![]() |
Dakota Reference Manual
Version 6.16
Explore and Predict with Confidence
|
Samples variables along points moving out from a center point
This keyword is related to the topics:
Alias: none
Argument(s): none
Child Keywords:
| Required/Optional | Description of Group | Dakota Keyword | Dakota Keyword Description | |
|---|---|---|---|---|
| Required | step_vector | Size of steps to be taken in each dimension of a centered parameter study | ||
| Required | steps_per_variable | Number of steps to take in each dimension of a centered parameter study | ||
| Optional | model_pointer | Identifier for model block to be used by a method | ||
Dakota's centered parameter study computes response data sets along multiple coordinate-based vectors, one per parameter, centered about the initial values from the variables specification. This is useful for investigation of function contours with respect to each parameter individually in the vicinity of a specific point (e.g., post-optimality analysis for verification of a minimum), thereby avoiding the cost associated with a multidimensional grid.
Default Behavior
By default, the centered parameter study operates over all types of variables.
The centered_parameter_study takes steps along each orthogonal dimension. Each dimension is treated independently. The number of steps are taken in each direction, so that the total number of points in the parameter study is
.
Expected Outputs
A centered parameter study produces a set of responses for each parameter set that is generated.
Expected HDF5 Output
If Dakota was built with HDF5 support and run with the hdf5 keyword, this method writes the following results to HDF5:
The following example is a good comparison to the examples on multidim_parameter_study and vector_parameter_study.
# tested on Dakota 6.0 on 140501
environment
tabular_data
tabular_data_file = 'rosen_centered.dat'
method
centered_parameter_study
steps_per_variable = 5 4
step_vector = 0.4 0.5
model
single
variables
continuous_design = 2
initial_point = 0 0
descriptors = 'x1' "x2"
interface
analysis_driver = 'rosenbrock'
fork
responses
response_functions = 1
no_gradients
no_hessiansThese keywords may also be of interest: