maxWithinClust: Return sorted data points using maximum within-cluster...

In lsheremet/imbc:

Description

Return sorted data points using maximum within-cluster distance

Usage

maxWithinClust(data, estep, previousResultIndicies, distanceMethod)

Arguments

data	the data you wish to use
estep	output from the mclust function "estep"
previousResultIndicies	indicies of rows the imbc algorithm previously queried the user on
distanceMethod	a method used to find distances between points. This must be one of "euclidean", "maximum", "manhattan", "canberra", "binary" or "minkowski", with a default of "euclidean."

Value

a vector of points sorted according to the maximum within-cluster distance method algorithm

Examples

#Load data
library(mclust)
data(banknote)

#Create new dataset with only continuous variables
bankdata <- banknote[,2:7]

#create Mclust object
object <- Mclust(bankdata)
#determine best model
model <- object$modelName
#extract model parameters
param <- object$parameters

#output from estep of EM algorithm
estepbank <- estep(model, bankdata, param)

#we assume we have not previously queried any points
#if this vector contained any row indicies, those would not show up in output sorted vector of the function
previousResultIndicies <- c()

#output from minimax algorithm: returns vector of points that are least confidently placed in their respective classes
maxWithinClust(bankdata, estep, previousResultIndicies)

#try with a different distance method
maxWithinClust(bankdata, estep, previousResultIndicies, distanceMethod = "manhattan")

lsheremet/imbc documentation built on May 20, 2019, 7:01 p.m.