Test a neighbours list for symmetry
Checks a neighbours list for symmetry/transitivity (if i is a neighbour of j,
then j is a neighbour of i). This holds for distance and contiguity based
neighbours, but not for k-nearest neighbours. The helper function
sym.attr.nb() calls is.symmetric.nb() to set the sym
attribute if needed, and make.sym.nb makes a non-symmetric list symmetric by adding neighbors. is.symmetric.glist checks a list of general weights corresponding to neighbours for symmetry for symmetric neighbours.
is.symmetric.nb(nb, verbose = NULL, force = FALSE) sym.attr.nb(nb) make.sym.nb(nb) old.make.sym.nb(nb) is.symmetric.glist(nb, glist)
nb |
an object of class |
verbose |
default NULL, use global option value; if TRUE prints non-matching pairs |
force |
do not respect a neighbours list |
glist |
list of general weights corresponding to neighbours |
TRUE if symmetric, FALSE if not; is.symmetric.glist returns a value with an attribute, "d", indicating for failed symmetry the largest failing value.
A new version of make.sym.nb by Bjarke Christensen is now included. The older version has been renamed old.make.sym.nb, and their comparison constitutes a nice demonstration of vectorising speedup using sapply and lapply rather than loops. When any no-neighbour observations are present, old.make.sym.nb is used.
Roger Bivand Roger.Bivand@nhh.no
columbus <- st_read(system.file("shapes/columbus.shp", package="spData")[1], quiet=TRUE)
col.gal.nb <- read.gal(system.file("weights/columbus.gal", package="spData")[1])
coords <- coordinates(as(columbus, "Spatial"))
ind <- row.names(as(columbus, "Spatial"))
print(is.symmetric.nb(col.gal.nb, verbose=TRUE, force=TRUE))
k4 <- knn2nb(knearneigh(coords, k=4), row.names=ind)
k4 <- sym.attr.nb(k4)
print(is.symmetric.nb(k4))
k4.sym <- make.sym.nb(k4)
print(is.symmetric.nb(k4.sym))Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.