How to make a dataset similar to that of murchison data in spatstat for ppm and AUc analysis - spatstat

Question

I have four variables: a point process pattern of species
occurrence, rivers, ponds polygons and land image data. I would like
to make a dataset similar to that of Murchison dataset using these
shape layers but I have failed to manoeuvre.
I need to make a data frame from these polygon shape layers of
rivers, ponds and land cover images together with the point pattern
data of species occurrences I tried using a hyper frame but I am
unable to use a distance function from the river or the ponds.
rivers <- readShapespatial("river.shp") ponds <-
readShapeSpatial(pond.shp") fro <- read.table("fro.txt",
header=TRUE) image <- raster("image.tif")
I would like to combine
these four files as a single spatstat object like that of Murchison
data which comes with spatstat package. if I can put them in a frame
then ponds, land cover, rivers are covariates.
I have used analyst function but return errors that they can not be
used as covariates, fore example x is a list can not be used as
covariates particularly for ponds and rivers when I call the dist
function.

Answer 1

Why do you need a hyperframe? You refer to murchison data and that is not
a hyperframe. It simply a standard R list (with extendend classes
listof, anylist and solist for better printing and plotting in
spatstat, but the actual data structure is just a plain list).
To recreate the murchison data:
library(spatstat)
P <- murchison$gold # Points
L <- murchison$faults # Lines
W <- murchison$greenstone # "Windows
mur <- solist(points = P, lines = L, windows = W)
mur
#> List of spatial objects
#>
#> points:
#> Planar point pattern: 255 points
#> window: rectangle = [352782.9, 682589.6] x [6699742, 7101484] metres
#>
#> lines:
#> planar line segment pattern: 3252 line segments
#> window: rectangle = [352782.9, 682589.6] x [6699742, 7101484] metres
#>
#> windows:
#> window: polygonal boundary
#> enclosing rectangle: [352782.9, 681699.6] x [6706467, 7100804] metres
To use the data in a model they don’t have to be collected in a single list,
but it may be convenient. The following two models are identical:
(mod1 <- ppm(P ~ W))
#> Nonstationary Poisson process
#>
#> Log intensity: ~W
#>
#> Fitted trend coefficients:
#> (Intercept) WTRUE
#> -21.918688 3.980409
#>
#> Estimate S.E. CI95.lo CI95.hi Ztest Zval
#> (Intercept) -21.918688 0.1666667 -22.24535 -21.592028 *** -131.51213
#> WTRUE 3.980409 0.1798443 3.62792 4.332897 *** 22.13252
(mod2 <- ppm(points ~ windows, data = mur))
#> Nonstationary Poisson process
#>
#> Log intensity: ~windows
#>
#> Fitted trend coefficients:
#> (Intercept) windowsTRUE
#> -21.918688 3.980409
#>
#> Estimate S.E. CI95.lo CI95.hi Ztest Zval
#> (Intercept) -21.918688 0.1666667 -22.24535 -21.592028 *** -131.51213
#> windowsTRUE 3.980409 0.1798443 3.62792 4.332897 *** 22.13252
If you insist on a hyperframe you should have a column for each measured
variable, but these are primarily used for when you have several replications
of an experiment, and is not of much use here. The function call is simply:
murhyp <- hyperframe(points = P, lines = L, windows = W)