R/reordclust.R

In optpart: Optimal Partitioning of Similarity Relations

reordclust <- function (clustering,from,to)
{ 
    clustering <- as.integer(clustify(clustering))

    nfrom <- length(table(from))
    nto <- length(table(to))
    if (length(from)!=length(to)) stop("membership vectors must be the same size")

    out <-rep(0,length(clustering))
    for (i in 1:nfrom) {
        out[clustering==from[i]] <- to[i]
    }
    out <- as.numeric(factor(out))
    out <- list(clustering=out)
    class(out) <- 'clustering'
    attr(out,'call') <- match.call()
    attr(out,'timestamp') <- date()
    out
}