[R-SIG-Finance] [R-sig-finance] http://www.market-topology.com/

[Yuri Volchik]

I managed to create something similar to subj by using igraph package:

adjMat<-(1-abs(cov.rob(dMat,cor=TRUE)$cor))^2
g2 <- graph.adjacency(adjMat,mode='plus',weighted=T, diag=F)
V(g2)$name<-Names
V(g2)$label<-Names
mst <- minimum.spanning.tree(g2)
plot(mst,layout=layout.fruchterman.reingold,  vertex.color="grey", 
vertex.size=20)


So far i'm quite happy with results i got (for small datasets), seem to
align pretty well with a priori assumptions.
I'm also wondering about what Spencer said about full rank correlation
matrix, if there is another suitable metric to use for computing distances
instead of correlation matrix?

Thanks



-- 
View this message in context: http://www.nabble.com/http%3A--www.market-topology.com--tp15786834p16527497.html
Sent from the Rmetrics mailing list archive at Nabble.com.