[R-sig-Geo] Join Count "correlogram"?

[amluiz]

Dear List,

I'm using an approach which calculates Manhattan distances in a matrix where
the columns represent the correlograms of species abundances, as follow:

require (vegan)
require (ncf)

data (mite)
data(mite.xy)
Y.hel<- decostand(mite, 'hel')
kk<- 10
inc<- 1
Correlog<- matrix(NA, nrow=ncol(Y.hel), ncol=kk)

for (i in 1:ncol(Y.hel)){
  cor<-correlog(x=mite.xy[,1], y=mite.xy[,2], z=Y.hel[,i], increment=inc,
resamp=1)
  nn<-  length (cor$correlation)
  Correlog[i,]<-as.matrix(cor$correlation)
  rownames(Correlog)<-colnames(Y.hel)
}
Correlog

However, my matrix is binary, i.e. rows are sites and columns species
presence (1) or absence (0) in a given site. 
1- Is there some alternative way to calculate an analogous matrix as in the
code above, with my presence/absence data? 
2- I've read that Moran's I does not deal with binary data, while Join Count
analysis would be more appropriated. But, is it possible generates
"correlograms" with join count analysis?
Any help will be appreciated.

Thanks in advance,

best,

Amom



--
View this message in context: http://r-sig-geo.2731867.n2.nabble.com/Join-Count-correlogram-tp7586143.html
Sent from the R-sig-geo mailing list archive at Nabble.com.