Package ai.timefold.solver.core.api.score.stream.common

Interface LoadBalance<Balanced_>

Type Parameters:

Balanced_ - type of the item being balanced

All Known Implementing Classes:

LoadBalanceImpl

public interface LoadBalance<Balanced_>

Calculates the unfairness measure for a particular input. It is the result of applying the load-balancing constraint collector.

Method Summary

Modifier and Type	Method	Description
@NonNull Map<Balanced_,Long>	loads()	Returns the items being balanced, along with their total load.
@NonNull BigDecimal	unfairness()	The unfairness measure describes how fairly the load is distributed over the items; the higher the number, the higher the imbalance.

Method Details

loads

@NonNull Map<Balanced_,Long> loads()

Returns the items being balanced, along with their total load. The iteration order of the map is undefined. For use in justifications, create a defensive copy of the map; the map itself is mutable and will be mutated by the constraint collector.

unfairness

@NonNull BigDecimal unfairness()

The unfairness measure describes how fairly the load is distributed over the items; the higher the number, the higher the imbalance. When zero, the load is perfectly balanced.

Unfairness is a dimensionless number which is solution-specific. Comparing unfairness between solutions of different input problems is not helpful. Only compare unfairness measures of solutions which have the same set of balanced items as input.

Returns:

never negative, six decimal places