Dakota Reference Manual  Version 6.4
Large-Scale Engineering Optimization and Uncertainty Analysis
 All Pages
dimension_adaptive


Perform anisotropic expansion refinement by preferentially adapting in dimensions that are detected to have higher `importance'.

Specification

Alias: none

Argument(s): none

Required/Optional Description of Group Dakota Keyword Dakota Keyword Description
Required
(Choose One)
dimension adaptivity estimation approach (Group 1) sobol

Estimate dimension preference for automated refinement of stochastic expansion using total Sobol' sensitivity indices.

decay

Estimate spectral coefficient decay rates to guide dimension-adaptive refinement.

generalized

Use the generalized sparse grid dimension adaptive algorithm to refine a sparse grid approximation of stochastic expansion.

Description

Perform anisotropic expansion refinement by preferentially adapting in dimensions that are detected to hold higher `importance' in resolving statistical quantities of interest.

Dimension importance must be estimated as part of the refinement process. Techniques include either sobol or generalized for stochastic collocation and either sobol, decay, or generalized for polynomial chaos. Each of these automated refinement approaches makes use of the max_iterations and convergence_tolerance iteration controls.