I have a multiple regression model with two predictors, where I wanted to visualize the interrelationship using the explained variances (partial, semi-partial and total I think it would have to be, although I am not the most savvy when it comes to regressions). I have been trying some packages (e.g. VennDiagram), but they all seem to only be able to do Venn diagrams based on counts of values, not variances.
Does anyone have a good idea on how to do this in R or elsewhere? Preferrably I would like the ares to adapt to the sizes of the explained variances.
Thanks in advance!
You are in luck, I was discussing it today. I give two options: (1.) via the modEvA package in R and one by hand (manual).
P.S. for simplicity I ignore all assumptions of the models.
I think if you plot them by yourself, it would be more proportional and beautiful. If you consider the size of the variances while plotting the circles, you obtain a more meaningful tool to illustrate the findings. Good luck. Babak Jamshidi
I tried doing one manually using https://app.diagrams.net. Not entirely pleased, but something like this is what I was looking and hoping for. Problem is, I have only guessed at the areas, not sure how to get at the proper proportions.
- Badges
- Science topic
- Similar topics
- Force