choosenbclust: Tune the number of clusters according to the partition...

In clusterMI: Cluster Analysis with Missing Values by Multiple Imputation

Tune the number of clusters according to the partition instability

Description

choosenbclust reports the cluster instability according to the number of clusters chosen.

Usage

choosenbclust(output, grid = 2:5, graph = TRUE, verbose = TRUE, nnodes = NULL)

Arguments

output	an output from the clusterMI function
grid	a vector indicating the grid of values tested for nb.clust. By default 2:5
graph	a boolean indicating if a graphic is plotted
verbose	if TRUE, choosenbclust will print messages on console
nnodes	number of CPU cores for parallel computing. By default, the value used in the call to the clusterMI function

Details

The choosenbclust function browses a grid of values for the number of clusters and for each one imputes the data and computes the instability.

Value

a list of two objects

nb.clust	the number of clusters in grid minimizing the instability
crit	a vector indicating the instability for each value in the grid

References

Audigier, V. and Niang, N., Clustering with missing data: which equivalent for Rubin's rules? Advances in Data Analysis and Classification <doi:10.1007/s11634-022-00519-1>, 2022.

See Also

imputedata

Examples

data(wine, package = "clusterMI")

require(parallel)
set.seed(123456)
ref <- wine$cult
nb.clust <- 3
wine.na <- wine
wine.na$cult <- NULL
wine.na <- prodna(wine.na)

# imputation
res.imp <- imputedata(data.na=wine.na, nb.clust = nb.clust, m = 5)

# pooling
nnodes <- 2 # number of CPU cores for parallel computing
res.pool <- clusterMI(res.imp, nnodes = nnodes, instability = FALSE)

# choice of nb.clust

choosenbclust(res.pool)

clusterMI documentation built on April 4, 2025, 12:55 a.m.