The new version of UQLab (V2.0) was released on February 1st, 2022.
This is an important milestone for us, as it includes the complete open source release of the software, including its core, which until now was closed.
UQLab is a general purpose Uncertainty Quantification software developed at ETH Zurich (Switzerland) that can solve all types of uncertainty quantification problems in engineering, economics and natural sciences: uncertainty propagation through Monte Carlo simulation (e.g. using surrogate models such as Gaussian processes and polynomial chaos expansions), sensitivity analysis, rare event estimation, Bayesian inference for inversion and model calibration, and much more.
With well over 4,000 unique users from 94 countries worldwide since its release in 2015, UQLab has become one of the reference uncertainty quantification software for research and teaching. As of this V2.0 version, UQLab also becomes accessible for free for companies and private entities.
In addition to the transition to a full-open source model (BSD 3-clause license), UQLab V2.0 packs several major new features:
- Full offline operation: the removal of the licensing system means that UQLab can now run without a live internet connection.
- Random fields module: a brand new module to easily represent, discretize and sample Gaussian and non-Gaussian random fields in multiple dimensions.
- Stochastic spectral embedding (SSE): fresh from its publication in 2021, UQLab provides now an implementation of this novel surrogate modeling technique, that can accurately represent models with highly inhomogeneous behavior throughout the input domain.
- Sampling-free Bayesian inversion: two new solvers based on the recently developed spectral likelihood expansion technique (SLE and SSLE), have been added to the arsenal of solvers of UQLab’s Bayesian inversion module. Reconstruct an analytical expression of the posterior distribution without the need of running MCMC sampling!
Visit our website www.uqlab.com
Register, download, enjoy!