Design space for 4 variables

[#1]

We designed an RSM model for the optimization of 4 variables that are used as a combination. Once the results are obtained how do we obtain the design space (sweet spot) for all 4 variables together? 

Please help.

Many Thanks!

[#2]

Hi @Amp,

Welcome in the Community !

When you have collected the results, you can then fit a model with the "Fit Model" platform, and once you have obtained a satisfactory model that explains results variation from a statistical and domain points of view with your factors and different effects, you can use the Design Space Profiler, accessible in the options from the red triangle of the Profiler. This will enable to identify an experimental area respecting the different specifications and objectives (upper/lower, or both) in your responses.

I hope this info will help you,

[#3]

Hi Victor,

Thanks a lot for your answer. I was able to obtain the design space profiler. However, what I am looking for is a design space for all the 4 variables combined. These are mixture variables and I need to find the ideal range of each variable when used in combination to maximize the output.

Many Thanks!

[#4]

Hi @Amp,

In the link I have provided, if you read the notes :

"The Design Space Profiler is not designed to handle mixture factors".

I would recommend using the Mixture Profiler. If you specify USL and LSL for your responses, you'll be able to determine the experimental area available, and the corresponding range/ratio of mixture factors.

[#5]

I'm a bit confused, so my response may make little or no sense.  I thought from your first post you ran some sort of RSM for 4 factors.  In your last post, you now say it was a mixture design.  Two different "animals".  If you ran a mixture design, you should be analyzing the data using the mixture profiler.  

https://www.jmp.com/support/help/en/17.2/?os=mac&source=application#page/jmp/mixture-designs.shtml%2...

JMP will give you a contour map for the response in the space created by your mixture design.