CEC Functions#

The CEC (Competition on Evolutionary Computation) benchmark suites are included in the core installation. This page covers additional setup that may be required for certain CEC functions.

Installation#

CEC functions are included in the core installation:

pip install surfaces

No additional dependencies required.

Available CEC Suites#

Surfaces includes functions from multiple CEC competitions:

CEC 2013#

28 benchmark functions for real-parameter optimization.

from surfaces.test_functions.cec import (
    # Basic functions
    Rastrigin,
    RotatedRastrigin,
    RotatedDiscus,
    RotatedBentCigar,
    # Composition functions
    CompositionFunction1,
    CompositionFunction2,
    # ... and more
)

CEC 2014#

30 benchmark functions with improved characteristics.

from surfaces.test_functions.cec import (
    # CEC 2014 functions
    # ...
)

CEC 2017#

Updated benchmark suite with modern test functions.

from surfaces.test_functions.cec import (
    # CEC 2017 functions
    # ...
)

Usage Example#

from surfaces.test_functions.cec import RotatedRastrigin

# Create a 10-dimensional function
func = RotatedRastrigin(n_dim=10)

# Evaluate
result = func(func.search_space_sample())
print(f"Result: {result}")

# Get search space
space = func.search_space()

What CEC Functions Offer#

Shifted Optima#

Unlike basic algebraic functions where the optimum is often at the origin, CEC functions have shifted optima at unknown locations. This tests whether your optimizer can find solutions anywhere in the search space.

Rotated Search Spaces#

Many CEC functions apply rotation matrices to make the landscape non-separable. This tests whether your optimizer can handle correlated variables.

Composition Functions#

CEC includes composition functions that combine multiple basic functions. These create complex multi-modal landscapes that are particularly challenging for optimization algorithms.

Next Steps#

CEC Functions - Detailed guide to CEC functions
/api_reference/test_functions/cec - API reference