cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Discussions

Solve problems, and share tips and tricks with other JMP users.
Choose Language Hide Translation Bar
frankderuyck
Level VII

SVEM Prediction formula

I made a SVEM analysis and would like to use the formula to comupute process control ranges using the Design Space Profiler. I saved the prediction formula and  can't save a prediction formula with process parameters, only a Mean formula here below? 

frankderuyck_0-1764687174977.png

Is  it possible to get an expression as a function of the process parameters that can be used in the design space profiler? Or can I use the Maximum Likelihood expression reported together with SVEM in the GENREG report?

13 REPLIES 13
frankderuyck
Level VII

Re: SVEM Prediction formula

HI Victor, I tried several times cfr. below 

frankderuyck_0-1765177559885.png

unfortunately this does't work 

frankderuyck_1-1765177643504.png

Strange that, when clicking on relaunch, the expand box is emty? 

frankderuyck_2-1765177780810.png

 

Victor_G
Super User

Re: SVEM Prediction formula

Hi @frankderuyck,

Can you share your datatable or an anonymized version ? Which JMP version are you using ?

I don't understand what is not working. I reproduce the same steps on JMP dataset Powder Metallurgy with a Weibull SVEM Forward regression. I have only the option "Save Resample Formulas" like you have, and it gives me a Pred Formula similar to what you showed :

Victor_G_0-1765184114415.png

Do you have also the column property "Expand Intermediates" with the value set at 1 ?

Victor_G_3-1765184279532.png

 

When I launch the Profiler from the Graph menu with the "Expand Intermediate Formulas" :

Victor_G_1-1765184208211.png

I obtain a Profiler with the response linked only to the predictors :

Victor_G_2-1765184252177.png

 

You can check you have the same settings as I show here (on JMP version 18.1.2 for me), and if it's not helpful, can you share a dataset so that I can try to see what's not working ?

Best,

Victor GUILLER

"It is not unusual for a well-designed experiment to analyze itself" (Box, Hunter and Hunter)
frankderuyck
Level VII

Re: SVEM Prediction formula

Find aatached DOE with SVEM analysis

Victor_G
Super User

Re: SVEM Prediction formula

Ok, I understand better.
Some of the "Purity sample prediction formula" in the Profiler corresponds to constant values. These 24 sample predictions values are constant, so they don't have influence when you change factor settings.

Example with Purity Sample 11 Prediction Formula :

Victor_G_0-1765194274720.png

 

What I would suggest is to replace these constant values predictions formula by their values in the Pred Formula Purity SVEM formula, so that the Profiler only includes prediction formulas linked to factors. With the option "Expand Intermediate Formulas" checked with this simplified formula, it should work :

Victor_G_1-1765195856927.png

 

Please find attached your datatable with the added simplified SVEM formula column and a script with the final Profiler.

Victor GUILLER

"It is not unusual for a well-designed experiment to analyze itself" (Box, Hunter and Hunter)

Recommended Articles