The HIPAM algorithm starts with one large cluster and, at each level, a given (parent) cluster is partitioned using PAM.
In this version of HIPAM, called $HIPAM_IMO$, the number k of (child) clusters is obtained by using the INCA (Index Number Clusters Atypical) criterion (Irigoien et al. (2008)) in the following way: at each node P, if there is k such that $INCA_k > 0.2$, then the k prior to the first largest slope decrease is selected. However, this procedure does not apply either to the top node or to the generation of the new partitions from which the Mean Split Silhouette is calculated. In these cases, even when all $INCA_k < 0.2$, k = 3 is fixed as the number of groups to divide and proceed. See Vinue et al. (2014) for more details.
The foundation and performance of the HIPAM algorithm is explained in hipamAnthropom.
getBestPamsamIMO(data,maxsplit,orness=0.7,type,ah,verbose,...)A list with the following elements:
medoids: The cluster medoids.
clustering: The clustering partition obtained.
asw: The asw of the clustering.
num.of.clusters: Number of clusters in the final clustering.
info: List that informs about the progress of the clustering algorithm.
profiles: List that contains the asw and sesw (stardard error of the silhouette widths) profiles at each stage of the search.
metric: Dissimilarity used (called 'McCulloch' because the dissimilarity function used is that explained in McCulloch et al. (1998)).
Data to be clustered.
The maximum number of clusters that any cluster can be divided when searching for the best clustering.
Quantity to measure the degree to which the aggregation is like a min or max operation. See weightsMixtureUB and getDistMatrix.
Option 'IMO' for using $HIPAM_IMO$.
Constants that define the ah slopes of the distance function in getDistMatrix. Given the five variables considered, this vector is c(23,28,20,25,25). This vector would be different according to the variables considered.
Boolean variable (TRUE or FALSE) to indicate whether to report information on progress.
Other arguments that may be supplied.
This function was originally created by E. Wit et al., and it is available freely on https://www.math.rug.nl/~ernst/book/smida.html. We have adapted it to incorporate the INCA criterion.
Vinue, G., Leon, T., Alemany, S., and Ayala, G., (2014). Looking for representative fit models for apparel sizing, Decision Support Systems 57, 22--33.
Wit, E., and McClure, J., (2004). Statistics for Microarrays: Design, Analysis and Inference. John Wiley & Sons, Ltd.
Wit, E., and McClure, J., (2006). Statistics for Microarrays: Inference, Design and Analysis. R package version 0.1. https://www.math.rug.nl/~ernst/book/smida.html.
Pollard, K. S., and van der Laan, M. J., (2002). A method to identify significant clusters in gene expression data. Vol. II of SCI2002 Proceedings, 318--325.
Irigoien, I., and Arenas, C., (2008). INCA: New statistic for estimating the number of clusters and identifying atypical units, Statistics in Medicine 27, 2948--2973.
Irigoien, I., Sierra, B., and Arenas, C., (2012). ICGE: an R package for detecting relevant clusters and atypical units in gene expression, BMC Bioinformatics 13 1--29.
McCulloch, C., Paal, B., and Ashdown, S., (1998). An optimization approach to apparel sizing, Journal of the Operational Research Society 49, 492--499.
hipamAnthropom