Followup fix to MOO refactor

Summary:
In D28869199 (facebook@595095e), I changed the multi objective data model, so that instead of relying on objective.metric.lower_is_better to indicate directionality, we can instead use objective.minimize.

I forgot to update the validation we have that errors if the directionality of an objective threshold doesn't match the directionality of an objective. The validation is fine, but the way we determine the directionality of the objective has to change (i.e. check minimize now instead of lower_is_better).

Reviewed By: Balandat

Differential Revision: D29165965

fbshipit-source-id: 4b935f2a018da2666d487e21ae7b659559e42a3f

ax/core/optimization_config.py

@@ -364,25 +364,9 @@ def _validate_optimization_config(
         outcome_constraints = outcome_constraints or []
         objective_thresholds = objective_thresholds or []
 
-        # Verify we aren't optimizing too many objectives.
-        objective_metrics_by_name = {
-            metric.name: metric for metric in objective.metrics
-        }
-        # Warn if thresholds on objective_metrics bound from the wrong direction.
-        for threshold in objective_thresholds:
-            metric_name = threshold.metric.name
-            if metric_name in objective_metrics_by_name:
-                lower_is_better = threshold.metric.lower_is_better
-                bounded_above = threshold.op == ComparisonOp.LEQ
-                is_aligned = lower_is_better == bounded_above
-                if not (is_aligned or lower_is_better is None):
-                    raise ValueError(
-                        make_wrong_direction_warning(
-                            metric_name=metric_name,
-                            bounded_above=bounded_above,
-                            lower_is_better=lower_is_better,
-                        )
-                    )
+        maybe_raise_wrong_direction_warning(
+            objective=objective, objective_thresholds=objective_thresholds
+        )
 
         unconstrainable_metrics = objective.get_unconstrainable_metrics()
         OptimizationConfig._validate_outcome_constraints(
@@ -402,12 +386,26 @@ def __repr__(self) -> str:
         )
 
 
-def make_wrong_direction_warning(
-    metric_name: str, bounded_above: bool, lower_is_better: Optional[bool]
-) -> str:
-    return (
-        f"Constraint on {metric_name} bounds from "
-        f"{'above' if bounded_above else 'below'} "
-        f"but {metric_name} is being "
-        f"{'minimized' if lower_is_better else 'maximized'}."
-    ).format(metric_name)
+def maybe_raise_wrong_direction_warning(
+    objective: Objective,
+    objective_thresholds: List[ObjectiveThreshold],
+) -> None:
+    """Warn if thresholds on objective_metrics bound from the wrong direction."""
+    if not isinstance(objective, MultiObjective):
+        return
+
+    objectives_by_name = {obj.metric.name: obj for obj in objective.objectives}
+    for threshold in objective_thresholds:
+        metric_name = threshold.metric.name
+        if metric_name in objectives_by_name:
+            minimize = objectives_by_name[metric_name].minimize
+            bounded_above = threshold.op == ComparisonOp.LEQ
+            is_aligned = minimize == bounded_above
+            if not is_aligned:
+                message = (
+                    f"Constraint on {metric_name} bounds from "
+                    f"{'above' if bounded_above else 'below'} "
+                    f"but {metric_name} is being "
+                    f"{'minimized' if minimize else 'maximized'}."
+                ).format(metric_name)
+                raise ValueError(message)

ax/core/tests/test_optimization_config.py

@@ -30,7 +30,7 @@
 MOOC_STR = (
     "OptimizationConfig(objective=MultiObjective(objectives="
     '[Objective(metric_name="m1", minimize=True), '
-    'Objective(metric_name="m2", minimize=True)]), '
+    'Objective(metric_name="m2", minimize=False)]), '
     "outcome_constraints=[OutcomeConstraint(m2 >= -0.25%), "
     "OutcomeConstraint(m2 <= 0.25%)], objective_thresholds=[])"
 )
@@ -189,7 +189,7 @@ def setUp(self):
         }
         self.objectives = {
             "o1": Objective(metric=self.metrics["m1"]),
-            "o2": Objective(metric=self.metrics["m2"], minimize=True),
+            "o2": Objective(metric=self.metrics["m2"], minimize=False),
             "o3": Objective(metric=self.metrics["m3"], minimize=False),
         }
         self.objective = Objective(metric=self.metrics["m1"], minimize=False)

ax/modelbridge/tests/test_factory.py

@@ -266,12 +266,12 @@ def test_MOO_PAREGO(self):
     def test_MOO_EHVI(self):
         single_obj_exp = get_branin_experiment(with_batch=True)
         metrics = single_obj_exp.optimization_config.objective.metrics
-        metrics[0].lower_is_better = True
         objective_thresholds = [
             ObjectiveThreshold(
                 metric=metrics[0], bound=0.0, relative=False, op=ComparisonOp.GEQ
             )
         ]
+        # ValueError: Multi-objective optimization requires multiple objectives.
         with self.assertRaises(ValueError):
             get_MOO_EHVI(
                 experiment=single_obj_exp,
@@ -280,12 +280,15 @@ def test_MOO_EHVI(self):
             )
         multi_obj_exp = get_branin_experiment_with_multi_objective(with_batch=True)
         metrics = multi_obj_exp.optimization_config.objective.metrics
-        metrics[0].lower_is_better = False
-        metrics[1].lower_is_better = True
         multi_objective_thresholds = [
-            ObjectiveThreshold(metric=metrics[0], bound=0.0, relative=False),
-            ObjectiveThreshold(metric=metrics[1], bound=0.0, relative=False),
+            ObjectiveThreshold(
+                metric=metrics[0], bound=0.0, relative=False, op=ComparisonOp.GEQ
+            ),
+            ObjectiveThreshold(
+                metric=metrics[1], bound=0.0, relative=False, op=ComparisonOp.GEQ
+            ),
         ]
+        # ValueError: MultiObjectiveOptimization requires non-empty data.
         with self.assertRaises(ValueError):
             get_MOO_EHVI(
                 experiment=multi_obj_exp,

ax/utils/testing/core_stubs.py

@@ -848,7 +848,6 @@ def get_branin_multi_objective() -> Objective:
             Objective(metric=get_branin_metric(name="branin_a")),
             Objective(metric=get_branin_metric(name="branin_b")),
         ],
-        minimize=False,
     )