Examples¶

Start with the Optimizer Catalog for a short, runnable setup for each optimizer. Install only the extra required by your chosen backend; SciPy and the iwopy optimizers are included in the base package.

Pareto flows¶

The catalog contains explicit unconstrained and constrained NSGA2/NSGA3 flows, including selection of a weighted compromise from iwopy Pareto results:

uv run --extra pymoo python examples/optimizers/pymoo_nsga2.py
uv run --extra pymoo python examples/optimizers/pymoo_nsga2_constrained.py

See pymoo NSGA3 and pymoo NSGA3 Constrained Flow for the reference-direction alternative.

Callback scripts¶

The examples/callbacks folder contains standalone SLSQP and pymoo scripts that print intermediate optimization results with a custom callback and record the same states with OptimizationHistory. Run them from the repository root:

uv run python examples/callbacks/run_slsqp.py
uv run --extra pymoo python examples/callbacks/run_pymoo.py

Both scripts save objective-history plots to the Git-ignored examples/callbacks/output directory.

Notebook workflows¶

The notebooks explore analytical and numerical derivatives, multi-objective results, and mixed variables. Some require pymoo or PyGMO; see their imports and Installation.