Is there a way in JMP to design an experiment with two equality constraints?
I am working on an experimental design that requires two Equality constraints.
There are 5 mixture variables.
Two Equality Constraints:
(x1) + (x2) + (x3) + (x4) + (x5) = 1
(x1)*14 + (x2)*14.4 + (x3)*13 + (x4)*9 + (x5)*12.5 = 13
I realize that part of the issue with two constraints is specifying three of the variables will resolve what the other two are.
I attempted to make a very tight inequality < and > inequality (>12.9 < 13.1); however, the JMP profiler results look very odd (CI don't show up).
Also, I made the inequalities equal <13 & >13. JMP warns there may be singularity issues. Then, when I created the design it only inserts 3/5 mixture variables. If I add in the other 2, I get the singularity issue and no p-values. This has something to do with the fact that when there are two ineqaulity constraints, only three variables are needed to solve for the other two. The dimensionality of the problem should decrease everytime a new equality constraint is specified.
I was also thinking about ways to reduce the two constraints into a single constraint, but was unsuccessful.
Also, is there a reason why JMP doesn't let you specify multiple equality constraints?
Thank you,
Nate
