![]() |
Dakota Reference Manual
Version 6.15
Explore and Predict with Confidence
|
Characteristic values to scale continuous design variables
Alias: cdv_scales
Argument(s): REALLIST
Default: vector values = 1 . (no scaling)
Each real value in scales
is a nonzero characteristic value to be used in scaling each continuous design variable. They only have effect when the associated method specifies scaling
.
This keyword is required for scale_types of 'value'
and optional for 'auto'
and 'log'
. When specified in conjunction with log, scale values are applied prior to the logarithm, to permit log scaling of strictly negative design variables.
If a single real value is specified it will apply to all continuous design variables. Otherwise, a scale value must be specified for each continuous design variable.
Usage Tips:
When scales
are specified, but not scale_types, the scaling type is assumed to be 'value' for all continuous design variables.
Use scale values of 1.0 to selectively avoid scaling a subset of the variables.
Scaling for linear constraints is applied after any continuous variable scaling.
See the scaling information under specific methods, e.g., method-*-scaling
for details on how to use this keyword.
Two continuous design variables, one scaled by the characteristic value 4.0, the other log-scaled without additional value scaling
continuous_design = 2 initial_point -1.2 1.0 lower_bounds -200 0.001 upper_bounds 200 2.0 descriptors 'x1' "x2" scale_types = 'value' 'log' scales = 4.0 1.0