pytwinnet.optimization.bayesopt

Functions

rbf_kernel(X1, X2, lengthscale, variance)

Classes

SimpleBayesOpt(bounds[, init_points, iters, ...])

Tiny Bayesian Optimization (RBF GP + Expected Improvement).

class pytwinnet.optimization.bayesopt.SimpleBayesOpt(bounds, init_points=8, iters=32, lengthscale=0.5, variance=1.0, noise=1e-06, seed=0, X=<factory>, y=<factory>)

Bases: object

Tiny Bayesian Optimization (RBF GP + Expected Improvement).

Parameters:

• bounds (List[Tuple[float, float]])

• init_points (int)

• iters (int)

• lengthscale (float)

• variance (float)

• noise (float)

• seed (int)

• X (List[List[float]])

• y (List[float])

X: List[List[float]]

ask(n=1)

Return type:

List[List[float]]

Parameters:

n (int)

bounds: List[Tuple[float, float]]

init_points: int = 8

iters: int = 32

lengthscale: float = 0.5

noise: float = 1e-06

run(evaluate)

Return type:

Dict[str, Any]

Parameters:

evaluate (Callable[[List[float]], float])

seed: int = 0

suggest(n=1)

Return type:

List[List[float]]

Parameters:

n (int)

tell(X, y)

Return type:

None

Parameters:

• X (List[List[float]])

• y (List[float])

variance: float = 1.0

y: List[float]

pytwinnet.optimization.bayesopt.rbf_kernel(X1, X2, lengthscale, variance)

Return type:

ndarray

Parameters:

• X1 (ndarray)

• X2 (ndarray)

• lengthscale (float)

• variance (float)