Request for two new features: Random assignment and indirect effects

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

jtextor / dagitty