What is the easiest way in JMP to arrange an array of boxplots of different variables?
Jun 18, 2019 1:37 AM(692 views)
I'm analyzing metabolomics data on a regular basis. My data tables typically consist of hundreds of variables (i.e., analyte signal intensities, in columns). The rows contain the observations which belong to different groups, for example fresh vs. stored. One frequent task is to create an array of boxplots, let's say 16 bopxlots in a 4 by 4 fashion, with the y axes being scaled individually. Using graph builder, I can easily arrange several boxplots in column-wise manner but, as far as I can see, creating an array of boxplots is not easily possible, at least not without a complicated rearrangement of the data table, the introduction of dummy variables and using the "group by X" function
This is what I have:
This is what I need (here arranged manually using a photo editor):
Is there a smart and quick way to solve this problem?
Since JMP can generate code, it's always a good strategy to do something 'by hand' and see what you get (even if you know JSL).
So, in this case, you could build the left-hand set interactively in one window (using Graph Builder) then do the same for the right-hand set in another. Then use the Combine Windows menu option. This will actually make a JMP Dashboard. This may be enough, but even if not, you will have most of the code you need.