topic Re: How do I model A -> B -> C behavior? in Discussions

How do I model A -> B -> C behavior?

JasonM — Thu, 08 Jun 2023 21:03:28 GMT

Hello, I was wondering if anyone else has had this problem. We have a DoE custom design for evaluating factors to optimize for product "B" over an impurity, "C." C isn't independent of B - it's a byproduct which forms only upon B forming, that is A -> B-> C. I see how in Column Info one can set independent limits for responses A and B, but is there a way to set constraints such that B + C < 100%? Thank you!

Re: How do I model A -> B -> C behavior?

statman — Tue, 06 Apr 2021 19:37:11 GMT

Not sure I understand completely, but here are my thoughts. Seems you Have A > B + C? In any case, you can add factors as mixture factors and impose linear constraints.

Re: How do I model A -> B -> C behavior?

ih — Tue, 06 Apr 2021 19:43:31 GMT

It sounds like both B and C are responses, rather than being controlled directly. If so, then would their measured values always be less than or equal to 100%? It seems like you might not need to introduce that constraint, and you could just try to maximize B and minimize C.