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


Access to methods in the DOT package

Topics

This keyword is related to the topics:

Specification

Alias: none

Argument(s): none

Required/Optional Description of Group Dakota Keyword Dakota Keyword Description
Required
(Choose One)
Group 1 frcg A conjugate gradient optimization method
mmfd Method of feasible directions
bfgs A conjugate gradient optimization method
slp Sequential Linear Programming
sqp Sequential Quadratic Program
Optional max_iterations

Stopping criterion based on number of iterations

Optional convergence_tolerance

Stopping criterion based on convergence of the objective function or statistics

Optional constraint_tolerance The maximum allowable value of constraint violation still considered to be feasible
Optional speculative Compute speculative gradients
Optional max_function_evaluations Stopping criteria based on number of function evaluations
Optional scaling Turn on scaling for variables, responses, and constraints
Optional linear_inequality_constraint_matrix Define coefficients of the linear inequality constraints
Optional linear_inequality_lower_bounds Define lower bounds for the linear inequality constraint
Optional linear_inequality_upper_bounds Define upper bounds for the linear inequality constraint
Optional linear_inequality_scale_types Specify how each linear inequality constraint is scaled
Optional linear_inequality_scales Define the characteristic values to scale linear inequalities
Optional linear_equality_constraint_matrix Define coefficients of the linear equalities
Optional linear_equality_targets Define target values for the linear equality constraints
Optional linear_equality_scale_types Specify how each linear equality constraint is scaled
Optional linear_equality_scales Define the characteristic values to scale linear equalities
Optional model_pointer

Identifier for model block to be used by a method

Description

The DOT library [82] contains nonlinear programming optimizers, specifically the Broyden-Fletcher-Goldfarb-Shanno (Dakota's dot_bfgs method) and Fletcher-Reeves conjugate gradient (Dakota's dot_frcg method) methods for unconstrained optimization, and the modified method of feasible directions (Dakota's dot_mmfd method), sequential linear programming (Dakota's dot_slp method), and sequential quadratic programming (Dakota's dot_sqp method) methods for constrained optimization.

Specialized handling of linear constraints is supported with DOT; linear constraint coefficients, bounds, and targets can be provided to DOT at start-up and tracked internally.

One of the five available methods in Group 1 must be specified.

All these methods take the same Optional Keywords , dealing with linear equality and inequality constraints.

Method Independent Controls - Stopping Critiera

Stopping critiera are set by:

Note: The convergence_tolerance criterion must be satisfied for two consecutive iterations before DOT will terminate.

Method Independent Controls - Output

The output verbosity specification controls the amount of information generated by DOT: the silent and quiet settings result in header information, final results, and objective function, constraint, and parameter information on each iteration; whereas the verbose and debug settings add additional information on gradients, search direction, one-dimensional search results, and parameter scaling factors.

Concurrency

DOT contains no parallel algorithms which can directly take advantage of concurrent evaluations. However, if numerical_gradients with method_source dakota is specified, then the finite difference function evaluations can be performed concurrently (using any of the parallel modes described in the Users Manual[5]). In addition, if speculative is specified, then gradients (dakota numerical or analytic gradients) will be computed on each line search evaluation in order to balance the load and lower the total run time in parallel optimization studies.