Demonstrate use of edge constraints in a causal discovery algorithm tutorial

[robertness]

Is your feature request related to a problem? Please describe. We need a tutorial that demonstrates the utility of constraints in causal discovery.

Describe the solution you'd like Update the notebook in this PR to work with constraints. Needs for this to be resolved first.

Describe alternatives you've considered I think we could have multiple notebooks to that address this, actually.

Additional context See links above.