SPARTAAS (version 1.2.7)

hclustcompro: hclustcompro

Description

Compromised Hierarchical bottom-up clustering method. The method uses at least two sources of information. The merging of the data sources is done by weigth that allows to weight each source. Exemple for two sources: $$D_\alpha = \alpha D_1 + (1-\alpha) D_2$$

Usage

hclustcompro(d, method = "complete", weight = NULL)

Value

See stats::hclust or fastcluster::hclust

Arguments

d

a list of dissimilarity structure as produced by dist.

method

the agglomeration method to be used. This should be (an unambiguous abbreviation of) one of "ward.D", "ward.D2", "single", "complete", "average" (= UPGMA), "mcquitty" (= WPGMA), "median" (= WPGMC) or "centroid" (= UPGMC).

weight

The parameter in order to generate a mixed matrix on which the classical hclust method is applied (same length as d).
Formula: $$D_\alpha = w_1 D_1 + w_2 D_2 + ...$$
where \(w_1\) is the first value of vector weight, \(D_1\) the first matrix of d and \(D_\alpha\) the mixed matrix used in hclust. Sum of weight must be 1.

Author

A. COULON

L. BELLANGER

P. HUSI

Details

CAH
Data fusion (parameter optimal value see hclustcompro_select_alpha). It is necessary to define the appropriate proportion for each data source. This is the first sensitive point of the method that the user has to consider. A tool is provided to help him in his decision.

Cut dendrogram
The division into classes (see hclustcompro_cutree) is the second crucial point. It has to be done based on the knowledge of the study area and some decision support tools such as the cluster silhouette or the calculation of the intra-cluster variability (WSS: Within Sum of Square).

Examples

data(datangkor)

#network stratigraphic data (Network)
network <- datangkor$stratigraphy

#contingency table
cont <- datangkor$contingency

#obtain the dissimilarities matrices
distance <- CAdist(cont, nPC = 11)
constraint <- adjacency(network)

#You can also run hclustcompro with the dist matrix directly
tree = hclustcompro(d = list(distance, constraint), weight = c(0.7, 0.3), method = "ward.D2")
cluster = hclustcompro_cutree(tree, k=5, reorder=TRUE, letter=TRUE)