optpp_g_newton
Newton method based least-squares calbration
Topics
package_optpp, local_optimization_methods
Specification
Alias: None
Arguments: None
Child Keywords:
Required/Optional |
Description of Group |
Dakota Keyword |
Dakota Keyword Description |
---|---|---|---|
Optional |
Select a search method for Newton-based optimizers |
||
Optional |
Balance goals of reducing objective function and satisfying constraints |
||
Optional |
Controls how close to the boundary of the feasible region the algorithm is allowed to move |
||
Optional |
Controls how closely the algorithm should follow the “central path” |
||
Optional |
Max change in design point |
||
Optional |
Stopping critiera based on L2 norm of gradient |
||
Optional |
Number of iterations allowed for optimizers and adaptive UQ methods |
||
Optional |
Stopping criterion based on objective function or statistics convergence |
||
Optional |
Compute speculative gradients |
||
Optional |
Number of function evaluations allowed for optimizers |
||
Optional |
Turn on scaling for variables, responses, and constraints |
||
Optional |
Identifier for model block to be used by a method |
Description
The Gauss-Newton algorithm is available as optpp_g_newton
and
supports unconstrained, bound-constrained, and generally-constrained
problems. When interfaced with the unconstrained, bound-constrained,
and nonlinear interior point full-Newton optimizers from the OPT++
library, it provides a Gauss-Newton least squares capability which –
on zero-residual test problems – can exhibit quadratic convergence
rates near the solution. (Real problems almost never have zero
residuals, i.e., perfect fits.)
See Package: OPT++ for info related to all optpp
methods.
Expected HDF5 Output
If Dakota was built with HDF5 support and run with the
hdf5
keyword, this method
writes the following results to HDF5:
Calibration (when
calibration_terms
are specified)Parameter Confidence Intervals (when
num_experiments
equals 1)