ClustOfVar: An R Package for the Clustering of Variables

Marie Chavent (1) , V. Kuentz Simonet (2) , Benoit Liquet (3) , Jérôme Saracco (4)

1 CQFD - Quality control and dynamic reliability
2 UR ADBX - Aménités et dynamiques des espaces ruraux
3 ISPED - Institut de Santé Publique, d'Epidémiologie et de Développement
4 IMB - Institut de Mathématiques de Bordeaux

Clustering of variables is as a way to arrange variables into homogeneous clusters, i.e., groups of variables which are strongly related to each other and thus bring the same information. These approaches can then be useful for dimension reduction and variable selection. Several specific methods have been developed for the clustering of numerical variables. However concerning qualitative variables or mixtures of quantitative and qualitative variables, far fewer methods have been proposed. The R package ClustOfVar was specifically developed for this purpose. The homogeneity criterion of a cluster is defined as the sum of correlation ratios (for qualitative variables) and squared correlations (for quantitative variables) to a synthetic quantitative variable, summarizing as "good as possible" the variables in the cluster. This synthetic variable is the first principal component obtained with the PCAMIX method. Two clustering algorithms are proposed to optimize the homogeneity criterion: iterative relocation algorithm and ascendant hierarchical clustering. We also propose a bootstrap approach in order to determine suitable numbers of clusters. We illustrate the methodologies and the associated package on small datasets.

Mots clés: STABILITY; MIXTURE OF QUANTITATIVE AND QUALITATIVE VARIABLES; K-MEANS CLUSTERING OF VARIABLES; HIERARCHICAL CLUSTERING OF VARIABLES; DIMENSION REDUCTION; ANALYSE STATISTIQUE; METHODOLOGIE; ALGORITHME

Domaines: Sciences de l'environnement

Marie Chavent, V. Kuentz Simonet, Benoit Liquet, Jérôme Saracco. ClustOfVar: An R Package for the Clustering of Variables. Journal of Statistical Software, 2012, 50 (13), pp.1-16. ⟨hal-00742795⟩