"Consistent" extension of PAGs?

[robertness]

Is your feature request related to a problem? Please describe.
There is an algorithm that takes a PDAG and returns a DAG that is a member of the equivalence class. The algorithm returns a "consistent extension" which might have some theoretical properties I'm not aware of. I'm not sure if such a "consistent extension" exists for PAGs, but it does seem useful to be able to get an ancestral graph member of the equivalence class the PAG represents.

Describe the solution you'd like
An algorithm that takes a PAG and returns a MAG.

Describe alternatives you've considered
Or just a ancestral graph in the equivalence class, may doesn't need to be maximal.

[aryan26roy]

@robertness doesn't #92 do exactly what you are asking for here?