Measure

This module defines evaluating measure of hardness and allows to set various limits. Both general methods of measure module uses in Solver implementations.

class Measure:
    def get_budget() -> Tuple[Optional[str], Optional[float]]:
    def check_and_get(stats: Dict[str, float], status: Optional[bool]) -> Tuple[Optional[float], Status]

The get_budget method returns a string identifier of the selected measure with the value of budget argument passed during initialization, as a tuple.

The check_and_get method selects the corresponding value of the selected measure from the passed stats dictionary and returns a tuple of this value and the option of Status enumerable type, which determines the resulting status of the processed task.

You can also specify common arguments for all implementations:

• budget – the maximum number of resources in the selected measure, that can be spent to solve one task instance.

• at_least – the minimum number of resources in the selected measure, that must be spent to solve one task instance.

Solving time

from function.module.measure import SolvingTime

measure = SolvingTime(
    budget: Optional[float]
    at_least: Optional[float]
)

Conflicts

from function.module.measure import Conflicts

measure = Conflicts(
    budget: Optional[float]
    at_least: Optional[float]
)

Propagations

from function.module.measure import Propagations

measure = Propagations(
    budget: Optional[float]
    at_least: Optional[float]
)

Learned literals

from function.module.measure import LearnedLiterals

measure = LearnedLiterals(
    at_least: Optional[float]
)

Note

The budget argument of Learned literals measure doesn’t support in available solvers!