How to change the display of optimization plot in Minitab 18

#1
Hi, Could someone kindly advise how to change the display of optimization plot?
I run a full DOE of 2 factors, each with 3 levels (all had set as numerical), and every treatment has 10 replicates (i.e. total 3x3x10=90 data).
I use minitab 18, and the first way I followed: Stat->DOE->Factorial; after "Analyze Factorial Design", ->Response Optimizer->the optimization plot shows "points", rather than the "lines" as expected. How can I change the display to "lines"?
The second way I tried: Stat->Regression->Regression->Fit Regression Model->Response Optimizer
Under this, I can change the display of optimization plot as long as I define the factor as "Continuous predictors" or "Categorical predictors".
The former display "lines" plot, and the later "points".
I know above two approached apply for different models, just would like to know how to change the display under DOE. Thanks!

Regards,
Magicstars
 

Miner

Forum Moderator
Staff member
Admin
#2
Since you stated that you used 3 levels, I presume that you must have used the General Full Factorial to create your design. The response optimizer for General Full Factorial is limited to points. This is because with more than 2 levels, it is possible that the model is not linear, yet the model makes no attempt to fit a nonlinear model. it simply tells you whether the factors are significant or not. Therefore, it can only optimize using the best combination of the levels that you actually ran. It cannot predict "in between" conditions.
 
#3
Since you stated that you used 3 levels, I presume that you must have used the General Full Factorial to create your design. The response optimizer for General Full Factorial is limited to points. This is because with more than 2 levels, it is possible that the model is not linear, yet the model makes no attempt to fit a nonlinear model. it simply tells you whether the factors are significant or not. Therefore, it can only optimize using the best combination of the levels that you actually ran. It cannot predict "in between" conditions.
Hi, Miner,
Unbelievable! Just one day I removed the puzzle out of my mind. Thanks so much!
Elsmar Cove is really a good place to come.
Regards,
Magicstars
 


Top Bottom