saealib#
saealib is a Python optimization library for objective functions whose evaluations take substantial time or cost. It combines a model that approximates the objective from past evaluations with direct evaluation to reduce the number of expensive evaluations.
Documentation#
Learn the common ways to get started.
Follow task-oriented procedures.
Explore the pieces that make up an optimization.
Explore contracts, ComponentGraph, Compiler, and Runtime.
Review algorithm configurations and sources.
Browse the public API.
Minimal example#
from saealib import minimize
from saealib.benchmarks import rastrigin
result = minimize(rastrigin(n_var=10), max_fe=300, seed=0)
print(result.x, result.f)
Configure which components to use and how to combine them. Configuration conflicts are validated before optimization starts. See What is saealib? and Optimization components for details.
Installation instructions are in Installation, and implementation sources are listed in References.