DOE Scripting 2: How to set up a full/fractional factorial design

[#1]

Hi & Sorry for bothering you again.

Somehow I can't find a way to create a full-factorial design for a variable number of factors (k). Same for fractional factorial design. I know how to do it if I know the number of factors, as I can use

full_factorial_design = DOE(Screening Design);

full_factorial_design << add factor;

full_factorial_design << add factor;

full_factorial_design << add factor;

full_factorial_design << add factor;

full_factorial_design << Make Design(4);

But this works only if i want to select the 4th design in the design list. This would not be the full-factorial design if I had 6 factors.

Is there any documentation about the scripting functions for the DOE-platforms in JMP? I could not find anything in the Scripting Guide or the DOE-pdf. Sadly the Scripting-Index isn't very helpful either.

Thanks for all hints!

Sebastian

[#2]

Continuing the way that you started, you could iterate for the necessary number of factors.

n = 4;

full_factorial_design = DOE(Screening Design);

For( f = 1, f <=n, f++, full_factorial_design << add factor);

full_factorial_design << Make Design(n);

[#3]

Hi Mark,

thanks for the reply. Sadly this does not solve my problem as Make Design(4) will be a full-factorial design for 4 factors (as it is the 4th design in the design list) but Make Design(5) is not the full-factorial design for 5 factors.

[#4 · Accepted solution]

If you want a full factorial design, you can use the Full Factorial designer:

DOE(
  Full Factorial Design,
  {Add Response( Maximize, "Y", ., ., . ),
  Add Factor( Continuous, {-1, 1}, "X1", 0 ),
  Add Factor( Continuous, {-1, 1}, "X2", 0 ),
  Add Factor( Continuous, {-1, 1}, "X3", 0 ),
  Add Factor( Continuous, {-1, 1}, "X4", 0 ),
  Add Factor( Continuous, {-1, 1}, "X5", 0 ),
  Make Design}
);

I don't see an obvious way to get the contents of the design choices from the Screening designer, so that you can see what each number design is.

-Jeff

[#5]

Thanks a lot! That does the trick for the full factorials.

Is there any documentation on the different designs? How would I do the same thing for fractional factorials, etc? How do I select a fractional design of a specific resolution? Usually I'm good at figuring this kind of stuff out on my own, but I couldn't find any documentation on that.

Kind regards,

Sebastian