4 refine tests code quality adherence to solid

app/services/model_store.py

@@ -6,6 +6,9 @@
 import joblib
 import numpy as np
 import pandas as pd
+from loguru import logger
+from sklearn.linear_model import SGDRegressor
+from sklearn.preprocessing import StandardScaler
 
 try:
     from services.features import get_feature_columns, make_features

app/tests/test_api.py

@@ -18,7 +18,7 @@ def test_status_endpoint(client):
     """Test the /status endpoint."""
     response = client.get("/status")
     assert response.status_code == 200
-
+    
     data = response.json()
     assert "models" in data
     assert isinstance(data["models"], list)
@@ -27,36 +27,18 @@ def test_status_endpoint(client):
 def test_train_endpoint_success(client):
     """Test the /train endpoint with valid data."""
     # Mock the CSV reading and training
-    mock_df = pd.DataFrame(
-        {
-            "tenant_id": ["tenant_a"] * 15,
-            "product": ["apples"] * 15,
-            "date": pd.date_range("2024-01-01", periods=15),
-            "units_sold": [10, 12, 15, 8, 20, 18, 22, 25, 19, 30, 28, 35, 32, 40, 38],
-            "price": [
-                1.0,
-                1.1,
-                1.05,
-                0.95,
-                1.2,
-                1.15,
-                1.25,
-                1.3,
-                1.18,
-                1.4,
-                1.35,
-                1.45,
-                1.42,
-                1.5,
-                1.48,
-            ],
-        }
-    )
-
-    with patch("pandas.read_csv", return_value=mock_df):
-        with patch("app.main.train_on_dataframe") as mock_train:
+    mock_df = pd.DataFrame({
+        'tenant_id': ['tenant_a'] * 15,
+        'product': ['apples'] * 15,
+        'date': pd.date_range('2024-01-01', periods=15),
+        'units_sold': [10, 12, 15, 8, 20, 18, 22, 25, 19, 30, 28, 35, 32, 40, 38],
+        'price': [1.0, 1.1, 1.05, 0.95, 1.2, 1.15, 1.25, 1.3, 1.18, 1.4, 1.35, 1.45, 1.42, 1.5, 1.48]
+    })
+
+    with patch('pandas.read_csv', return_value=mock_df):
+        with patch('app.main.train_on_dataframe') as mock_train:
             response = client.post("/train")
-
+            
             assert response.status_code == 200
             data = response.json()
             assert data["status"] == "ok"
@@ -67,9 +49,9 @@ def test_train_endpoint_success(client):
 def test_train_endpoint_handles_error(client):
     """Test the /train endpoint when training fails."""
     # Mock read_csv to raise an exception
-    with patch("pandas.read_csv", side_effect=Exception("File not found")):
+    with patch('pandas.read_csv', side_effect=Exception("File not found")):
         response = client.post("/train")
-
+        
         assert response.status_code == 500
         data = response.json()
         assert "detail" in data
@@ -78,13 +60,14 @@ def test_train_endpoint_handles_error(client):
 def test_predict_endpoint_all_targets(client):
     """Test the /predict endpoint with all valid target options."""
     targets = ["next_week", "this_month", "next_month"]
-
+    
     for target in targets:
-        response = client.get(
-            "/predict",
-            params={"tenant_id": "tenant_a", "product": "apples", "target": target},
-        )
-
+        response = client.get("/predict", params={
+            "tenant_id": "tenant_a", 
+            "product": "apples", 
+            "target": target
+        })
+
         assert response.status_code == 200
         data = response.json()
         assert data["tenant_id"] == "tenant_a"
@@ -99,34 +82,33 @@ def test_predict_endpoint_different_tenants_products(client):
     test_cases = [
         ("tenant_a", "apples"),
         ("tenant_b", "bananas"),
-        ("tenant_c", "laptops"),
+        ("tenant_c", "laptops")
     ]
-
+    
     for tenant_id, product in test_cases:
-        response = client.get(
-            "/predict",
-            params={"tenant_id": tenant_id, "product": product, "target": "next_week"},
-        )
-
-        # Some combinations might not exist, so we accept 200 or 404
+        response = client.get("/predict", params={
+            "tenant_id": tenant_id,
+            "product": product,
+            "target": "next_week"
+        })
+
         assert response.status_code in [200, 404]
 
 
 def test_predict_endpoint_forecast_structure(client):
     """Test the structure of forecast data."""
-    response = client.get(
-        "/predict",
-        params={"tenant_id": "tenant_a", "product": "apples", "target": "next_week"},
-    )
-
+    response = client.get("/predict", params={
+        "tenant_id": "tenant_a",
+        "product": "apples", 
+        "target": "next_week"
+    })
+
     if response.status_code == 200:
         data = response.json()
         forecast = data["forecast"]
-
-        # next_week should return 7 days
+
         assert len(forecast) == 7
-
-        # Check structure of each forecast item
+
         for day in forecast:
             assert "date" in day
             assert "predicted_units" in day
@@ -135,85 +117,75 @@ def test_predict_endpoint_forecast_structure(client):
 
 def test_predict_endpoint_this_month_structure(client):
     """Test the structure of this_month forecast."""
-    response = client.get(
-        "/predict",
-        params={"tenant_id": "tenant_a", "product": "apples", "target": "this_month"},
-    )
-
+    response = client.get("/predict", params={
+        "tenant_id": "tenant_a",
+        "product": "apples",
+        "target": "this_month"
+    })
+
     if response.status_code == 200:
         data = response.json()
         forecast = data["forecast"]
-
-        # this_month should return remaining days in current month
+
         assert len(forecast) > 0
-
+        
         for day in forecast:
             assert "date" in day
             assert "predicted_units" in day
 
 
 def test_predict_endpoint_next_month_structure(client):
     """Test the structure of next_month forecast."""
-    response = client.get(
-        "/predict",
-        params={"tenant_id": "tenant_a", "product": "apples", "target": "next_month"},
-    )
-
+    response = client.get("/predict", params={
+        "tenant_id": "tenant_a",
+        "product": "apples",
+        "target": "next_month"
+    })
+
     if response.status_code == 200:
         data = response.json()
         forecast = data["forecast"]
-
-        # next_month should return all days in next month
+
         assert len(forecast) >= 28  # At least February
-        assert len(forecast) <= 31  # At most 31 days
-
+        assert len(forecast) <= 31   # At most 31 days
+        
         for day in forecast:
             assert "date" in day
             assert "predicted_units" in day
 
 
 def test_cors_middleware(client):
     """Test that CORS middleware is properly configured."""
-    # Test preflight request
+
     response = client.options("/predict")
-    # OPTIONS requests might return 405 Method Not Allowed, which is fine
     assert response.status_code in [200, 405]
 
-    # Check CORS headers if response is successful
     if response.status_code == 200:
         assert "access-control-allow-origin" in response.headers
         assert response.headers["access-control-allow-origin"] == "*"
 
 
 def test_app_startup(client):
     """Test that app startup loads models and starts polling."""
-    # The fixture already tests startup by creating the client
-    # This test explicitly verifies the startup behavior
     response = client.get("/status")
     assert response.status_code == 200
-
-    # Should have loaded some models
+
     data = response.json()
     assert len(data["models"]) > 0
 
 
 def test_predict_endpoint_parameter_validation(client):
     """Test various parameter combinations for predict endpoint."""
-    # Test with missing tenant_id
-    response = client.get(
-        "/predict", params={"product": "apples", "target": "next_week"}
-    )
+
+    response = client.get("/predict", params={"product": "apples", "target": "next_week"})
     assert response.status_code == 422
-
-    # Test with missing product
-    response = client.get(
-        "/predict", params={"tenant_id": "tenant_a", "target": "next_week"}
-    )
+
+    response = client.get("/predict", params={"tenant_id": "tenant_a", "target": "next_week"})
     assert response.status_code == 422
-
-    # Test with empty strings
-    response = client.get(
-        "/predict", params={"tenant_id": "", "product": "apples", "target": "next_week"}
-    )
-    # This should return 404 (model not found) rather than validation error
+    
+    response = client.get("/predict", params={
+        "tenant_id": "",
+        "product": "apples",
+        "target": "next_week"
+    })
     assert response.status_code == 404