clustatis_kmeans {ClustBlock}

Compute the CLUSTATIS partitioning algorithm on different blocks of quantitative variables. Can be performed using a multi-start strategy or initial partition provided by the user.

Description

Partitioning algorithm for quantitative variables. Each cluster is associated with a compromise computed by the STATIS method. Moreover, a noise cluster can be set up.

Usage

clustatis_kmeans(Data, Blocks, clust, nstart=100, rho=0, NameBlocks=NULL,
Itermax=30,Graph_groups=TRUE, Graph_weights=FALSE,
 scale=FALSE, print_attempt=FALSE)

Arguments

Data

data frame or matrix. Correspond to all the blocks of variables merged horizontally

Blocks

numerical vector. The number of variables of each block. The sum must be equal to the number of columns of Data

clust

numerical vector or integer. Initial partition or number of starting partitions if integer. If numerical vector, the numbers must be 1,2,3,...,number of clusters

nstart

integer. Number of starting partitions. Default: 100

rho

numerical between 0 and 1. Threshold for the noise cluster. Default:0

NameBlocks

string vector. Name of each block. Length must be equal to the length of Blocks vector. If NULL, the names are B1,...Bm. Default: NULL

Itermax

numerical. Maximum of iterations by partitionning algorithm. Default: 30

Graph_groups

logical. Should each cluster compromise be plotted? Default: TRUE

Graph_weights

logical. Should the barplot of the weights in each cluster be plotted? Default: FALSE

scale

logical. Should the data variables be scaled? Default: FALSE

print_attempt

logical. Print the number of remaining attempts in the multi-start case? Default: FALSE

Value

a list with:

• group: the clustering partition. If rho>0, some blocks could be in the noise cluster ("K+1")

• rho: the threshold for the noise cluster

• homogeneity: percentage of homogeneity of the blocks in each cluster and the overall homogeneity

• rv_with_compromise: RV coefficient of each block with its cluster compromise

• weights: weight associated with each block in its cluster

• comp_RV: RV coefficient between the compromises associated with the various clusters

• compromise: the W compromise of each cluster

• coord: the coordinates of objects of each cluster

• inertia: percentage of total variance explained by each axis for each cluster

• rv_all_cluster: the RV coefficient between each block and each cluster compromise

• criterion: the CLUSTATIS criterion error

• param: parameters called

• type: parameter passed to other functions

References

Llobell, F., Cariou, V., Vigneau, E., Labenne, A., & Qannari, E. M. (2018). Analysis and clustering of multiblock datasets by means of the STATIS and CLUSTATIS methods. Application to sensometrics. Food Quality and Preference, in Press.
Llobell, F., Vigneau, E., Qannari, E. M. (2019). Clustering datasets by means of CLUSTATIS with identification of atypical datasets. Application to sensometrics. Food Quality and Preference, 75, 97-104.

See Also

plot.clustatis, clustatis, summary.clustatis, statis

Examples

 data(smoo)
 NameBlocks=paste0("S",1:24)
 #with multi-start
 cl_km=clustatis_kmeans(Data=smoo,Blocks=rep(2,24),NameBlocks = NameBlocks, clust=3)
 #with an initial partition
 cl=clustatis(Data=smoo,Blocks=rep(2,24),NameBlocks = NameBlocks,
 Graph_dend=FALSE)
 partition=cl$cutree_k$partition3
 cl_km2=clustatis_kmeans(Data=smoo,Blocks=rep(2,24),NameBlocks = NameBlocks,
 clust=partition, Graph_weights=FALSE, Graph_groups=FALSE)
 graphics.off()