Closed ArcticG closed 4 years ago
I'm glad your finding it useful @ArcticG. Have you tried using set.seed(some number)
? It should make everything reproducible. Unfortunately, it's not otherwise possible to control the output since the results are based on optimization methods.
Thank you for your reply. You are correct and setting seed solves the reproducibility issue. I simply tried a sequence of ten seeds and picked the one that produces the correct diagram.
As a side suggestion, it might be helpful if the package could notify/warn the user if all the overlapping areas are not included in the generated diagram. Although I do not know if it would be feasible/possible behind_the_scenes or worth the effort.
Thank you again for the package and your reply and happy new year!
Dear John Larsson, Thank you for your help with eulerr package. The package is really useful and has been doing great with venn diagrams and 3 way euler diagrams. I have noticed an issue with more complex euler diagrams in my hands, leading to reproducibility issue.
For example: After running the same code 6 times I obtained somewhat different euler diagrams with the same data. The resulting plots are shown as E1-E6 in below link with E3/E1 being the closest to the graph I was hoping for, but graphs like E2 being the most recurring ones:
https://ibb.co/DgcrK3H
Notes:
I wonder if there is a way/option to make the results more consistent/reproducible for more complex euler diagrams.
I would be happy to answer any questions about system/setting used or send the raw data via email if it would be helpful. Thank you