my.dcirc.measure: Weight border points for their local points

In ryamada22/Ronlyryamada:

Description

This function weights points with fraction of points closest to them among all border points

Usage

my.dcirc.measure(x, r1, r2 = 1)

Arguments

x	is a 0/1 matrix.
r1	is a positive real value of width to measure circumference.
r2	is a positive real value to construct border points

Examples

library(igraph)
n <- 2^5
x <- y <- 1:n
xy <- as.matrix(expand.grid(x,y))
xy.val <- rep(0,length(xy[,1]))
n.circle <- 30
ctr <- matrix(sample(20:40,replace=TRUE,n.circle*2),ncol=2)
r <- sample(1:20,replace=TRUE,n.circle)

for(i in 1:n.circle){
	tmp.x <- xy[,1] - ctr[i,1]
	tmp.y <- xy[,2] - ctr[i,2]
	s <- which(tmp.x^2+tmp.y^2 < r[i]^2)
	xy.val[s] <- 1
}
xy.mat <- matrix(xy.val,ncol=n)
yyy <- my.dcirc.measure(xy.mat,10,1)
bdr <- yyy$border
bdr <- rbind(bdr,bdr[2,])
diff.bdr <- apply(bdr,2,diff)
#diff.bdr <- diff.bdr[-length(diff.bdr[,1]),]
diff.bdr.1 <- diff.bdr[-length(diff.bdr[,1]),]
diff.bdr.2 <- diff.bdr[-1,]
diff.bdr.1.cplx <- diff.bdr.1[,1] + diff.bdr.1[,2]*1i
diff.bdr.2.cplx <- diff.bdr.2[,1] + diff.bdr.2[,2]*1i
direc <- diff.bdr.2.cplx/diff.bdr.1.cplx
arg.direc <- Arg(direc)
cumsum.arg <- cumsum(arg.direc)
plot(cumsum.arg,type="l")
plot(yyy[[1]])
plot(yyy[[3]])
col <- yyy[[1]]
col <- (max(col)-col)/(max(col)-min(col))
col <- 0.2 + 0.6*col
plot(yyy[[3]],pch=20,col=gray(col))
loc <- cumsum(yyy[[1]])
plot(loc)
plot(loc,cumsum.arg)

ryamada22/Ronlyryamada documentation built on May 28, 2019, 10:43 a.m.