Implement goals

ttdlgc/main.py

@@ -8,7 +8,7 @@
 import typer
 
 from .events import Event, Solution
-from .milp import create_milp, extract_solution, Constraint
+from .milp import create_milp, extract_solution, Constraint, Goal
 from .simulation import Simulation
 
 SUCCESS = 0
@@ -28,7 +28,12 @@ def solve(
         typer.Option(help="Filepath to write the generated choices to."),
     ] = None,
     constraint: Annotated[
-        Optional[list[Constraint]], typer.Option(help="Constraints to add to the MILP.")
+        Optional[list[Constraint]],
+        typer.Option(help="Hard constraints to add to the MILP."),
+    ] = None,
+    goal: Annotated[
+        Optional[Goal],
+        typer.Option(help="Goal/preference to have the solver optimize towards."),
     ] = None,
     verbose: bool = False,
 ) -> None:
@@ -42,7 +47,7 @@ def solve(
     if constraint is None:
         constraint = []
 
-    problem = create_milp(events, constraint)
+    problem = create_milp(events, constraint, goal)
     logger.debug(problem)
 
     solver = pulp.PULP_CBC_CMD(msg=0)

ttdlgc/milp.py

@@ -40,6 +40,12 @@ class Constraint(enum.Enum):
     ChaosLv5AtEnd = "ChaosLv5AtEnd"
 
 
+class Goal(enum.Enum):
+    MaximizeLaw = "MaximizeLaw"
+    MaximizeGrey = "MaximizeGrey"
+    MaximizeChaos = "MaximizeChaos"
+
+
 def extract_solution(events: list[Event], problem: pulp.LpProblem) -> Solution:
     variables = problem.variablesDict()
 
@@ -72,7 +78,7 @@ def extract_solution(events: list[Event], problem: pulp.LpProblem) -> Solution:
 
 
 def create_milp(
-    events: list[Event], constraints: Iterable[Constraint]
+    events: list[Event], constraints: Iterable[Constraint], goal: Optional[Goal]
 ) -> pulp.LpProblem:
     problem = pulp.LpProblem("LGC_Alignment", pulp.LpMaximize)
 
@@ -95,9 +101,14 @@ def create_milp(
     chaos_chapter_impacts: ImpactsDict = collections.defaultdict(lambda: [])
 
     # Objective function
-    problem += law_chapter_variables[-1], "Maximize law alignment"
-    # problem += grey_chapter_variables[-1], "Maximize grey alignment"
-    # problem += chaos_chapter_variables[-1], "Maximize chaos alignment"
+    if goal is None:
+        problem += 0 == 0, "No goal"
+    elif goal == Goal.MaximizeLaw:
+        problem += law_chapter_variables[-1], "Maximize law alignment"
+    elif goal == Goal.MaximizeGrey:
+        problem += grey_chapter_variables[-1], "Maximize grey alignment"
+    elif goal == Goal.MaximizeChaos:
+        problem += chaos_chapter_variables[-1], "Maximize chaos alignment"
 
     # Chapter 5 route
     law_route = pulp.LpVariable("chapter_5_route_law", cat=pulp.const.LpBinary)