momst web app

Try the R package in your browser: multi-objective minimum spanning trees with NSGA-II and Pareto local search.

Parameters

Instance

Random Upload file

Nodes (n)

Objectives

Instance seed

Weights, objective 1

Weights, objective 2

Weights, objective 3

Solver

Variants: NSGA-II alone or combined with a local search.

Base NSGA-II Path Relinking Pareto LS Tabu Search

Independent runs

Population size (even)

Max generations

Solver seed

NSGA-II parameters

Crossover rate

Mutation rate

Tournament size

Convergence window

Global Pareto front by variant

Click a point to draw its spanning tree.

Run log

Press Run solver to start.

What it does

momst solves the Multi-Criteria Minimum Spanning Tree problem on complete weighted graphs with NSGA-II, optionally combined with one of three Pareto local search operators. Chromosomes are Prufer sequences, so every individual decodes to a valid spanning tree.

Variant	Local search
`base`	None (pure NSGA-II)
`PR`	Path Relinking
`PLS`	Pareto Local Search
`TS`	Tabu Search

How this page works

This page runs a JavaScript port of the R package, including R's random number generator. With the same instance, parameters and seed it returns the same Pareto front as the R package, which is checked automatically on every deployment.

The same analysis in R

library(momst)
inst <- generate_instance(n = 10, num_obj = 2, seed = 12345)
res  <- run_momst(instance = inst, n = 10, num_obj = 2, variant = "PLS",
                  iterations = 2, pop_size = 30, max_generations = 30,
                  seed = 2026, verbose = FALSE)
plot_pareto_front(res)

# Or the interactive app, locally
momst::run_app()

Reference

Parraga-Alava, J., Inostroza-Ponta, M. and Dorn, M. (2017). Using local search strategies to improve the performance of NSGA-II for the Multi-Criteria Minimum Spanning Tree problem. IEEE CEC 2017, pp. 1818 to 1825. doi:10.1109/CEC.2017.7969432

Documentation: jorgeklz.github.io/momst · Source code: github.com/jorgeklz/momst