typal: Identification of Typal Samples in a Partition

In optpart: Optimal Partitioning of Similarity Relations

Description

Identifies samples that typify clusters in a partition based on dissimilarity.

Usage

typal(clustering,dist,k=1)

Arguments

clustering	a vector of integers or an object of class ‘clustering’, ‘partition’, or ‘partana’
dist	and object of class ‘dist’ from dist, dsvdis, or vegdist
k	number of typal species/cluster to identify

Details

typal calculates two versions of typal species based on silhouette analysis (see silhouette) and partana analysis (see partana). With respect to silhouette analysis, the function returns k species with the largest positive silhouette width for each cluster. With respect to the partana analysis the function returns the k species with the highest mean similarity to the cluster.

Value

Returns a list with two data.frames. The first, partana, gives the clusters as rows and typal samples as columns from the perspective of the partana ratio. The second, silhouette, also gives the clusters as rows and typal samples as columns but from the perspective of silhouette widths.

Author(s)

David W. Roberts [e-mail]

Examples

data(shoshveg)
dis.bc <- dsvdis(shoshveg,'bray')
opt.5 <- bestopt(dis.bc,5,20)
typal(opt.5,dis.bc,3)

optpart documentation built on March 26, 2020, 6:18 p.m.