Skip to contents

K-mean parameter defaults.

params_kmeans_defaults.Rd

Helper function to generate defaults for the k-mean clustering were more control is needed.

Usage

params_kmeans_defaults()

Value

A named list with the following elements:

  • k_means_iter - Integer. The number of iterations to use for the clustering. Defaults to 30L.

  • k_means_init - String. The initialisation. One of c("parallel", "random"). Defaults to "parallel".

  • gemm - Boolean or NULL. Controls which CPU implementation is used by the method. GEMM is faster with large dimensionality. Defaults to FALSE.

  • hamerly - Boolean or NULL. Shall a faster exact method be used leveraging the triangle inequality. Faster on large data sets with large numbers of centroids. Defaults to TRUE.