saealib.ProbabilityOfFeasibility#

class saealib.ProbabilityOfFeasibility(obj_idx=0, reference=None)[source]#

Bases: PointwiseAcquisition

Probability of Feasibility (PoF) acquisition function.

Estimates the probability that a candidate satisfies a constraint g(x) <= 0, using a surrogate model that predicts the constraint value.

PoF(x) = Phi((0 - mu(x)) / sigma(x))

Typically used in combination with another acquisition function (e.g., EI * PoF) to handle black-box constraints.

Requires a surrogate that provides uncertainty estimates (std).

Parameters:

  • obj_idx (int) – Index of the predicted constraint to evaluate. Default: 0.

  • reference (Any)

References

[31]: Schonlau, M. (1997). Computer Experiments and Global Optimization (PhD thesis, University of Waterloo). (Earliest formulation.)

[32]: Gelbart, M. A., Snoek, J., & Adams, R. P. (2014). Bayesian optimization with unknown constraints. Proceedings of the 30th Conference on Uncertainty in Artificial Intelligence (UAI). (Constraint-handling with independent surrogates.)

Methods

__init__

compute_reference

Return fixed reference if set, otherwise None.

contract

Return the pointwise acquisition contract.

evaluate

Reject prediction is None, resolve the reference, then score().

prepare

Delegate to compute_reference(archive, rng=ctx.rng if ctx else None).

score

Compute Probability of Feasibility scores.

Method Details

ProbabilityOfFeasibility.__init__(obj_idx=0, reference=None)[source]#
Parameters:

  • obj_idx (int)

  • reference (Any)

ProbabilityOfFeasibility.compute_reference(archive, rng=None)[source]#

Return fixed reference if set, otherwise None.

Parameters:

  • archive (Archive)

  • rng (np.random.Generator | None)

Return type:

Any

ProbabilityOfFeasibility.contract()#

Return the pointwise acquisition contract.

Return type:

ComponentContract

ProbabilityOfFeasibility.evaluate(candidates_x, prediction, archive, ctx=None, *, prepared=<object object>)#

Reject prediction is None, resolve the reference, then score().

See AcquisitionFunction.evaluate() for the shared parameter contract, including the _UNSET vs. explicit prepared=None distinction.

Parameters:

  • candidates_x (np.ndarray)

  • prediction (SurrogatePrediction | None)

  • archive (Archive)

  • ctx (OptimizationState | None)

  • prepared (Any)

Return type:

AcquisitionResult

ProbabilityOfFeasibility.prepare(archive, ctx=None)#

Delegate to compute_reference(archive, rng=ctx.rng if ctx else None).

Parameters:

  • archive (Archive)

  • ctx (OptimizationState | None)

Return type:

Any

ProbabilityOfFeasibility.score(prediction, reference=None, rng=None)[source]#

Compute Probability of Feasibility scores.

Parameters:

  • prediction (SurrogatePrediction) – Surrogate predictions of constraint values. Must have std (has_uncertainty == True).

  • reference (Any) – Not used. Accepted for interface compatibility.

  • rng (Generator | None)

Returns:

PoF scores in [0, 1]. shape: (n_samples,) Higher scores indicate a higher probability of feasibility.

Return type:

np.ndarray

Raises:

TypeError – If prediction does not contain uncertainty estimates.