pystra.options.SimulationOptions#

class SimulationOptions(*, n_samples=100000, block_size=1000, target_cov=0.05, sampling_std=1.0, bins=None, transform=None, rosenblatt_order=None)[source]#

Bases: object

Settings for the simulation methods and active learning.

Parameters:

  • n_samples (int) – Sample budget: points for crude Monte Carlo, importance sampling and distribution analysis, lines for line sampling, and samples per level for subset simulation.

  • block_size (int) – Points passed to the limit-state function per call, and crude Monte Carlo’s convergence-check interval.

  • target_cov (float) – Crude Monte Carlo and importance sampling stop once the estimate’s coefficient of variation reaches this value; 0 uses the whole budget.

  • sampling_std (float) – Standard deviation of the sampling density in standard coordinates.

  • bins (int | None) – Histogram bins for distribution analysis; chosen from the sample size if omitted.

  • transform (str | None) – Isoprobabilistic transformation; None selects it from the model.

  • rosenblatt_order (tuple[int, ...] | None) – Conditioning order of the variables for the Rosenblatt transformation.

Methods