surrogate package
Notebooks
surrogate.bootstrap.bootstrap module
- surrogate.bootstrap.bootstrap.bootstrap_ci(xdata, ydata, estimator, n_boot, **kwargs)[source]
Bootstrap with replacement for confidence interval
- Parameters:
xdata (numpy.ndarray) – input data (N, D) with N number samples D dimension
ydata (numpy.ndarray) – output data (N, 1)
estimator (BaseEstimator) – UQ model
n_boot (int) – number of bootstraps
n_samples (int, optional) – The bootstrap sample size. If not provided, n_samples = N
seed (int, optional) – The seed of the numpy random number generator.
- Returns:
y_lo (N, 1) evaluated at xdata y_up (N, 1) evaluated at xdata
- Return type:
[numpy.ndarray, numpy.ndarray]
- surrogate.bootstrap.bootstrap.bootstrap_sobol(xdata, ydata, estimator, n_boot, **kwargs)[source]
Bootstrap with replacement of Sobol indices
- Parameters:
xdata (numpy.ndarray) – input data (N, D) with N number samples D dimension
ydata (numpy.ndarray) – output data (N, 1)
estimator (BaseEstimator) – UQ model
n_boot (int) – number of bootstraps
n_samples (int, optional) – The bootstrap sample size. If not provided, n_samples = N
seed (int, optional) – The seed of the numpy random number generator.
Notes
Implemented according to G. E. B. Archer, A. Saltelli & I. M. Sobol (1997) Sensitivity measures, anova-like Techniques and the use of bootstrap, Journal of Statistical Computation and Simulation, 58:2,99-120, DOI: 10.1080/00949659708811825
- Returns:
Two lists of CI upper bounds, i.e. main sensitivities, total sensitivites
- Return type:
list