Good afternoon all; I'm pretty new to JMP and am most familiar with using it for multivariate (particularly canonical discriminant) analysis of geochemical data. I'm running into an obstacle to visualizing some data as a parallel plot. I have a dataset containing about 45 rock samples for each of which I measured the concentration of about 30 elements. The samples come from three different rock formations. I have already run discriminant analysis to determine which elements are best at separating the three rock formations but I want to try and visualize this as a parallel plot. I have tried graphing all of the samples and elements on a single parallel plot (elements on x, element concentrations on y, each line color coded by which formation the sample belongs to), or one three graphs (one for each rock formation) but this is still a bit cluttered looking.
What I would like to do is create a single parallel plot that shows the mean value for each element for each rock formation (with rock formations color coded), with a standard deviation bar for each of those points. So, each element on the x axis would have three points, a mean for that element for each rock formation. This would help visualize for my collaborators which elements overlap in concentration across formations and are thus not useful for distinguishing among the formations. Is there a straightforward way to do this in JMP 12?
Many thanks!