PySwarm

Particle Swarm Optimization (PSO) for Python

pyswarm is a gradient-free, evolutionary optimization library for Python that implements Particle Swarm Optimization (PSO) with built-in support for constraints. It is lightweight, easy to use, and suitable for a wide range of optimization problems where gradient information is unavailable or impractical to compute.

Quick example

from pyswarm import pso


def objective(x):
    x1, x2 = x
    return x1**4 - 2 * x2 * x1**2 + x2**2 + x1**2 - 2 * x1 + 5


lb = [-3, -1]
ub = [2, 6]

result = pso(objective, lb, ub)

print("Optimal solution:", result.x)  # → [1.0, 1.0]
print("Function value:", result.fun)  # → 4.0
print("Converged:", result.success)
print("Reason:", result.message)
print("Iterations:", result.nit)
print("Evaluations:", result.nfev)

pso returns an OptimizeResult — a dict subclass with attribute access, compatible with scipy.optimize.OptimizeResult.

Installation

pip install pyswarm

Requires Python 3.10+ and NumPy. See the full installation guide for uv, poetry, and source builds.

Documentation

• Theory — mathematical background, hierarchical basis, algorithms
• Quickstart — runnable examples
• API Reference — class and function signatures
• References — literature citations

License

BSD-3-Clause — see LICENSE.