Create a Point Matching
Creates an object of class "pppmatching" representing
a matching of two planar point patterns (objects of class "ppp").
pppmatching(X, Y, am, type = NULL, cutoff = NULL, q = NULL,
mdist = NULL)X,Y |
Two point patterns (objects of class |
am |
An |
type |
A character string giving the type of the matching.
One of |
cutoff, q |
Numerical values specifying the cutoff value > 0 for interpoint distances and
the order q in [0,Inf] of the average that is applied to them.
|
mdist |
Numerical value for the distance to be associated with the matching. |
The argument am is interpreted as a "generalized adjacency matrix":
if the [i,j]-th entry is positive, then the i-th point
of X and the j-th point of Y are matched and the
value of the entry gives the corresponding weight of the match. For
an unweighted matching all the weights should be set to 1.
The remaining arguments are optional and allow to save
additional information about the matching. See the help files for
pppdist and matchingdist for details on
the meaning of these parameters.
Dominic Schuhmacher dominic.schuhmacher@stat.unibe.ch http://dominic.schuhmacher.name/
# a random unweighted complete matching
X <- runifrect(10)
Y <- runifrect(10)
am <- r2dtable(1, rep(1,10), rep(1,10))[[1]]
# generates a random permutation matrix
m <- pppmatching(X, Y, am)
summary(m)
m$matrix
plot(m)
# a random weighted complete matching
X <- runifrect(7)
Y <- runifrect(7)
am <- r2dtable(1, rep(10,7), rep(10,7))[[1]]/10
# generates a random doubly stochastic matrix
m2 <- pppmatching(X, Y, am)
summary(m2)
m2$matrix
plot(m2)
m3 <- pppmatching(X, Y, am, "ace")
m4 <- pppmatching(X, Y, am, "mat")Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.