatial'), block.x=3, overlap.percent=2)
plot(tl)
# create sample polygons
g <- st_sfc(st_point(c(1,2)), st_point(c(5,6)), st_point(c(2,4)), st_point(c(3,3)), st_point(c(3,4)), st_point(c(4,5)))
g.b <- st_buffer(g,0.6)
p <- st_sf(value = ceiling(10*runif(6)),
geometry = st_sfc(g.b))
plot(p, add=TRUE)
Cropping the polygons (or parts of polygons) that are within each tile works fine , i guess. My actual goal is to process the data tile by tile. It looks like the following loop does this and also joins the parts of the polygons to entire polygons. Well, the output (cr) is the same as the input (p)... I thought that this step might be necessary in order to rasterize the resulting (parts of) polygons in the next step. i didn´t expect that the polygons would be joined. I was trying to build a loop based on the intermediate result, therefore following lines are add:
# crop polygons (or parts) for each tile
result <- p
for(i in 1:length(tl)) {cr <- rbind(result, st_crop(p, tl[i]))}
cr <- cr[-7,]
library(scales)
plot(st_geometry(cr), col=alpha("white", 0.5), add=TRUE)
However, I struggle rasterizing the (intermediate) results.
Try this:
myF <- function(i){return(st_crop(p, tl[i]))}
cr <- do.call(rbind, lapply(1:length(tl), myF))
I need to calculate the magnitude-per-unit area of polylines that fall within a radius around each cell. Essentially I need to calculate a km/km2 road density within a 500m pixel search radius. ArcMap has a quick and easy tool that handles this, but I need a pure R solution.
Here is a link on how line density works: http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-analyst-toolbox/how-line-density-works.htm
And this is how to use it in a python (arcpy) script: http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-analyst-toolbox/line-density.htm
I currently execute a backwards approach using raster::focal function, calculating a density of burned in road features. I then convert the km2/km2 output to km/km2.
#Import libraries
library(raster)
library(rgdal)
library(gdalUtils)
#Read-in an already created raster mask (cells are all set to 0)
mask <- raster("x://path to raster mask...")
#Make a copy of the mask to burn features in, keeping the original untouched
roads_mask <- file.copy(mask, "x://output path ...//roads.tif")
#Read-in road features (shapefile format)
roads_sldf <- readOGR("x://path to shapefile" , "roads")
#Rasterize spatial lines data frame ie. burn road features into mask
#Where road features get a value of 1, mask extent gets a value of 0
roads_raster <- gdalUtils::gdal_rasterize(src_datasource = roads_sldf,
dst_filename = "x://output path ...//roads.tif", b = 1,
burn = 1, l = "roads", output_Raster = TRUE)
#Run a 1km circular radius density function (be mindful of edge effects)
weight <- raster::focalWeight(roads_raster,1000,type = "circle")
1km_rdDensity <- raster::focal(roads_raster, weight, fun=sum, filename = '',
na.rm=TRUE, pad=TRUE, NAonly=FALSE, overwrite=TRUE)
#Convert km2/km2 road density to km/km2
#Set up the moving window
weight <- raster::focalWeight(roads_raster,1000,type = "circle")
#Count how many records in each column of the moving window are > 0
columnCount <- apply(weight,2,function(x) sum(x > 0))
#Get the sum of the column count
number_of_cells <- sum(columnCount)
#multiply km2/km2 density by number of cells in the moving window
step1 <- roads_raster * number_of_cells
#Rescale step1 output with respect to cell size(30m) and radius of a circle
final_rdDensity <- (step1*0.03)/3.14159265
#Write out final km/km2 road density raster
writeRaster(final_rdDensity,"X://path to output...", datatype = 'FLT4S', overwrite = TRUE)
After some more research I think I may be able to use a kernel function, however I don't want to apply the smoothing algorithm... As well the output is an 'im' object which I would need to write to as a 'tif'
#Import libraries
library(spatstat)
library(rgdal)
#Read-in road features (shapefile format)
roads_sldf <- readOGR("x://path to shapefile" , "roads")
#Convert roads spatial lines data frame to psp object
psp_roads <- as.psp(roads_sldf)
#Apply kernel density, however this is where I am unsure of the arguments
road_density <- spatstat::density.psp(psp_roads, sigma = 0.01, eps = 500)
Cheers.
See this question https://gis.stackexchange.com/questions/138861/calculating-road-density-in-r-using-kernel-density
Tried to mark as a duplicate but doesn't work because the other Q is on gis stack exchange
Short answer is use spatstat.geom::pixellate()
I also needed spatstat.geom::as.psp(sf::st_geometry(x)) to convert an sf lines object to the correct format and maptools::as.im.RasterLayer(r) to convert a raster. I was able to convert the result to RasterLayer with raster::raster(pix_res)
Perhaps you can use terra::rasterizeGeom which is available in the development version that you can install with install.packages('terra', repos='https://rspatial.r-universe.dev')
Example data
library(terra)
f <- system.file("ex/lux.shp", package="terra")
v <- vect(f) |> as.lines()
r <- rast(v, res=.1)
Solution
x <- rasterizeGeom(v, r, fun="length", "km")
And then use focal sum, but you would not have a perfect circle.
What you could do instead, if your dataset is not too large, is create a circle for each grid cell and use intersect. Something like this:
p <- xyFromCell(r, 1:ncell(r)) |> vect(crs="+proj=longlat")
p$id <- 1:ncell(r)
b <- buffer(p, 10000)
values(v) <- NULL
i <- intersect(v, b)
x <- aggregate(perim(i), list(id=i$id), sum)
r[x$id] <- x[,2]
I would like to sample a big raster by creating In small raster 100x100 cells.
I don't know how to do that so any ideas are welcome
My actual lead :
library(raster)
library(spatstat)
library(polyCub)
r <- raster(ncol=1000,nrow=1000) # create empty raster
r[] <- 1:(1000*1000) # Raster for testing
e <- extent(r) # get extend
# coerce to a SpatialPolygons object
p <- as(e, 'SpatialPolygons')
nc <- as.owin.SpatialPolygons(p) #polyCub
pts <- rpoint(50, win = nc)
plot(pts)
Now I need to generate 100x100 cell square around my 50 points and I would like to crop r using those square and stack each small raster individually ...
The answer by #adrian-baddeley basically has the ingredients to do what
you want. If you simply want a list of small im objects that contain
the 100x100 box you simply subset im objects by owin objects to
extract the relevant region. Here is an example (with fewer points to
avoid overplotting)
library(raster)
library(spatstat)
library(maptools)
r <- raster(ncol=1000,nrow=1000) # create empty raster
r[] <- 1:(1000*1000) # Raster for testing
e <- extent(r) # get extend
# coerce to a SpatialPolygons object
p <- as(e, 'SpatialPolygons')
nc <- as.owin.SpatialPolygons(p)
set.seed(42)
pts <- rpoint(7, win = nc)
rim <- as.im.RasterLayer(r)
Box <- owin(c(-50,50) * rim$xstep, c(-50,50) * rim$ystep)
The following is a list of im objects of size 100x100
imlist <- solapply(seq_len(npoints(pts)),
function(i) rim[shift(Box, pts[i])])
Here is a plot of the im objects in the region and the points on top
plot(pts)
for(i in imlist) plot(i, add = TRUE)
plot(pts, pch = 19, add = TRUE)
You can convert to a list of raster layers with
rasterList <- lapply(imlist, as, Class = "RasterLayer")
PS: The following is a list of im objects of the original size with
NA outside the 100x100 box if you need that format instead
imlist <- solapply(seq_len(npoints(pts)),
function(i) rim[shift(Box, pts[i]), drop = FALSE])
If you want to use spatstat then you need to convert the raster object r into an object of class im supported by spatstat. You can do this conversion in the maptools package. Call this image object rim. Then you can do as follows
Box <- owin(c(-50,50) * rim$xstep, c(-50,50) * rim$ystep)
BoxesUnion <- MinkowskiSum(pts, Box)
W <- intersect.owin(as.mask(rim), BoxesUnion)
This would give you the subset of the raster that is covered by the squares.
If you want to keep the squares separate, do something like
M <- as.mask(rim)
BoxList <- solapply(seq_len(npoints(pts)),
function(i) intersect.owin(M, shift(Box, pts[i])))
Then BoxList is a list of the individual sub-rasters.