Update main.py

main.py

@@ -5,66 +5,70 @@ from fastapi import FastAPI, Request, Form
 from fastapi.responses import HTMLResponse
 from fastapi.templating import Jinja2Templates
 
-app = FastAPI()
-# Templates folder
+app = FastAPI(title="Student Score Predictor")
 templates = Jinja2Templates(directory="templates")
 
-# Load trained model (ensure student_performance_model.pkl is in /app)
+# Load the trained model at startup
 MODEL_PATH = os.getenv('MODEL_PATH', 'student_performance_model.pkl')
-with open(MODEL_PATH, 'rb') as f:
-    model = pickle.load(f)
+try:
+    with open(MODEL_PATH, 'rb') as f:
+        model = pickle.load(f)
+except Exception as e:
+    raise RuntimeError(f"Failed to load model: {e}")
+
+# List of input fields and types
+FIELDS = [
+    ('Age', int),
+    ('Gender', str),
+    ('HoursOfStudyPerDay', float),
+    ('SchoolAttendanceRate', float),
+    ('TuitionAccess', str),
+    ('AveragePreviousScores', float),
+    ('HoursOfSleep', float),
+    ('BreakfastDaily', str),
+    ('ScreenTimeHours', float),
+    ('PhysicalActivityHours', float),
+    ('PlaysSport', str),
+    ('MentalHealthScore', float),
+    ('ParentalEducationLevel', str),
+    ('HouseholdIncomeLevel', str),
+    ('StudyEnvironmentRating', float),
+    ('FriendSupportScore', float),
+    ('ParticipatesInClubs', str),
+    ('PartTimeWork', str),
+]
 
 @app.get('/', response_class=HTMLResponse)
-async def read_form(request: Request):
+async def get_form(request: Request):
     return templates.TemplateResponse('index.html', {
         'request': request,
-        'predicted': None
+        'predicted': None,
+        'values': {}
     })
 
 @app.post('/predict', response_class=HTMLResponse)
-async def predict(request: Request,
-                  Age: int = Form(...),
-                  Gender: str = Form(...),
-                  HoursOfStudyPerDay: float = Form(...),
-                  SchoolAttendanceRate: float = Form(...),
-                  TuitionAccess: str = Form(...),
-                  AveragePreviousScores: float = Form(...),
-                  HoursOfSleep: float = Form(...),
-                  BreakfastDaily: str = Form(...),
-                  ScreenTimeHours: float = Form(...),
-                  PhysicalActivityHours: float = Form(...),
-                  PlaysSport: str = Form(...),
-                  MentalHealthScore: float = Form(...),
-                  ParentalEducationLevel: str = Form(...),
-                  HouseholdIncomeLevel: str = Form(...),
-                  StudyEnvironmentRating: float = Form(...),
-                  FriendSupportScore: float = Form(...),
-                  ParticipatesInClubs: str = Form(...),
-                  PartTimeWork: str = Form(...)):
-    # Organize input into DataFrame
-    data = pd.DataFrame([{
-        'Age': Age,
-        'Gender': Gender,
-        'HoursOfStudyPerDay': HoursOfStudyPerDay,
-        'SchoolAttendanceRate': SchoolAttendanceRate,
-        'TuitionAccess': TuitionAccess,
-        'AveragePreviousScores': AveragePreviousScores,
-        'HoursOfSleep': HoursOfSleep,
-        'BreakfastDaily': BreakfastDaily,
-        'ScreenTimeHours': ScreenTimeHours,
-        'PhysicalActivityHours': PhysicalActivityHours,
-        'PlaysSport': PlaysSport,
-        'MentalHealthScore': MentalHealthScore,
-        'ParentalEducationLevel': ParentalEducationLevel,
-        'HouseholdIncomeLevel': HouseholdIncomeLevel,
-        'StudyEnvironmentRating': StudyEnvironmentRating,
-        'FriendSupportScore': FriendSupportScore,
-        'ParticipatesInClubs': ParticipatesInClubs,
-        'PartTimeWork': PartTimeWork
-    }])
+async def post_predict(request: Request, **form_data):
+    # Convert form inputs to correct types
+    record = {}
+    for name, dtype in FIELDS:
+        raw = form_data.get(name)
+        if raw is None:
+            # missing field
+            record[name] = None
+            continue
+        record[name] = dtype(raw)
+
+    # Prepare DataFrame
+    df = pd.DataFrame([record])
     # Predict
-    pred = model.predict(data)[0]
+    try:
+        score = model.predict(df)[0]
+        predicted = round(score, 2)
+    except Exception as e:
+        predicted = f"Error: {e}"
+
     return templates.TemplateResponse('index.html', {
         'request': request,
-        'predicted': round(pred, 2)
+        'predicted': predicted,
+        'values': record
     })