Hi,
I would like to use CP Optimizer’s conflict refiner with a distinction between hard constraints and soft/member constraints.
In the CPLEX MP API, this is possible through conflict refiner groups/preferences, for example with docplex.mp.conflict_refiner.ConstraintsGroup(preference=..., cts=[...]). In particular, a preference of 0.0 can be used for constraints that should be treated as required/hard.
For CP Optimizer, the underlying Concert API appears to support conflict refinement with constraint preferences, but I could not find an equivalent exposed in the Python docplex.cp API. Could docplex.cp expose a way to pass conflict refiner preferences to CpoModel.refine_conflict()?
Kind regards
Orestis
Hi,
I would like to use CP Optimizer’s conflict refiner with a distinction between hard constraints and soft/member constraints.
In the CPLEX MP API, this is possible through conflict refiner groups/preferences, for example with
docplex.mp.conflict_refiner.ConstraintsGroup(preference=..., cts=[...]). In particular, a preference of0.0can be used for constraints that should be treated as required/hard.For CP Optimizer, the underlying Concert API appears to support conflict refinement with constraint preferences, but I could not find an equivalent exposed in the Python docplex.cp API. Could
docplex.cpexpose a way to pass conflict refiner preferences toCpoModel.refine_conflict()?Kind regards
Orestis