Generate.clusters: Multiple clusterings generation from the corresponding trees

In clusterv: Assessment of Cluster Stability by Randomized Maps

Multiple clusterings generation from the corresponding trees

Description

Multiple clusterings generation from the corresponding trees for a given cut (number of clusters).

Usage

Generate.clusters(tr, c = 3)

Arguments

tr	a list of trees as returned by the hclust algorithm
c	number of clusters

Value

A list of lists. Each list represents a clustering. Each cluster is a list of vectors, whose elements are the labels of the examples.

See Also

Achlioptas.hclustering.tree, PMO.hclustering.tree,

Norm.hclustering.tree,RS.hclustering.tree

Examples

# list of clusterings generated using Achlioptas random projections, 
# using cuts corresponding to 3, 4 and 10 clusters
M <- generate.sample0(n=10, m=2, sigma=1, dim=800)
list.trees <- Achlioptas.hclustering.tree(M, dim=100, hmethod = "average", 
                                          n = 20, scale = TRUE)
list.clusters3 <- Generate.clusters(list.trees, c = 3)
list.clusters4 <- Generate.clusters(list.trees, c = 4)
list.clusters10 <- Generate.clusters(list.trees, c = 10)

clusterv documentation built on June 8, 2025, 10:21 a.m.