Combining two point pattern objects in spatstat creating a combined mark

Question

Suppose there are two point patterns in spatstat. I understand we can superimpose these two point patterns to get a single point pattern. Now if there are common coordinates for these patterns but with different marks, then the points will be duplicated when superimposed. Is there a way in spatstat to get a unique set of points while creating a common mark for the coordinates that coincide?

Answer 1

I’m not aware of a built-in solution to do this, so you have to do a bit of manual work as detailed below.

Load package and make example data with overlapping points:

library(spatstat)
X1 <- cells[1:22]
marks(X1) <- factor("a")
X2 <- cells[20:42]
marks(X2) <- factor("b")
plot(superimpose(X1, X2), main = "")

For each point in X1 find the nearest point in X2:

nn <- nncross(X1, X2)
tail(nn)
#>         dist which
#> 17 0.1386110     4
#> 18 0.1802776     5
#> 19 0.1069766     5
#> 20 0.0000000     1
#> 21 0.0000000     2
#> 22 0.0000000     3
id1 <- which(nn$dist==0) ## Tests EXACT equality. Consider small tolerance.
id2 <- nn$which[id1]

Add extra mark level to X1 and assign it to points with duplicates in X2:

levels(marks(X1)) <- c("a", "c")
marks(X1)[20:22] <- factor("c")
X <- superimpose(X1, X2[-id2])
plot(X, main = "")