Version 3.2.0
Major Changes:
- UQ Options:
- Support for a gradient-free optimization method
- Support for stochastic Kriging
- Fast global sensitivity analysis for very high dimensional output (tested on 2 million QoIs)
- FEM
- "None" option for FEM that can be used when creating surrogates if user has data for which surrogate is required.
- Misc
- New option to discard working directories after each model simulation
- Significantly enhanced speed of surrogate validation and prediction
- Improved user interface including error bounds of the surrogate prediction
- Renaming of options:
(1) OpenseesPy->python (FEM), (2) Parameters estimation -> deterministic calibration (UQ/Dakota), (3) Inverse problem -> Bayesian calibration (UQ/Dakota)