Convert InstanceFeatures into dataclass

miplearn/components/lazy_static.py

@@ -64,7 +64,7 @@ class StaticLazyConstraintsComponent(Component):
         assert features.instance is not None
         assert features.constraints is not None
 
-        if not features.instance["Lazy constraint count"] == 0:
+        if not features.instance.lazy_constraint_count == 0:
             logger.info("Instance does not have static lazy constraints. Skipping.")
         logger.info("Predicting required lazy constraints...")
         self.enforced_cids = set(self.sample_predict(features, training_data))

miplearn/components/objective.py

@@ -80,7 +80,7 @@ class ObjectiveValueComponent(Component):
         assert features.instance is not None
         x: Dict[Hashable, List[List[float]]] = {}
         y: Dict[Hashable, List[List[float]]] = {}
-        f = list(features.instance["User features"])
+        f = list(features.instance.user_features)
         if "LP value" in sample and sample["LP value"] is not None:
             f += [sample["LP value"]]
         for c in ["Upper bound", "Lower bound"]:

miplearn/features.py

@@ -120,7 +120,7 @@ class FeaturesExtractor:
         for (cid, cdict) in features.constraints.items():
             if cdict["Lazy"]:
                 lazy_count += 1
-        return {
+        return InstanceFeatures(
-            "User features": user_features,
+            user_features=user_features,
-            "Lazy constraint count": lazy_count,
+            lazy_constraint_count=lazy_count,
-        }
+        )

miplearn/types.py

@@ -78,14 +78,11 @@ LearningSolveStats = TypedDict(
     total=False,
 )
 
-InstanceFeatures = TypedDict(
+
-    "InstanceFeatures",
+@dataclass
-    {
+class InstanceFeatures:
-        "User features": List[float],
+    user_features: List[float]
-        "Lazy constraint count": int,
+    lazy_constraint_count: int = 0
-    },
-    total=False,
-)
 
 
 @dataclass

tests/components/test_lazy_static.py

@@ -12,7 +12,12 @@ from miplearn import LearningSolver, InternalSolver, Instance
 from miplearn.classifiers import Classifier
 from miplearn.classifiers.threshold import Threshold, MinProbabilityThreshold
 from miplearn.components.lazy_static import StaticLazyConstraintsComponent
-from miplearn.types import TrainingSample, Features, LearningSolveStats
+from miplearn.types import (
+    TrainingSample,
+    Features,
+    LearningSolveStats,
+    InstanceFeatures,
+)
 
 
 @pytest.fixture
@@ -25,9 +30,10 @@ def sample() -> TrainingSample:
 @pytest.fixture
 def features() -> Features:
     return Features(
-        instance={
+        instance=InstanceFeatures(
-            "Lazy constraint count": 4,
+            user_features=[0],
-        },
+            lazy_constraint_count=4,
+        ),
         constraints={
             "c1": {
                 "Category": "type-a",

tests/components/test_objective.py

@@ -9,7 +9,7 @@ from numpy.testing import assert_array_equal
 
 from miplearn import GurobiPyomoSolver, LearningSolver, Regressor
 from miplearn.components.objective import ObjectiveValueComponent
-from miplearn.types import TrainingSample, Features
+from miplearn.types import TrainingSample, Features, InstanceFeatures
 from tests.fixtures.knapsack import get_knapsack_instance
 
 import numpy as np
@@ -18,9 +18,9 @@ import numpy as np
 @pytest.fixture
 def features() -> Features:
     return Features(
-        instance={
+        instance=InstanceFeatures(
-            "User features": [1.0, 2.0],
+            user_features=[1.0, 2.0],
-        }
+        )
     )
 
 

tests/test_features.py

@@ -4,7 +4,7 @@
 
 from miplearn import GurobiSolver
 from miplearn.features import FeaturesExtractor
-from miplearn.types import VariableFeatures
+from miplearn.types import VariableFeatures, InstanceFeatures
 from tests.fixtures.knapsack import get_knapsack_instance
 
 
@@ -50,7 +50,7 @@ def test_knapsack() -> None:
                 "User features": [0.0],
             }
         }
-        assert instance.features.instance == {
+        assert instance.features.instance == InstanceFeatures(
-            "User features": [67.0, 21.75],
+            user_features=[67.0, 21.75],
-            "Lazy constraint count": 0,
+            lazy_constraint_count=0,
-        }
+        )