kerndenscluster {otrimle}

Aggregated distance to elliptical unimodal density over clusters

Description

This calls kerndensp for computing and aggregating density- and principal components-based distances between multivariate data and a unimodal elliptical distribution about the data mean for all clusters in a mixture-based clustering as generated by otrimle or rimle. For use in otrimleg.

Usage

  kerndenscluster(x,fit,maxq=qnorm(0.9995),kernn=100)

Arguments

x

something that can be coerced into a matrix. Dataset.

fit

output object of otrimle or rimle.

maxq

positive numeric. One-dimensional densities are evaluated between mean(x)-maxq and mean(x)+maxq.

kernn

integer. Number of points at which the one-dimensional density is evaluated, input parameter n of density. This should be even.

Details

See Hennig and Coretto (2021), Sec. 4.2. kerndenscluster calls kerndensp for all clusters and aggregates the resulting measures as root sum of squares.

Value

A list with components ddpi, ddpm, measure.

ddpi

list of outputs of kerndensp for all clusters.

ddpm

vector of measure-components of ddpi.

measure

Final aggregation result.

Author(s)

Christian Hennig [e-mail] https://www.unibo.it/sitoweb/christian.hennig/en/

References

Hennig, C. and P.Coretto (2021). An adequacy approach for deciding the number of clusters for OTRIMLE robust Gaussian mixture based clustering. To appear in Australian and New Zealand Journal of Statistics, https://arxiv.org/abs/2009.00921.

See Also

kerndensp, kerndensmeasure, otrimle, rimle

Examples

   data(banknote)
   selectdata <- c(1:30,101:110,117:136,160:161)
   set.seed(555566)
   x <- banknote[selectdata,5:7]
   ox <- otrimle(x, G=2, ncores=1)
   kerndenscluster(x,ox)$measure