-
From the Flow section of the action bar,
click DOE
, and drop it on the
process diagram.
-
Double-click DOE
.
The DOE Editor appears.
-
From the General tab's DOE Technique list, select Adaptive DOE.
- Enter the Number of Points that you want to study.
For each defined
factor, Optimization
Process Composer varies the values between the specified lower and upper
values to maximally fill the search space. - Enter the Significant Step Size to control the minimum distance between any
two points in the search space.
The size is a fraction of the
domain (upper bound minus lower bound) of each factor and must be a positive number. To find more subtle gradients, you should enter a small value, such as 0.001. To filter noise, you should enter a large value, such as 0.05.
- Optional: Select Use a fixed
random seed, and enter the seed.
If you do not select this option, Optimization
Process Composer determines the seed
randomly. Note:
You can also specify a fixed seed for the entire model. If you specify a seed for a DOE component, it takes precedence over the seed defined for the model.
- Optional: Select Read Prior Results to prescribe an arbitrary set of
previously run design points that the Adaptive DOE space-filling algorithm should avoid when
calculating the design matrix.
The points are read from a file in the working directory called DOE_Results.txt. - Optional: Select Write Current Results to save the current design points in DOE_Results.txt in the working directory. You can use the file to seed the next iteration of the Adaptive DOE.
- Click Ok to save your changes and to close the DOE Editor.
|