jorgeparraga-arch.r-universe.dev

[jorgeklz] momst 0.1.1

jorge.parraga@utm.edu.ec (Jorge A. Parraga-Alava) — Sat, 10 Oct 2026 13:03:10 GMT

Solves the Multi-Criteria Minimum Spanning Tree (mc-MST) problem on complete weighted graphs by combining the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with optional Pareto local search operators. Chromosomes are represented as Prufer sequences so that every random individual decodes to a valid spanning tree (Cayley's theorem), avoiding repair operators. Four solver variants are provided: pure NSGA-II ("base"), Path Relinking ("PR"), Pareto Local Search ("PLS"), and Tabu Search ("TS"). The package supports 2 and 3 objective formulations and provides convenience functions to plot Pareto fronts and best-compromise spanning trees. This package is the reference implementation of the method described in Parraga-Alava, Inostroza-Ponta and Dorn (2017) .

[jorgeklz] moc.gapbk 0.3.0

jorge.parraga@utm.edu.ec (Jorge Parraga-Alava) — Fri, 15 May 2026 18:17:59 GMT

Implements the Multi-Objective Clustering Algorithm Guided by a-Priori Biological Knowledge ('MOC-GaPBK') proposed by Parraga-Alava and others (2018) . The algorithm performs gene clustering using 'NSGA-II' as the underlying multi-objective evolutionary engine, together with Path-Relinking and Pareto Local Search as intensification and diversification strategies. Two versions of the Xie-Beni validity index are used as objective functions, one per distance matrix, so that prior biological knowledge can be incorporated through the second matrix.

[jorgeklz] mstknnclust 1.0.0

jorge.parraga@utm.edu.ec (Jorge Parraga-Alava) — Mon, 11 May 2026 12:58:45 GMT

Implements the MST-kNN clustering algorithm proposed by Inostroza-Ponta (2008) . The algorithm determines the number of clusters automatically by recursively intersecting the Minimum Spanning Tree (MST) and the k-Nearest Neighbor (kNN) proximity graphs constructed from a pairwise distance matrix. The value of k is selected via a connectivity criterion (the smallest k such that the kNN graph is connected, bounded by floor(log(n))). The package requires only a distance matrix as input and returns cluster assignments, an igraph network, and partition metadata.