Why as.ppp.sf attach only the first column as mark?

[agila5]

Dear all, I was wondering why the following code

sf/R/spatstat.R

Lines 141 to 154 in 6f247a5

	as.ppp.sf = function(X, ...) {
	check_spatstat("spatstat.geom", X)
	pp = spatstat.geom::as.ppp(st_geometry(X), ...)
	if (st_dimension(X[1,]) == 2)
	X = X[-1,]
	st_geometry(X) = NULL # remove geometry column
	if (ncol(X) > 1)
	warning("only first attribute column is used for marks")
	if (ncol(X) == 0)
	pp
	else
	spatstat.geom::setmarks(pp, X[1])
	}

forces the conversion sf -> ppp to keep only the first column of mark. What are the downside of keeping all the columns (if any) of X? AFAICT, the argument value in spatstat.geom::setmarks accepts a vector but also a data.frame object. For example:

library(sf)
library(spatstat)

X <- st_read(system.file("csv/pt.csv", package="sf"))
pp <- spatstat.geom::as.ppp(st_geometry(X)) 
st_geometry(X) = NULL
spatstat.geom::setmarks(pp, X) 
#> Marked planar point pattern: 3 points
#> Mark variables: Id, Int64, Int32, WKT 
#> window: rectangle = [0, 44] x [0, 55] units

^{Created on 2024-10-03 with reprex v2.0.2}

EDIT: Maybe this wasn't always the case, but I cannot find any proper reference in the NEWS file of spatstat

[edzer]

Yes, good point, I vaguely recollect that marks were univariate, but may have overlooked this. PR welcome!