Segue o script e os dados em anexo.

event=Caso22
x <- c(0,6,6,0)
y <- c(0,0,8,8)
poly=cbind(x,y)
s=seq(1,5,1)
tm=seq(1,5,1)

anaksenv<-function(event, poly, s, tm, nsim = 10){
#
# This function implements monte-carlo simulations
# to create and plot simulation envelopes to test
# against the hypothesis of complete temporal randomness (CTR)
# of an observed spatial-temporal point pattern.
# It uses the spatial k-function and the cramer-vonMises measure of distance.
#
# OBS: You need splancs loaded
#
# Where:
#   - event: A 3xn matrix that contains the coordinates and times of the events
#   - poly:    A polygon enclosing the points
#   - s:    A vector of spatial distances for the analysis.
#   - tm:    A vector of times for the analysis
#   - nsim: Number of monte-carlo simulations
#
#
    library(splancs)
#
#    Starting values
#
    x<-event[,1]
    y<-event[,2]
    t<-event[,3]
    maxx<-max(x)
    maxy<-max(y)
    minx<-min(x)
    miny<-min(y)
    lent<-length(x)
    pontos<-cbind(x,y)
    tl<-cbind(min(t),max(t))
#
#     Get the K-function for space and time of the observed point pattern
#
      ksto <- stkhat(pts=pontos, times=t, poly=poly, tlimits=tl, s=s, tm=tm)
    ko <- ksto$ks
#
# Doing Monte-Carlo simulation under the CSR
#
      hold<-matrix(0,nsim,length(s))
         for(i in (1:nsim)) {
        tt<-floor(runif(lent,min(t),max(t)))
        xx<-floor(runif(lent,minx,maxx))
        yy<-floor(runif(lent,miny,maxy))
        pont<-cbind(xx,yy)
        tll<-cbind(min(tt),max(tt))
        kstmc<-stkhat(pts=pont, times=tt, poly=poly, tlimits=tll, s=s, tm=tm)
            kko<-kstmc$ks
    hold[i,]<-kko
}
#
        up<- apply(hold,2,max)
        dow<- apply(hold,2,min)
#
# Creating plot
#
    par(pty = "s")
        plot(s, up, type = "n", xlab = "Distância", ylab = "Função K espacial")
    lines(s, ko, col="blue")
        lines(s, up, lty=2, col="red")
        lines(s, dow,lty=2,col="red")
}

Desde já muito obrigada!!