Cut a tree into groups of ordered sizes

Cut a tree into groups of ordered sizes

Usage

cutreeIntoOrderedGroups(tree, k = NULL, h = NULL, decreasing = TRUE)

Arguments

tree

a tree as produced by hclust, will be passed to cutree

k

an integer scalar or vector with the desired number of groups

h

numeric scalar or vector with heights where the tree should be cut.

decreasing

logical, should be the first group the largest?

Cut a tree, e.g. as resulting from hclust, into groups, with the groups being ordered by their size.

Value

A named integer vector of cluster assignments, ordered by cluster size (largest first by default). If multiple values of k or h are provided, a matrix with one column per value.

See also

cutree

Examples

hc <- hclust(dist(USArrests))
hck5 <- cutreeIntoOrderedGroups(hc, k = 5)
table(hck5)
#> hck5
#>  1  2  3  4  5 
#> 14 14 10 10  2 
## compare with cutree, which does not order the groups
table(cutree(hc, k=5))
#> 
#>  1  2  3  4  5 
#> 14 14 10  2 10 

hck25 <- cutreeIntoOrderedGroups(hc, k = 2:5)
apply(hck25, 2, table)
#> $`2`
#> 
#>  1  2 
#> 34 16 
#> 
#> $`3`
#> 
#>  1  2  3 
#> 20 16 14 
#> 
#> $`4`
#> 
#>  1  2  3  4 
#> 20 14 14  2 
#> 
#> $`5`
#> 
#>  1  2  3  4  5 
#> 14 14 10 10  2 
#>