What inspired this wish list request? Occasionally I come across a data distribution that I need to calculate a spec limit for and the non-parametric fit is the only one that works well. I can fit this distribution a couple of different ways but only the Legacy Fitters in the Distribution platform under Smooth Curve allow using a K-Sigma Multiplier to find the appropriate spec limits for a desired Ppk.
This fir is not as polished and I'm sure at some point will go away.
What is the improvement you would like to see?
Please add the Non-Parametric fit to the options under "Continuous Fits" in the distribution platform.
Why is this idea important?
Fitting distributions is critical for Process Capabilities and reverse engineering spec limits for Producibility Proposals for customers so having additional fits to handle less common situations add robustness in JMP's capability and more opportunities for additional user value generation.