How can I calculate area under the curve in jmp pro?

[#1]

Hello!

In a recent study, I collected blood samples at baseline (fasted) and four timepoints following consumption of one of two types of fiber (each subject received both types of fiber, but on separate days).  I analyzed the blood samples to determine the concentrations of various compounds at each time point.  I would like to calculate the area under the curve for each compound following both types of fiber, for each subject.  Is there an easy way to do this in jmp pro?  I am a novice and not experienced with using code...

Any advice would be very much appreciated- thanks!

[#2]

Please provide more details.  Are you wanting to calculate the AUC for an ROC curve?  It doesn't sound like that since your dependent variable appears to be continuous.  If it is indeed continuous, then you need to explain what you mean by the "area under the curve," specifically what curve are you referring to?  If you really mean the AUC, then JMP Pro provides this automatically for a number of platforms, but your dependent variable should be a nominal (categorical) variable.  From your description, it is not clear to me what you would be analyzing as a nominal dependent variable.

[#3]

Thanks for the valuable insight. I ended up using Prism instead of JMP and managed to get the AUC, so I diverted having to learn how to do this in JMP.  Sometimes it's easier to just use what you're used to!

[#4]

It would help other Community members who are following this discussion to understand what you were doing. We still do not know what you mean by a 'curve' in your data or your analysis. That way we can 'close the loop' with a solution for JMP users who do not use Prism.

[#5]

Mark:

Just so happens that I am trying to do the same thing.
FDE course > Capstone Study > Polymer Development example (attached below). I have molecular weight distribution data just like that. I want to know how much of the mass has MW > 10^6. Since I don't have a function, I need the numeric integral to answer the question. Is there some existing JMP platform that can do this? Or is it time to break out my numerical methods textbook and write some formulae...?

[#6]

Perhaps you won't have the 'crack open' that text book after all.

[#7]

Mark:

Thanks! Integrate() takes an expression as argument, f(x). I don't have an expression, just raw (x,y) data. I don't want to fit a smooth curve to it since I am really interested in the bumps in the distributions.

[#8]

I think that you have to resort to a summation. The Summation() function is useful for this purpose. It should work with rows in the data table or rows in a matrix. I can't think of anything else.