testing updates to rasa's

A4/conftest.py

@@ -17,14 +17,19 @@ def models_dir(repo_root):
     return os.path.join(repo_root, "A3", "models")
 
 
+@pytest.fixture
+def a4_models_dir(repo_root):
+    return os.path.join(repo_root, "A4", "models")
+
+
 @pytest.fixture
 def regression_model_path(models_dir):
     return os.path.join(models_dir, "champion_model_final_2.pkl")
 
 
 @pytest.fixture
-def classification_model_path(models_dir):
-    return os.path.join(models_dir, "final_champion_model_A3.pkl")
+def classification_model_path(a4_models_dir):
+    return os.path.join(a4_models_dir, "weaklink_classifier_rf.pkl")
 
 
 @pytest.fixture
@@ -79,4 +84,11 @@ def sample_classification_features(classification_artifact):
 # expected values
 @pytest.fixture
 def expected_classification_classes():
-    return ["Lower Body", "Upper Body"]
+    # all 14 weaklink categories
+    return [
+        'ExcessiveForwardLean', 'ForwardHead', 'LeftArmFallForward',
+        'LeftAsymmetricalWeightShift', 'LeftHeelRises', 'LeftKneeMovesInward',
+        'LeftKneeMovesOutward', 'LeftShoulderElevation', 'RightArmFallForward',
+        'RightAsymmetricalWeightShift', 'RightHeelRises', 'RightKneeMovesInward',
+        'RightKneeMovesOutward', 'RightShoulderElevation'
+    ]

A4/models/champion_model_final_2.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2aa2439a4b32f522de5f0cef2f5267d02c39f14a36fdbcbfb08273cfa7fd6178
-size 9202

A4/models/final_champion_model_A3.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:29b0e9b7f4f950904687023290cc0e5e044426f821778ed3d220bba708621716
-size 563557

A4/test_models.py

@@ -76,7 +76,8 @@ class TestClassificationModelLoading:
         assert "feature_columns" in classification_artifact
 
     def test_classification_artifact_has_classes(self, classification_artifact):
-        assert "classes" in classification_artifact
+        # weaklink categories for the 14 classes
+        assert "weaklink_categories" in classification_artifact
 
     def test_classification_model_has_predict_method(self, classification_artifact):
         model = classification_artifact["model"]
@@ -85,7 +86,7 @@ class TestClassificationModelLoading:
     def test_classification_classes_match_expected(
         self, classification_artifact, expected_classification_classes
     ):
-        classes = list(classification_artifact["classes"])
+        classes = list(classification_artifact["weaklink_categories"])
         assert sorted(classes) == sorted(expected_classification_classes)
 
 
@@ -95,7 +96,11 @@ class TestClassificationModelPrediction:
         self, classification_artifact, sample_classification_features
     ):
         model = classification_artifact["model"]
-        prediction = model.predict(sample_classification_features)
+        scaler = classification_artifact.get("scaler")
+        features = sample_classification_features
+        if scaler is not None:
+            features = scaler.transform(features)
+        prediction = model.predict(features)
         assert isinstance(prediction, np.ndarray)
 
     def test_classification_prediction_shape(
@@ -103,7 +108,11 @@ class TestClassificationModelPrediction:
     ):
         # one class per sample
         model = classification_artifact["model"]
-        prediction = model.predict(sample_classification_features)
+        scaler = classification_artifact.get("scaler")
+        features = sample_classification_features
+        if scaler is not None:
+            features = scaler.transform(features)
+        prediction = model.predict(features)
         assert prediction.shape[0] == len(sample_classification_features)
 
     def test_classification_prediction_is_valid_class(
@@ -112,7 +121,11 @@ class TestClassificationModelPrediction:
     ):
         # should be a valid class
         model = classification_artifact["model"]
-        prediction = model.predict(sample_classification_features)[0]
+        scaler = classification_artifact.get("scaler")
+        features = sample_classification_features
+        if scaler is not None:
+            features = scaler.transform(features)
+        prediction = model.predict(features)[0]
         assert prediction in expected_classification_classes
 
 
@@ -122,7 +135,7 @@ class TestModelArtifactStructure:
         assert "test_metrics" in regression_artifact
 
     def test_classification_artifact_has_metrics(self, classification_artifact):
-        assert "test_metrics" in classification_artifact
+        assert "test_performance" in classification_artifact
 
     def test_regression_metrics_has_r2(self, regression_artifact):
         metrics = regression_artifact.get("test_metrics", {})