run_kmeans: Conduct k-means partitioning

In opm: Analysing Phenotype Microarray and Growth Curve Data

Description

Run a k-means partitioning analysis. This function is used by discrete in ‘gap’ mode to automatically determine the range of ambiguous data. If applied to such one-dimensional data, it uses an exact algorithm from the Ckmeans.1d.dp package.

Usage

  ## S4 method for signature 'matrix,numeric'
run_kmeans(object, k, cores = 1L, nstart = 10L, ...)
  ## S4 method for signature 'numeric,numeric'
run_kmeans(object, k, cores = 1L)

Arguments

object	Numeric vector or matrix.
k	Numeric vector. Number of clusters requested.
nstart	Numeric scalar. Ignored if ‘Ckmeans.1d.dp’ is called. Otherwise passed to kmeans from the stats package.
cores	Numeric scalar indicating the number of cores to use.
...	List of optional arguments passed to kmeans from the stats package.

Value

S3 object of class kmeanss, basically a named list of kmeans objects.

References

Wang, H., Song, M. 2011 Ckmeans.1d.dp: Optimal k-means clustering in one dimension by dynamic programming. The R Journal 3, p. 29–33.

See Also

stats::kmeans Ckmeans.1d.dp::Ckmeans.1d.dp

Other kmeans-functions: borders, calinski, hist.Ckmeans.1d.dp, hist.kmeans, hist.kmeanss, plot.kmeanss, to_kmeans,

Examples

x <- as.vector(extract(vaas_4, as.labels = NULL, subset = "A"))
summary(x.km <- run_kmeans(x, k = 1:10)) # => 'kmeanss' object
stopifnot(inherits(x.km, "kmeanss"), length(x.km) == 10)
stopifnot(sapply(x.km, class) == "kmeans", names(x.km) == 1:10)

opm documentation built on May 2, 2019, 6:08 p.m.