bestfit {optpart}

Identify the Goodness-of-Fit of Cluster Members

Description

Sorts the members of clusters by maximum similarity goodness-of-fit

Usage

bestfit(x,cluster)

Arguments

x

an object of class ‘partana’ or ‘silhouette’

cluster

a specific cluster number

Details

Simply finds all members of a specific cluster and lists them in order of (1) mean similarity to their cluster (if x is an object of class ‘partana’) or silhouette width (if x is an object of class ‘silhouette’ as produced by functions in package ‘cluster’)

Value

returns a data.frame with cluster members in column ‘ID’ and goodness-of-fit in column ‘fit’

Author(s)

David W. Roberts [e-mail]

References

http://ecology.msu.montana.edu/labdsv/

See Also

typal

Examples

data(shoshveg)                      # returns vegetation matrix
dis.bc <- dsvdis(shoshveg,'bray')   # Bray/Curtis dissimilarity matrix 
opt.5 <- optpart(5,dis.bc)          # 5 cluster partition
print(class(opt.5))
fit <- bestfit(opt.5,1)             # goodness-of-fit for cluster 1
sil.5 <- silhouette(opt.5,dis.bc)   # calculate silhouette widths
fit2 <- bestfit(sil.5,1)            # goodness-of-fit for cluster 1