Adam Li

Perhaps it might make sense to also add a function that takes in a mixed-edge graph with directed and/or bidirected edges (e.g. an ADMG without undirected edges) and checks if...

This has been resolved. However, another issue I noticed is we should be able to handle drawing of purely undirected graphs.

Perhaps checking validity of a MAG(?). There are some graphical examples in the paper.

@robertness can you link a reference, or algorithm for this? If it's easy enough, it's a good issue for a new contributor.

> Just to list the procedure out: > > * Convert the PAG to MAG > * Check if the MAG is valid or not > * Convert the MAG...

Same skeleton as MAG so checking acyclicity, ancestral, and maximalist of the non-circle edge subgraph. A PAG could be fully oriented which results in no circle edges.

> > * Check if the converted and original PAGs are equivalent or not > > For this, I think I have to determine that the two PAGs have the...

Nice! Congrats. Can you remind me what part requires the acyclicity check?

Ah I see. Yeah it's just the same check as a MAG That's cuz we can't have cycles in the directed edges or almost directed cycles. You would only do...

Yep!