cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
  • New to JMP? Join us Sept. 23-24 for the Early User Edition of Discovery Summit, tailor-made for new users. Register now for free!
  • Use World Cup data to build models, explore spatial relationships, and create informative visualizations in JMP. Register. July 17, 2 pm US Eastern Time.
  • Your voice matters! Tell us how you prefer to receive JMP updates, so we can tailor our communication to your needs. Take short survey.

JMP Wish List

We want to hear your ideas for improving JMP. Share them here.
Choose Language Hide Translation Bar
0 Kudos

Allow Interval Formulas to be Saved with Nested Models

What inspired this wish list request? 

If I fit a nested model in JMP using the Fit Model platform, the Prediction Profiler shows accurate predictions along with confidence intervals. However, if I try to use the Save Columns -> Prediction and Interval Formulas option so that I can recreate the results manually in the Graph -> Profiler menu, it is not allowed with an error message (Std Error Formula does not support Nested Terms). In JMP 19.1 this actually causes JMP to crash, but has been fixed in more recent versions based on support I received from @PatrickGiuliano

 

What is the improvement you would like to see? 

Allow the Interval Formulas to be saved alongside the Prediction Formula so that the Graph -> Profiler menu can be used to show the exact same output as the original prediction profiler in Fit Model.

 

Why is this idea important?  

It would be great to have it save so that I can combine that model with the appropriate confidence intervals with models fit in other platforms to overlay those fits in a more general Profiler menu output. I was recently unable to do this for an example where for Y1 I had a nested model and Y2 I had a slightly different model parameterization and I was unable to get a combined Profiler with the confidence intervals with Y1 & Y2 together. It's clear from the initial model fit that the results are easily estimable across the design space but just needs to be more flexible.