pystra.active_learning.estimation.EnrichmentEstimator#

Use it: User guide · Worked example · Theory

class EnrichmentEstimator[source]#

Bases: ReliabilityEstimator

Estimator that also constructs a new learning population after each fit.

The runner restarts an exploration RNG from the same seed for each fit (common random numbers), and uses a separate stream for final estimation. Implementations must not retain mutable run state. Sampling weights and dependence belong in returned diagnostics, not in the learning function.

Methods

`estimate`	Estimate conditional Pf from a frozen surrogate, returning a snapshot.
`explore`	Return at most n_candidates selectable points and fit diagnostics.

abstractmethod explore(predict, *, dimension, rng, n_candidates)[source]#

Return at most n_candidates selectable points and fit diagnostics.

Return type:: EnrichmentResult

abstractmethod estimate(predict, *, dimension, rng)#

Estimate conditional Pf from a frozen surrogate, returning a snapshot.

Return type:: ReliabilityEstimate