You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
You have three structures Problem one in the packageexplain, one in the package maxsat and one in the package solver.
However, It seems that it could be convenient to first solve a problem and if the result is unsatisfiable, then find an explanation (MUS or just an unsatisfiable subset).
A simple (but inefficient) way is to do it as follows:
But having only one type of structure for both packages seems more convenient in my opinion.
Same principle with the problem structure for MaxSAT problem.
If you accept this "improvement", I will propose a PR for it.
That's a very good idea! That kind of thing was on my TODO-list anyway.
There's one potential pitfal you must be aware: in the solver package, problems are simplified as soon as they are parsed (meaning the problem (a or b or c) and not a will be transmitted as b or c to the solver).
That's not something you want when searching for MUSes or for MAXSAT problems though, so there are probably a few things to modify here and there in the solver package.
You have three structures
Problem
one in the packageexplain
, one in the packagemaxsat
and one in the packagesolver
.However, It seems that it could be convenient to first solve a problem and if the result is unsatisfiable, then find an explanation (MUS or just an unsatisfiable subset).
A simple (but inefficient) way is to do it as follows:
But having only one type of structure for both packages seems more convenient in my opinion.
Same principle with the problem structure for MaxSAT problem.
If you accept this "improvement", I will propose a PR for it.
Keep me informed @fdelorme
The text was updated successfully, but these errors were encountered: