ClusterStrings {TreeSearch}

Cluster similar strings

Description

Calculate string similarity using the Levenshtein distance and return clusters of similar strings.

Usage

ClusterStrings(x, maxCluster = 12)

Arguments

x

Character vector.

maxCluster

Integer specifying maximum number of clusters to consider.

Value

NameClusters() returns an integer assigning each element of x to a cluster, with an attribute med specifying the median string in each cluster, and silhouette reporting the silhouette coefficient of the optimal clustering. Coefficients < 0.5 indicate weak structure, and no clusters are returned. If the number of unique elements of x is less than maxCluster, all occurrences of each entry are assigned to an individual cluster.

Author(s)

Martin R. Smith ([e-mail])

See Also

Other utility functions: QuartetResolution(), WhenFirstHit()

Examples

ClusterStrings(c(paste0("FirstCluster ", 1:5),
                 paste0("SecondCluster.", 8:12),
                 paste0("AnotherCluster_", letters[1:6])))