Append spatstat intensity outputs across tesselated surface (derived from shapefile) back to SpatialPolygonsDataFrame for plotting in R

Sean Basquill 0
Viewed 20

Anyone have advice for retaining (or appending) attributes from a Spatial Polygons Data Frame after conversion to an owin object for pattern analysis in Spatstat?

Using spatstat to calculate point intensities across a tesselated surface. No issue with this except I lose the polygon labels when I convert the Spatial Polygons Data Frame to an owin object (and subsequently a tess object). See sample code. Thanks!


SF <- readOGR(dsn="D:/",layer="sf")
class(SF) # "SpatialPolygonsDataFrame"

# Convert SpatialPolygonsDataFrame to OWIN object for use in spatstat package

y <- as(SF, "SpatialPolygons")
p <- slot(y, "polygons")
v <- lapply(p, function(z) { SpatialPolygons(list(z)) })
winlist <- lapply(v, as.owin)

# establish window needed for creating tesseslated surface
SF2 <- as(SF, "owin")

# create tesselates surface
e <-tess(tiles = winlist, window = SF2)

## compute counts and density values, across tessellated area, of point data 

# B - point pattern object created from *.csv of point locations

QB   <- quadratcount(B, tess = e)  # tally counts of 'biotic pseudo-variables' by tess tiles
Qi.B <- intensity(QB)  # Compute intensity

# Plot results
plot(intensity(QB, image=TRUE), las=1, main=NULL) # plot tess units using intensity for legend
plot(B, pch=20, cex=0.5, col="white", add=TRUE)```
spatstat geospatial r
1 Answers

You will need to manually extract the polygon labels from the SpatialPolygonsDataFrame object SF as a character vector, and then assign them to the names of the tiles in the tessellation object e.

To extract the polygon labels you can probably type nama <- rownames(SF) but it depends which labels you require.

There are two ways to assign the tile names:

  1. Assign names to the list of tiles before creating the tessellation. In the command e <- tess(tiles=winlist, window=SF2), if names(winlist) is not NULL then the names will be retained as the names of the tiles. So you could first do names(winlist) <- nama and then e <- tess(tiles=winlist, window=SF2).

  2. Assign tile names after creating the tessellation using "tilenames<-",as in tilenames(e) <- nama where nama is a character vector.

In either case, once the tiles of e have names, the results of calculations such as Qi.B will also have these names.

Adrian Baddeley0
Related