Optimize

Concept

The Optimize pragmatic extends constraint solving. Instead of just finding any valid assignment, it searches for the assignment that produces the highest or lowest value for a given objective function (an attribute or an aggregation).

Syntax

Maximize <Objective> [Subject to <Constraints>].
Minimize <Objective> [Subject to <Constraints>].

Examples

// Direct Attribute Optimization
Maximize the price of ?product
Subject to:
    ?product is a Laptop.
    ?product is in-stock.

// Aggregation Optimization
Minimize the total weight of ?route
Subject to:
    ?route is a DeliveryPath.
    ?route is active.

Objective Functions

The objective must evaluate to a numeric value:

• Direct Attribute: Maximize the capacity of ?server. (Requires a variable binding).
• Aggregation: Minimize the sum of weight of ?items. (Supports total of, sum of, average of, number of).

Execution Logic

1. Constraint Resolution: The `Subject to` clause is converted into a constraint satisfaction problem (same as Solve).
2. Evaluation: For each valid variable assignment, the objective function is computed.
3. Selection: The system tracks and returns the solution(s) with the optimal score.

Reference

Defined in DS23 - Optimization.