Constrained landscape features and benchmarks for constrained continuous optimization. Implements the constrained landscape features defined by Malan et al. (2015) and a Pythonic interface to the CEC 2010 constrained benchmark suite.

Developed by the HELIX at AGH University of Krakow.

Quickstart

import numpy as np
from clandscape.cec2010 import CEC2010Problem01
from clandscape.constrained_landscape_features import (
    ConstrainedLandscapeFeatures,
)

problem = CEC2010Problem01(dim=10)
features = ConstrainedLandscapeFeatures(
    bounds=problem.bounds,
    fitness=problem.fitness,
    violation=problem.violation,
)
df = features.compute_features(num_runs=10)
print(df.describe())

License

MIT — see LICENSE.