Update app.py

app.py

@@ -9,22 +9,41 @@ from sklearn.pipeline import Pipeline
 # Load Model
 try:
     with open("final_model_1.pkl", "rb") as f:
-        model = pickle.load(f)
+        model, ct = pickle.load(f)  # Load both the model and the fitted transformer
     st.success("✅ Model loaded successfully!")
 except FileNotFoundError:
     st.error("❌ Model file not found! Please upload `final_model.pkl`.")
     model = None
+    ct = None
 
 # Define your preprocessing pipeline (to match the one used during training)
-nom_pl = Pipeline(steps=[('Encoding', OneHotEncoder(drop='first', sparse_output=False, handle_unknown='ignore'))])
-
-# Create a ColumnTransformer for the preprocessing steps
-ct = ColumnTransformer(
-    transformers=[
-        ('nom_pl', nom_pl, [0, 4]),  # Apply OneHotEncoder on 'POSTED_BY' and 'BHK_OR_RK'
-    ], 
-    remainder='passthrough'  # Leave other columns as they are
-)
+if model and ct is None:
+    nom_pl = Pipeline(steps=[('Encoding', OneHotEncoder(drop='first', sparse_output=False, handle_unknown='ignore'))])
+
+    # Create a ColumnTransformer for the preprocessing steps
+    ct = ColumnTransformer(
+        transformers=[
+            ('nom_pl', nom_pl, [0, 4]),  # Apply OneHotEncoder on 'POSTED_BY' and 'BHK_OR_RK'
+        ], 
+        remainder='passthrough'  # Leave other columns as they are
+    )
+
+    # Here we simulate fitting the transformer on a sample of data
+    train_data = pd.DataFrame({
+        'POSTED_BY': ['Owner', 'Dealer', 'Builder'],
+        'UNDER_CONSTRUCTION': [1, 0, 1],
+        'RERA': [1, 0, 1],
+        'BHK_NO_': [2, 3, 2],
+        'BHK_OR_RK': ['BHK', 'RK', 'BHK'],
+        'SQUARE_FT': [1200, 1500, 1300],
+        'READY_TO_MOVE': [1, 0, 1],
+        'RESALE': [1, 0, 1],
+        'LONGITUDE': [20.75, 21.00, 19.80],
+        'LATITUDE': [77.32, 78.00, 76.50]
+    })
+
+    # Fit the transformer on sample data
+    ct.fit(train_data)
 
 # Title of the application
 st.markdown("<h1 class='title'>🏡 House Price Predictor</h1>", unsafe_allow_html=True)