Say there are two experiments, test/test/iris/1/iris and test/test/iris/1/means. The experiment metadata for each of these experiments will be found in the same folder, test/test/iris/1. Say that these experiments use the same analyzer as well.

If you plot these experiments , the last one to be plotted will overwrite the other experiment's plot files. This is because there is no per-experiment unique key added to the plot image file names - instead the file names are only based off of the output names in the analyzer.