randomclustersim: Simulation of validity indexes based on random clusterings

In fpc: Flexible Procedures for Clustering

Simulation of validity indexes based on random clusterings

Description

For a given dataset this simulates random clusterings using stupidkcentroids, stupidknn, stupidkfn, and stupidkaven. It then computes and stores a set of cluster validity indexes for every clustering.

Usage

  randomclustersim(datadist,datanp=NULL,npstats=FALSE,useboot=FALSE,
                      bootmethod="nselectboot",
                      bootruns=25, 
                      G,nnruns=100,kmruns=100,fnruns=100,avenruns=100,
                      nnk=4,dnnk=2,
                      pamcrit=TRUE, 
                      multicore=FALSE,cores=detectCores()-1,monitor=TRUE)

Arguments

datadist	distances on which validation-measures are based, dist object or distance matrix.
datanp	optional observations times variables data matrix, see npstats.
npstats	logical. If TRUE, distrsimilarity is called and the two statistics computed there are added to the output. These are based on datanp and require datanp to be specified.
useboot	logical. If TRUE, a stability index (either nselectboot or prediction.strength) will be involved.
bootmethod	either "nselectboot" or "prediction.strength"; stability index to be used if useboot=TRUE.
bootruns	integer. Number of resampling runs. If useboot=TRUE, passed on as B to nselectboot or M to prediction.strength.
G	vector of integers. Numbers of clusters to consider.
nnruns	integer. Number of runs of stupidknn.
kmruns	integer. Number of runs of stupidkcentroids.
fnruns	integer. Number of runs of stupidkfn.
avenruns	integer. Number of runs of stupidkaven.
nnk	nnk-argument to be passed on to cqcluster.stats.
dnnk	nnk-argument to be passed on to distrsimilarity.
pamcrit	pamcrit-argument to be passed on to cqcluster.stats.
multicore	logical. If TRUE, parallel computing is used through the function mclapply from package parallel; read warnings there if you intend to use this; it won't work on Windows.
cores	integer. Number of cores for parallelisation.
monitor	logical. If TRUE, it will print some runtime information.

Value

List with components

nn	list, indexed by number of clusters. Every entry is a data frame with nnruns observations for every simulation run of stupidknn. The variables of the data frame are avewithin, mnnd, cvnnd, maxdiameter, widestgap, sindex, minsep, asw, dindex, denscut, highdgap, pearsongamma, withinss, entropy, if pamcrit=TRUE also pamc, if npstats=TRUE also kdnorm, kdunif. All these are cluster validation indexes; documented as values of clustatsum.
fn	list, indexed by number of clusters. Every entry is a data frame with fnruns observations for every simulation run of stupidkfn. The variables of the data frame are avewithin, mnnd, cvnnd, maxdiameter, widestgap, sindex, minsep, asw, dindex, denscut, highdgap, pearsongamma, withinss, entropy, if pamcrit=TRUE also pamc, if npstats=TRUE also kdnorm, kdunif. All these are cluster validation indexes; documented as values of clustatsum.
aven	list, indexed by number of clusters. Every entry is a data frame with avenruns observations for every simulation run of stupidkaven. The variables of the data frame are avewithin, mnnd, cvnnd, maxdiameter, widestgap, sindex, minsep, asw, dindex, denscut, highdgap, pearsongamma, withinss, entropy, if pamcrit=TRUE also pamc, if npstats=TRUE also kdnorm, kdunif. All these are cluster validation indexes; documented as values of clustatsum.
km	list, indexed by number of clusters. Every entry is a data frame with kmruns observations for every simulation run of stupidkcentroids. The variables of the data frame are avewithin, mnnd, cvnnd, maxdiameter, widestgap, sindex, minsep, asw, dindex, denscut, highdgap, pearsongamma, withinss, entropy, if pamcrit=TRUE also pamc, if npstats=TRUE also kdnorm, kdunif. All these are cluster validation indexes; documented as values of clustatsum.
nnruns	number of involved runs of stupidknn,
fnruns	number of involved runs of stupidkfn,
avenruns	number of involved runs of stupidkaven,
kmruns	number of involved runs of stupidkcentroids,
boot	if useboot=TRUE, stability value; stabk for method nselectboot; mean.pred for method prediction.strength.

Author(s)

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

References

Hennig, C. (2019) Cluster validation by measurement of clustering characteristics relevant to the user. In C. H. Skiadas (ed.) Data Analysis and Applications 1: Clustering and Regression, Modeling-estimating, Forecasting and Data Mining, Volume 2, Wiley, New York 1-24, https://arxiv.org/abs/1703.09282

Akhanli, S. and Hennig, C. (2020) Calibrating and aggregating cluster validity indexes for context-adapted comparison of clusterings. Statistics and Computing, 30, 1523-1544, https://link.springer.com/article/10.1007/s11222-020-09958-2, https://arxiv.org/abs/2002.01822

See Also

stupidkcentroids, stupidknn, stupidkfn, stupidkaven, clustatsum

Examples

  set.seed(20000)
  options(digits=3)
  face <- rFace(10,dMoNo=2,dNoEy=0,p=2)
  rmx <- randomclustersim(dist(face),datanp=face,npstats=TRUE,G=2:3,
    nnruns=2,kmruns=2, fnruns=1,avenruns=1,nnk=2)
## Not run: 
  rmx$km # Produces slightly different but basically identical results on ATLAS

## End(Not run)
  rmx$aven
  rmx$fn
  rmx$nn
  

fpc documentation built on Sept. 24, 2024, 9:07 a.m.