OptaPyException: A problem occurred when wrapping the python problem

[DavidGarHeredia]

Hi!

I was trying to solve a school timetabling problem using Optapy, but I got the following error:

solution = solver.solve(problem_to_solve)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/d-future/miniconda3/envs/Minerva/lib/python3.11/site-packages/optapy/optaplanner_api_wrappers.py", 
line 416, in solve
    wrapped_problem = PythonSolver.wrapProblem(object_class, problem)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
org.optaplanner.optapy.org.optaplanner.optapy.OptaPyException: org.optaplanner.optapy.OptaPyException: 
A problem occurred when wrapping the python problem
(<jpype._jproxy.proxy.OptaPlannerSchedule object at 0x7ff372e86440>). 
Maybe an annotation was passed an incorrect type (for example, @problem_fact_collection_property(str)  
on a function that return a list of int).

I have reviewed the code multiple times to find an error like the one described in the msg but I did not succeed. Could you have a quick look at the code below to see if I am missing something? From the error msg, it seems to be something trivial, but it keeps eluding me :S

Thanks in advance!

@dataclass(frozen=True)
class Slot:
    id: int
    t0: datetime
    tf: datetime
    duration: timedelta = field(init=False)

    def __post_init__(self):
        object.__setattr__(self, "duration", self.tf - self.t0)


@problem_fact
class SlotFact:
    def __init__(self, slot: Slot):
        self.slot = slot

    @planning_id
    def get_id(self) -> int:
        return self.slot.id

    @property
    def duration(self):
        return self.slot.duration


@planning_entity
class AssignmentEntity:
    def __init__(
        self,
        id: int,
        teacher: OptaTeacher,
        group: OptaGroup,
        subject: Subject,
        room: Classroom,
        slot: Optional[SlotFact] = None,
    ) -> None:
        self.id = id
        self.teacher = teacher
        self.group = group
        self.subject = subject
        self.room = room
        self.slot = slot

    @planning_id
    def get_id(self):
        return self.id

    @planning_variable(SlotFact, ["SlotRange"])
    def get_slot(self):
        return self.slot

    def set_slot(self, new_slot: SlotFact):
        self.slot = new_slot

@planning_solution
class OptaPlannerSchedule:
    def __init__(
        self, slots: list[SlotFact], assignments: list[AssignmentEntity], score=None
    ):
        self.slots = slots
        self.assignments = assignments
        self.score = score

    @problem_fact_collection_property(SlotFact)
    @value_range_provider("SlotRange")
    def get_slots(self) -> list[SlotFact]:
        return self.slots

    @planning_entity_collection_property(AssignmentEntity)
    def get_assignments(self) -> list[AssignmentEntity]:
        return self.assignments

    @planning_score(HardSoftScore)
    def get_score(self):
        return self.score

    def set_score(self, score):
        self.score = score

[mmmvvvppp]

There's a note at the bottom of section 3.3.1 that warns:

Planning entity __hash__(self) implementations must remain constant. Therefore entity __hash__(self) must not depend on any planning variables. Pay special attention when using data structures with auto-generated __hash__(self) as entities, such as Python @dataclass.

Perhaps this is an origin of your problem?

[DavidGarHeredia]

Hey! Thanks for the response!

I tried defining a __hash__(self) method for the @planning_entity class, but still not working. Actually, I think the error may have a different nature as @planning_entity is defined as a class, not as @dataclass.

According to the error msg, the issue seems to happen when wrapping the @planning_solution class. However, given your warning about the usage of @dataclass, I wonder if there may be any other side effects when using these with Optapy? I will try to investigate in that direction.

In the meanwhile, if you come up with any other ideas, just let me know, please.

Thanks a lot!