pyMCMA: Uniformly distributed Pareto-front representation

Abstract

pyMCMA is the Python implementation of a novel method for autonomous computation of the Pareto-front representation composed of efficient solutions distributed uniformly in terms of distances between neighbor Pareto solutions. pyMCMA supports scientific, i.e. objective, model analysis by providing preference-free Pareto front representation. pyMCMA seamlessly integrates independently developed substantive models. The computed Pareto-front, also for more than two criteria, is visualized by interactive parallel coordinate plot, as well as by charts of criteria pairs. Moreover, pyMCMA optionally exports the results for problems-specific analysis in the substantive model’s variables space. The pyMCMA functionality is illustrated by an analysis of China’s liquid fuel production model.

Keywords

• Pareto-front representation
• Multiple-criteria model analysis
• Pareto-front visualization
• Pyomo modeling language
• Structured modeling