R/silinfo.R

In tidydr: Unify Dimensionality Reduction Results

##' Choose best K (number of clusters)
##' 
##' This function calculate the silhouette scores of each K (number of clusters). 
##' The output object can be used to choose the best K (via `summary()` or `autoplot()` methods)
##' @title nk
##' @param data input data (a matrix or data frame)
##' @param k a vector of candidate number of clusters
##' @return a `silinfo` object, which contains 'data' (original data), 'silinfo' (silhouette scores), and k (the input k vector)
##' @importFrom cluster pam
##' @export
##' @examples
##' x <- nk(iris[,-5], 2:8)
##' summary(x)
##' # to visualize the average silhouete score (y axis) with k (x axis)
##' autoplot(x)
##' # to visualize a PCA plot color by the choosing k
##' autoplot(x, k=3)
##' @author Guangchuang Yu
nk <- function(data, k) {
    y = lapply(k, function(i) {
        # x <- cluster::silhouette(cluster::pam(data, k=i))
        # mean(x[, "sil_width"])

        x <- pam(data, k=i)
        # x$silinfo$avg.width
        x$silinfo
    })
    structure(list(data = data,
                silinfo = y, 
                k=k),
        class = "silinfo"
    )
}

##' @method print silinfo
##' @export
print.silinfo <- function(x, ...) {
    y <- summary(x)
    k <- y[which.max(y[,2]), 1]

    msg <- c(
        sprintf("Silhouette information for K = %s", paste0(range(x$k), collapse=":")),
        sprintf("Best K (number of clusters) is %s", k)
        )
    cat(msg, sep = "\n")
}

##' @method summary silinfo
##' @export
summary.silinfo <- function(object, ...) {
    K <- object$k
    y <- vapply(object$silinfo, function(x) {
            x$avg.width
        }, FUN.VALUE = numeric(1))

    res <- data.frame(K=K, 
            Silhouette = y)
    return(res)
}