Subscribe Bookmark RSS Feed

How do I plot "spiderdiagrams" in JMP


Community Member


Jan 10, 2018


I have a large geochemical dataset that I am attempting to plot in JMP13. I use JMP for most of my geochemical plots but I have yet to figure out how to do so-called "spiderdiagrams." Following the advice of another post I found, I transposed my data table and was able to plot things that looked like spiderdiagrams using the graph overlay tool. This allowed me to plot the data and would work fine as a solution if I had a small dataset. Using this method, everything is grouped by element rather than sample ID so when I click on one point I select all of the points of that element group (i.e. if I click on one Au datapoint, all of the Au datapoints are selected rather than the points that correspond to one sample). Ideally, I would like to be able to select all of the datapoints for a given sample (i.e. Pt-Pd-Rh-Ru-Au points) so that I can begin to divide all of the samples into subgroups.



 Please, does anyone have any suggestions for how I can group the connected points so I can begin to sort them into lines of similar slope and appearance.


Thank you!




Super User


Jun 23, 2011

You can try clustering the data using the approach outlined here (thanks Jeff P and David B!)





Jul 2, 2014

I recently was working with a similar type data set and I used the Parallel Plot in Graph Builder to achieve a plot like the one you show below.  In Graph Builder highlight the columns of interest along with sample ID and drag them to the X axis drop zone.  From there select the Parallel Plot icon on the upper right side of the icons list.  This should generate the plot.  From there go back to your data table and select Rows > Data Filter.  Add sample ID to filter on.  When the filter option comes up make sure you have "Select" checked.  Now you have to make sure both your Data Filter and Graph Builder are visible together.  You should be able to put them side-by-side in the same window.  Go to your list of sample ID's in the Data Filter and click on the one you want to see.  You should have a single line in your plot showing with all others grayed out.  I tried local data filter in Graph Builder but could not get the single sample view I was hoping for.  Maybe someone else out there has a better idea there.

If you only want to see the one line make sure to check "Show" and "Include" in the  data filter as well.


@pmroz made a great suggestion for clustering as well that will help you see "like" samples.  You can then drop the cluster variable into Group X or Group Y to further break things out.