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
Alex83
Level I

Issue with disallowed combinations filters in DOE

Dear Community,

I have two factors for which I want disallowed combination i a Custom DOE.

If the "Accel step 2 = 0  , there can't be any "speed from step 1"

if the Accel step 2 > 0, there cannot be any "speed from step 1" at 0.

I tried to translate like this :

Accel step 2 == 0 & Relative speed from step 1 > 0

| Accel step 2 > 0 &Relative speed from step 1 == 0

The DOE generated is indeed removing these two conditions but also removing other conditions that I need in the DOE, like : 

Accel step 2 = 0 and speed from step 1 >0

what is the reason of it and what could I do to solve it ?

Thank you so much for your support.

 

 

 

4 REPLIES 4
Victor_G
Super User

Re: Issue with disallowed combinations filters in DOE

Hello @Alex83,

Welcome in the Community !

Without more details about the factors type, ranges, any constraints you may have between other factors, assumed model, etc..., it's hard to help you. Can you provide your design table so that we can use the same factors settings and see what doesn't work ?

Best,

Victor GUILLER

"It is not unusual for a well-designed experiment to analyze itself" (Box, Hunter and Hunter)
Alex83
Level I

Re: Issue with disallowed combinations filters in DOE

Dear Victor, 

 

Thank you for your feedback and your support !

Here is my config : 

factors are continuous and start from zero.

These factor are not independent:  if acceleration is zero, the speed cannot be bigger than zero. 

same if the speed is bigger than 0, acceleration cannot be bigger than 0.

However, acceleration and speed can be both 0.

Alex83_0-1764916615043.png

 

I have put the DOE enclosed.

Best Regards,

Alex

 

Victor_G
Super User

Re: Issue with disallowed combinations filters in DOE

Hi @Alex83,

I don't know your topic or system you want to study with this DoE, but looking at the factor I have some questions.
Do you really need all these factors and can they really be changed independantly ? I ask this question because I see in your DOE that you have  acceleration, speed (or relative speed) and time at step 1 and 2 in your factors. I would think "naively" that acceleration is linked to the speed and time at each of these steps, so can you really change each of these factors independantly (apart from the constraint you want to include) ?

In your design, your factors do not start all at 0. Since the low level of speed step 1 is different from 0, you can't create the disallowed combination "If the "Accel step 2 = 0  , there can't be any speed from step 1". Changing the low levels of your speed and acceleration factors to 0 may help define the constraint. 

I think the easiest way would be to use the Candidate set approach and use it in the Custom design : Candidate Set Designs: Tailoring DOE Constraints to the Problem (2021-EU-30MP-784) 
Start by creating the full table of possible combinations between your factors, and then use the option "Select Covariate Factors" in Custom Design platform to select the factors from your candidate set table. The design generated will only include runs available from your candidate set table: if you have enforced your constraints in this table by only allowing/including runs respecting this constraint, then the resulting design will also respect your constraints.

You may also read similar topics related to disallowed combinations :
Using disallowed combinations to remove a factor from the DoE 
Adding a complex constrain to a Custom Design with 4 factors 

Hope this answer will help you,

Victor GUILLER

"It is not unusual for a well-designed experiment to analyze itself" (Box, Hunter and Hunter)
Alex83
Level I

Re: Issue with disallowed combinations filters in DOE

Dear Victor,

Thank you for your precious inputs. True, we were about to remove the time, as accel and speed are enough to define what we want.

What do you mean by "In your design, your factors do not start all at 0 ? I set the factors starts at 0, as you can see in the picture enclosed.

Alex83_0-1764950040608.png

I will try the candidate set approach by creating the table first, and see if it solve the issues.

Best Regards,

Alex

Recommended Articles