What inspired this wish list request? I am working on a couple projects with 'messy' data and multiple responses. For these projects, there is not a single optimum solution, but a contour of acceptable solutions. One way to demonstrate this to clients is to use the Output random table in the Profiler to show a variety of different solutions. However, with this messy data, we have extrapolation concerns. It would be great to understand which of the solutions are possible extrapolations, which are not, and by how much.
What is the improvement you would like to see? With the Output random table from the Profiler, we have check boxes for adding noise and confidence and prediction intervals. It would be great to have a check box to add extrapolation details to the output table.
Why is this idea important? With this information in the output table, the analyst can quickly and easily (1) understand which of the acceptable solutions is a possible extrapolation and (2) visualize what part of the entire space is a possible extrapolation for each model. This allows the analyst more information to suggest experiments that would confirm the acceptable solution region and make the models stronger in those areas. In the end, we would like to provide the operators some freedom to run the equipment within this acceptable region.
I am using JMP Pro 19.