Added functionality to translate a process tree into a Petri net#48
Conversation
| assert_eq!(4, net.places.len()); | ||
| assert_eq!(3, net.transitions.len()); | ||
| assert_eq!(6, net.arcs.len()); |
There was a problem hiding this comment.
Just checking number of places, transitions, and arcs feels weak. Could make it a bit more precise by looking at counts of silent/visible transitions, PlaceTransition/TransitionPlace arcs, etc.
Could look at pre/postsets of transitions (sizes and intersections with pre/postsets of other transitions) or identify certain places based on that (initial marking place has empty preset, then the connected transition is labelled "a", etc.)
Because of the Uuids, I'm not sure how to do an elegant, perfect, comparison without graph bisimulation or GED or something..
There was a problem hiding this comment.
Suggestion: We could create another Petri net class that represents each place by the set of its ingoing and outgoing transition labels. Like the eST-Miner does. What is your opinion on that, @cpitsch? It may not be fully sufficient for any Petri net comparison, but maybe it's a start and also useful for other projects?
- Fixed minor errors - simplified code - improved documentation Co-Authored-By: cpitsch <43351758+cpitsch@users.noreply.github.com>
…rs such that the comment references are enabled again
2 participants
No description provided.