ar.splitcl {Rarmec} | R Documentation |
Wrapper to group potential candidates by cluster analysis (ar.clust
)). Clusters can be formed by either choosing a fixed number of clusters (argument k
) or setting a dissimilarity threshold (argument h
).
ar.splitcl(hc, sol,x, k = NULL, h = 0.3)
hc |
Object from ar.clust |
sol |
Matrix from ar.getlist |
x |
Original data matrix |
k |
Number of clusters to be generated |
h |
Dissimilarity based threshold to define the clusters (1-correlation) |
List of clusters - Each contains
cor |
Correlation matrix |
met |
Characteristics of each solution as for ar.getlist |
David Enot dle@aber.ac.uk>
### Check ar.writecl for complete example