Request for two new features: Random assignment and indirect effects

[HeikoBreitsohl]

Hi Johannes,

I have two suggestions for new features that may make DAGitty even more useful:

• Variable mode "randomly assigned" (or "randomly manipulated" or something). When applied to a variable, this would block all arrows into that variable. Having this feature would help convey the advantages of experiments. Currently, the only option to represent an experiment (as far as I know) is to manually delete arrows into the variable, which is not as elegant as DAGitty is otherwise.
• Causal effect identification for indirect effects. Many researchers are interested in indirect (mediated) effects, but they can be very tricky to identify. Being able to model this would be a massive improvement!

Thanks for the great work!
Heiko