I want to assess if the observations in my data are spatially randomly distributed over the sampling area (Sweden). I wanted to reproduce the example given in this answer: Spatial Autocorrelation Analysis (Global Moran's I) in R
Here is a small subset of my data, and the spatial polygon I used. Note that the coordinates are in SWEREF99 (ESPG: 3006)
## spatial polygon of Sweden
library(rworldmap)
library(sp)
worldmap <- getMap(resolution = "high")
sweden <- worldmap[which(worldmap$SOVEREIGNT == "Sweden"),]
plot(sweden)
sweden
## conversion to EPSG: 3006 (SWEREF99 TM) (https://spatialreference.org/ref/epsg/3006/)
crs.laea <- CRS("+proj=utm +zone=33 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs")
sweden_proj <- spTransform(sweden, crs.laea)
## Data subset
x <- c(669894, 669894, 669894, 671088, 671117, 671117, 671117, 670513, 670513, 670513, 669921, 669310, 669310, 669310, 669303, 629720, 630318, 630925, 630925, 630925)
y <- c(7116684, 7116684, 7116684, 7116706, 7114900, 7114900, 7114900, 7114896, 7114896, 7114896, 7114888, 7115473, 7115473, 7115473, 7116075, 7131172, 7131180, 7131190, 7131190, 7131190)
library(spatstat)
coords.ppp_1 <- ppp( x , y , xrange = c(280227, 911417) , yrange = c(6142436, 7605020) )
coords.ppp <- unique(coords.ppp_1)
### plot data and Sweden map for check
plot(coords.ppp_1)
plot(sweden_proj, add=T)
So far it seems ok. Then I convert the spatial polygon to an owin object, simulate random data for comparison, and do the analysis.
library(maptools)
sw <- as.owin.SpatialPolygons(sweden_proj)
# Generate completely spatially random point patterns to compare against the observed
n <- coords.ppp_1$n
ex <- expression(runifpoint( n , sw))
# Compute a simulation envelope using Gest, which estimates the nearest neighbour distance distribution function G(r)
set.seed(1)
res <- envelope( coords.ppp , Gest , nsim = 99, simulate = ex ,verbose = FALSE, savefuns = TRUE )
plot(res)
With the envelope() I get the following error message: "In envelopeEngine(X = X, fun = fun, simul = simrecipe, nsim = nsim, : Window containing simulated patterns is not a subset of data window" I suspect that there is a problem with the conversion between sp and owin, but I couldn't figure out what the issue really is.
Any advice?
