Optimal Options¶
Random starts: The number of times the selection process will repeat. The best design will be chosen from among the repeats.
Number of threads: The number of computation cores in the CPU that can be assigned a task by Stat-Ease 360. The default is to use one less than the number of cores on the local machine (or just one if the computer is that old).
Coordinate Exchange¶
Start distance: The maximum distance in coded units a point can be moved to attempt to improve the current design criterion.
Stop distance: The minimum distance in coded units a point will be moved.
If no improvements to the design criterion are found between the start and stop size for a given point, the algorithm with move to the next point or terminate if on the last point.
Point Exchange¶
Max Loops: The maximum number of iterations to improve from each random start.
Run until convergence: Checking this box allows the algorithm to run until no further improvement can be found. This can take a long time but does result in slightly more optimal designs.
Excursion size: The maximum number of points that will be swapped at a time to improve the design criterion. Higher numbers produce a better design, but take longer to create.
Click the Defaults button to reset the parameters above to the original settings.