![]() |
Dakota Reference Manual
Version 6.16
Explore and Predict with Confidence
|
Activate adaptive procedure for determining the best basis order
Alias: none
Argument(s): none
Default: false
This option activates a cross validation procedure to select the best basis order for representing the current set of data. It searches from start_order to max_order in increments of kick_order.
Default Behavior
No cross validation for basis order.
This example shows specification of an order adaptation starting at order 2, incrementing by 2, and limited at order 6.
model,
id_model = 'FT'
surrogate global function_train
start_rank = 4
adapt_order start_order = 2 kick_order = 2 max_order = 6
solver_tolerance = 1e-12
rounding_tolerance = 1e-12
dace_method_pointer = 'SAMPLING'
Note that adapt_rank and adapt_order can either be combined or used separately.
These keywords may also be of interest: