Correctly store features and training data for file-based instances

miplearn/problems/tsp.py

@@ -128,6 +128,7 @@ class TravelingSalesmanInstance(Instance):
     """
 
     def __init__(self, n_cities, distances):
+        super().__init__()
         assert isinstance(distances, np.ndarray)
         assert distances.shape == (n_cities, n_cities)
         self.n_cities = n_cities

miplearn/solvers/learning.py

@@ -135,8 +135,6 @@ class LearningSolver:
 
         # Initialize training sample
         training_sample: TrainingSample = {}
-        if not hasattr(instance, "training_data"):
-            instance.training_data = []
         instance.training_data += [training_sample]
 
         # Initialize stats
@@ -151,7 +149,8 @@ class LearningSolver:
 
         # Extract features
         extractor = FeaturesExtractor(self.internal_solver)
-        instance.features = extractor.extract(instance)
+        instance.features.clear()  # type: ignore
+        instance.features.update(extractor.extract(instance))
 
         callback_args = (
             self,

tests/solvers/test_learning_solver.py

@@ -91,12 +91,14 @@ def test_solve_fit_from_disk():
         solver.solve(instances[0])
         instance_loaded = read_pickle_gz(instances[0].filename)
         assert len(instance_loaded.training_data) > 0
+        assert len(instance_loaded.features) > 0
 
         # Test: parallel_solve
         solver.parallel_solve(instances)
         for instance in instances:
             instance_loaded = read_pickle_gz(instance.filename)
-            assert len(instance.training_data) > 0
+            assert len(instance_loaded.training_data) > 0
+            assert len(instance_loaded.features) > 0
 
         # Delete temporary files
         for instance in instances: