DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) is a general-purpose software toolkit for performing systems analysis and design on high performance computers. It provides algorithms for design optimization, uncertainty quantification, parameter estimation, design of experiments, and sensitivity analysis, as well as a range of parallel computing and simulation interfacing services.
|Operating Systems||Windows Cygwin POSIX Unix|
The previous URL for Sandia National Labs' DAKOTA project won't work. If you use this project, please make a note that it is now at http://dakota.sandia.gov.
Release Notes: Minor bugfixes and enhancements.
Release Notes: The license has changed from GPL to LGPL with this release. There were numerous analysis, framework, and performance improvements along with bugfixes. New uniform and adaptive polynomial order refinement ("p-refinement") functions were provided for PCE and SC using tensor and sparse grids.
Release Notes: The license has been changed to the GNU Lesser General Public License. The JAGUAR 2.0 graphical user interface was added for creating, editing, and running DAKOTA input files. Additional discrete range and discrete set types within design, uncertain, and state variables were added. Anisotropic sparse grids, numerically-generated orthogonal polynomials, and improved expansion tailoring for stochastic expansion UQ methods were added. New methods were added for epistemic and mixed aleatory-epistemic uncertainty quantification. Other enhancements and bugfixes were made.
Release Notes: A new stochastic collocation method based on Lagrange interpolation. Extended quadrature and sparse grid capabilities. Additional random variable support in LHS and Nataf transformation. Generalized incremental Monte Carlo sampling. Extended polynomial chaos-based design optimization capability. New capabilities for model calibration under uncertainty. Enhanced parallelism in sequential hybrid strategy. A direct interface to APPSPACK with support for linear/nonlinear constraints. Improved support for Darwin PPC and Cygwin platforms, Intel and PGI compilers, and OpenMPI message passing.
Release Notes: Wiener-Askey generalized polynomial chaos expansions, efficient global reliability analysis (EGRA), incremental Latin hypercube and incremental Monte Carlo sampling, and (multimodal/adaptive) importance sampling were added. The optimization methods were added and extended. A Python interface, support for failure detection/mitigation, and mixed surrogate/truth models were added. Numerous bugs were fixed.