Frederico,

This is not exactly what you're after, but perhaps it will help.  In this 
example I fit a cluster analysis to the data, then I "cut the tree" at a 
height of 3 (you would do this with your data at a height of 40).  It's 
not a perfect solution, but it might be good enough, depending on the 
spatial distribution of your points.

# example data frame with x and y (on the same scale)
df <- data.frame(x = rnorm(100), y = rnorm(100))

# cluster analysis
tree <- hclust(dist(df[, c("x", "y")], method="euclidean"), 
method="complete")

# define groups as those that are at least 3 units apart
df$group <- cutree(tree, h=3)

# plot the data, using color and symbol to identify group membership
eqscplot(df$x, df$y, col=df$group, pch=df$group)

Jean


"Frederico Mestre" <mestre.freder...@gmail.com> wrote on 07/30/2012 
07:29:00 AM:
> 
> Hello:
> 
> What I want to do is quite simple, but I can't find a way.
> 
> I have a data frame with several points (x and y coords). I want to add
> another column with cluster membership. For example aggregate all the 
points
> that stand within a distance of 40 from each other. 
> 
> I've tried using "nncluster" from the package nnclust, but the results 
are
> not correct, for some reason (probably my mistake).
> 
> This is what I did:
> 
> x <- nncluster(as.matrix(dframe[,1:2]), threshold=35, fill = 1, maxclust 
=
> NULL, give.up = 500,verbose=FALSE,start=NULL)#avaliar as clusters 
> 
> Thanks,
> 
> Frederico

        [[alternative HTML version deleted]]

______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to