pyMetaheuristic Lab

Optimization Laboratory

Choose a workflow to begin. Each mode exposes different configuration options
and result visualizations tailored to the analysis you need.

🔬

Single Algorithm

Standard · Constrained · Binary

Run one metaheuristic on a continuous, constrained, or binary objective. Three sub-modes: standard optimization, constraint-handling (Deb/penalty/repair), and binary via BinaryAdapter with chaotic transfer functions.

Convergence chartsDiversityExploitation ratioPopulation snapshotFeasibility bannerBinary solution

Select →

📊

Benchmark Runner

Fast sweeps · Tables · Timing

Run a lightweight multi-algorithm × multi-problem sweep. Best when you need quick comparison tables, rankings, timing analysis, and CSV export.

Mean±std tableRankingsTimingCSV export

Select →

📈

Benchmark Study

Scientific protocol · Ranks · Tests

Run a reproducible benchmark study with long-format records, rank tables, scientific summaries, Friedman/Wilcoxon tests, and convergence data.

Rank tableStatistical testsScientific summaryRaw records

Select →

🏝️

Island System

Cooperative · Rules · Bandit · Portfolio

Build a heterogeneous island system and choose the coordination policy: fixed cooperation, rule-based orchestration, bandit controller, or portfolio-adaptive controller.

Migration networkIsland convergenceController decisionsProblem table

Select →

⌕

—

Loading algorithms…

◈

Select an algorithm from the list
to explore its capabilities and defaults.

—

— — —

Capabilities

Default Parameters

Continue to configuration →

—

∿

Click Run Optimization to start.
Results appear here live as the job runs.