Update app.py

app.py

@@ -12,7 +12,7 @@ FEATURES = artifact["features"]
 FIXED_FARE = artifact["fixed_fare"]
 RATE_PER_KM = artifact["rate_per_km"]
 
-# ===================== WEATHER (SAFE) =====================
+# ===================== WEATHER =====================
 API_KEY = "YOUR_API_KEY"
 CITY = "Visakhapatnam,IN"
 
@@ -47,14 +47,16 @@ def predict_dynamic_price(
     is_holiday, is_festival
 ):
 
+    # --- Weather ---
     weather_factor, temperature = get_weather_features()
 
+    # --- Base Fare ---
     base_price = FIXED_FARE + (distance_km * RATE_PER_KM)
 
-    # Initialize all features = 0
+    # --- Initialize feature vector ---
     row = {f: 0.0 for f in FEATURES}
 
-    # Fill known inputs
+    # --- Fill primary inputs ---
     inputs = {
         "zone_id": zone_id,
         "hour": hour,
@@ -71,19 +73,42 @@ def predict_dynamic_price(
         "traffic_index": traffic_index,
         "distance_km": distance_km,
         "duration_min": duration_min,
-        "base_fare": base_price,
-        "demand_supply_ratio": (demand + 1) / (supply + 1)
+        "base_fare": base_price
     }
 
     for k, v in inputs.items():
         if k in row:
             row[k] = float(v)
 
-    # Create dataframe with EXACT feature order
+    # =========================================================
+    # CRITICAL FIX — FORCE DEMAND/SUPPLY EFFECT STRONG
+    # =========================================================
+    ratio = (demand + 1) / (supply + 1)
+    row["demand_supply_ratio"] = np.clip(ratio, 0, 50)
+
+    # =========================================================
+    # HANDLE SEASON FEATURES (avoid all-zero vector)
+    # =========================================================
+    if "season_winter" in row:
+        row["season_winter"] = 0
+    if "season_summer" in row:
+        row["season_summer"] = 1   # set one season active
+    if "season_monsoon" in row:
+        row["season_monsoon"] = 0
+    if "season_autumn" in row:
+        row["season_autumn"] = 0
+
+    # --- Create dataframe in correct order ---
     df_row = pd.DataFrame([[row[f] for f in FEATURES]], columns=FEATURES)
 
-    # Predict
+    # --- Predict ---
     surge = float(model.predict(df_row)[0])
+
+    # =========================================================
+    # ADD EXTRA RESPONSE FROM DEMAND/SUPPLY (if model too flat)
+    # =========================================================
+    surge += 0.15 * (ratio - 1)
+
     surge = np.clip(surge, 1.0, 3.5)
 
     final_price = base_price * surge
@@ -118,7 +143,8 @@ demo = gr.Interface(
     fn=predict_dynamic_price,
     inputs=inputs,
     outputs=outputs,
-    title="Dynamic Pricing (Stable Version)"
+    title="Dynamic Pricing (Demand/Supply Fixed)"
 )
 
 demo.launch()
+