How to input non-linear Effect restrictions into Prediction Profiler, JMP 14
Jul 18, 2019 8:59 AM(278 views)
I am attempting to use JMP 14 to predict optomized recipes to get a part with specific characteristics using a tool I am using (I'm aware that JMP Pro is probably better for this, I'm looking into it). The tool has certain restrictions. Let's say that I have X1, X2, X3 and X4. I know that X1 and X2 are related in certain ways that have effects on my outputs, so instead of putting both values into my JMP table, I've put X1/X2 and X1+X2 in. The same is true for X3 and X4 (I have them in as X3/X4 and X3+X4). My tool, however, can not allow X1 > 1000, X3 >1000 or X4 < 200. I know I'd have to input something like:
X1+X2 - ((X1+X2)/(1+X1/X2)) >/ 1000
X3+X4 - ((X3+X4)/(1+X3/X4)) >/ 1000
(X3+X4/(1+X3/X4)) </ 200
but I don't understand where to put this. I'm sure it's a very simple thing, but I can't really find anything about this type of case online. Please let me know how to do this, thanks!