app: Apply a function to the cells of a SpatRaster

Description Usage Arguments Details Value See Also Examples

Description

Apply a function to the values of each cell of a SpatRaster. Similar to apply – think of each layer in a SpatRaster as a column (or row) in a matrix.

This is generally used to summarize the values of multiple layers into one layer; but this is not required.

app calls function fun with the raster data as first argument. Depending on the function supplied, the raster data is represented as either a matrix in which each layer is a column, or a vector representing a cell. The function should return a vector or matrix that is divisible by ncell(x). Thus, both "sum" and "rowSums" can be used, but "colSums" cannot be used.

You can also apply a function fun across datasets by layer of a SpatRasterDataset. In that case, summarization is across SpatRasters, not across layers.

Usage

1
2
3
4
5
## S4 method for signature 'SpatRaster'
app(x, fun, ..., cores=1, filename="", overwrite=FALSE, wopt=list())

## S4 method for signature 'SpatRasterDataset'
app(x, fun, ..., cores=1, filename="", overwrite=FALSE, wopt=list())

Arguments

x

SpatRaster or SpatRasterDataset

fun

a function that operates on a vector or matrix. This can be a function that is defined in base-R or in a package, or a function you write yourself (see examples). Functions that return complex output (e.g. a list) may need to be wrapped in your own function to simplify the output to a vector or matrix. The following functions have been re-implemented in C++ for speed: "sum", "mean", "median", "modal", "which", "which.min", "which.max", "min", "max", "prod", "any", "all", "sd", "std", "first". To use the base-R function for say, "min", you could use something like fun=function(i) min(i) or the equivalent fun = \(i) min(i)

...

additional arguments for fun. These are typically numerical constants. They should *never* be another SpatRaster

cores

positive integer. If cores > 1, a 'parallel' package cluster with that many cores is created and used. Ignored for functions that are implemented by terra in C++ (see under fun)

filename

character. Output filename

overwrite

logical. If TRUE, filename is overwritten

wopt

list with named options for writing files as in writeRaster

Details

To speed things up, parallelization is supported, but this is often not helpful, and it may actually be slower. There is only a speed gain if you have many cores (> 8) and/or a very complex (slow) function fun. If you write fun yourself, consider supplying a cppFunction made with the Rcpp package instead (or go have a cup of tea while the computer works for you).

Value

SpatRaster

See Also

lapp, tapp, math

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
r <- rast(ncols=10, nrows=10)
values(r) <- 1:ncell(r)
x <- c(r, sqrt(r), r+50)
s <- app(x, fun=sum)
s
# for a few generic functions like 
# "sum", "mean", and "max" you can also do
sum(x)

## SpatRasterDataset
sd <- sds(x, x*2, x/3)
a <- app(sd, max)
a
# same as 
max(x, x*2, x/3)

## also works for a single layer
f <- function(i) (i+1) * 2 * i + sqrt(i)
s <- app(r, f)
# same as above, but that is not memory-safe
# and has no filename argument 
s <- f(r)

## Not run: 
#### multiple cores 
test0 <- app(x, sqrt) 
test1 <- app(x, sqrt, cores=2)

testfun <- function(i) { 2 * sqrt(i) }
test2 <- app(x, fun=testfun, cores =2)

## this fails because testfun is not exported to the nodes
# test3 <- app(x, fun=function(i) testfun(i), cores=2)
## to export it, add it as argument to fun
test3 <- app(x, fun=function(i, ff) ff(i), cores =3, ff=testfun)

## End(Not run)

terra documentation built on Aug. 20, 2021, 9:07 a.m.