ds.varSelLcm: Conducts clustering using the VarSelLCM R function

In FlorianSchw/dsClusterAnalysisClient: DataSHIELD client side functions for the dsClusterAnalysisClient package

Conducts clustering using the VarSelLCM R function

Description

This function is similar to the native R function 'kmeans' from stats

Usage

ds.varSelLcm(
  df = NULL,
  num.clust = NULL,
  vbleSelec = TRUE,
  crit.varsel = "BIC",
  initModel = 50,
  nbcores = 1,
  nbSmall = 250,
  iterSmall = 20,
  nbKeep = 50,
  iterKeep = 1000,
  tolKeep = 1e-07,
  num.iterations = 5,
  newobj = NULL,
  datasources = NULL
)

Arguments

df	is a string character of the data set
num.clust	specifies the number of clusters for the computation
vbleSelec	specifies the max. number of iterations allowed
crit.varsel	relates to the number of random sets if clusters is a number and not a set of initial cluster centers
initModel	refers to the algorithm of calculating the kmeans and can be either 'Hartigan-Wong', 'Lloyd', 'Forgy' or 'MacQueen'
nbcores	is the name of the new object which is created with this function
nbSmall	is a logical or integer specifying whether tracing information on the progress of the algorithm is procuded for the Hartigan-Wong algorithm
iterSmall	is a logical or integer specifying whether tracing information on the progress of the algorithm is procuded for the Hartigan-Wong algorithm
nbKeep	is a logical or integer specifying whether tracing information on the progress of the algorithm is procuded for the Hartigan-Wong algorithm
iterKeep	is a logical or integer specifying whether tracing information on the progress of the algorithm is procuded for the Hartigan-Wong algorithm
tolKeep	represents the number at which point two successive models are defined to be converged; default is 1e-7
num.iterations	the number of iterations for finding SLMA clusters in each respective datasource
newobj	is a logical or integer specifying whether tracing information on the progress of the algorithm is procuded for the Hartigan-Wong algorithm
datasources	a list of DSConnection-class objects obtained after login

Details

The function calls the server-side function kmeansDS that computes the k-means clustering of a data set (type 'data.frame' or 'matrix'). The function creates a new object on the server-side, which is of class 'kmeans'. The new object is named by the user using the newobj argument, otherwise it is named kmeans.newobj by default.

Value

the object specified by the newobj argument of ds.kmeans or default name kmeans.newobj

Author(s)

Florian Schwarz for the German Institute of Human Nutrition

FlorianSchw/dsClusterAnalysisClient documentation built on Feb. 8, 2025, 10:32 a.m.