pystra.active_learning.surrogates.Surrogate#

Use it: User guide · Worked example · Theory

class Surrogate[source]#

Bases: ABC

Predict a scalar limit state in independent standard normal space.

Both methods use row-wise points, shape (n_samples, n_variables). Implementations must replace fitted state on every fit.

Methods

`fit`	Fit finite scalar observations, shape (n_samples,).
`predict`	Return finite mean and nonnegative std arrays, each (n_samples,).

abstractmethod fit(points, values)[source]#

Fit finite scalar observations, shape (n_samples,).

abstractmethod predict(points)[source]#

Return finite mean and nonnegative std arrays, each (n_samples,).