Update Streamlit app

app.py

@@ -40,40 +40,36 @@ def engineer_features(df):
     """Apply feature engineering to match training pipeline exactly"""
     df_enhanced = df.copy()
     
-    # STEP 1: Rename columns to match training convention
-    rename_mapping = {
-        "Lub oil pressure": "Lube Oil Pressure",
-        "lub oil temp": "Lube Oil Temperature",
-        "Coolant temp": "Coolant Temperature",
-        "Engine rpm": "Engine RPM",
-        "Fuel pressure": "Fuel Pressure",
-        "Coolant pressure": "Coolant Pressure"
-    }
+    # Create the three derived features that the model was trained with
+    # Lub_Stress_Index = Lub oil pressure * lub oil temp
+    df_enhanced['Lub_Stress_Index'] = (
+        df_enhanced['Lub oil pressure'] * df_enhanced['lub oil temp']
+    )
     
-    for old_name, new_name in rename_mapping.items():
-        if old_name in df_enhanced.columns:
-            df_enhanced.rename(columns={old_name: new_name}, inplace=True)
+    # Thermal_Efficiency = Coolant pressure / (Coolant temp + 1e-5)
+    df_enhanced['Thermal_Efficiency'] = (
+        df_enhanced['Coolant pressure'] / (df_enhanced['Coolant temp'] + 1e-5)
+    )
     
-    # STEP 2: Get sensor columns (all columns except 'Engine Condition')
-    sensor_columns = [col for col in df_enhanced.columns if col != 'Engine Condition']
+    # Power_Load_Index = Engine rpm * Fuel pressure
+    df_enhanced['Power_Load_Index'] = (
+        df_enhanced['Engine rpm'] * df_enhanced['Fuel pressure']
+    )
     
-    # STEP 3: Add ratio features
-    if 'Lube Oil Pressure' in df_enhanced.columns and 'Coolant Pressure' in df_enhanced.columns:
-        df_enhanced['Oil_Coolant_Pressure_Ratio'] = (
-            df_enhanced['Lube Oil Pressure'] / (df_enhanced['Coolant Pressure'] + 1)
-        )
-    
-    if 'Lube Oil Temperature' in df_enhanced.columns and 'Coolant Temperature' in df_enhanced.columns:
-        df_enhanced['Oil_Coolant_Temp_Diff'] = (
-            df_enhanced['Lube Oil Temperature'] - df_enhanced['Coolant Temperature']
-        )
-    
-    # STEP 4: Add squared features for EACH sensor column
-    for col in sensor_columns:
-        if col in df_enhanced.columns:
-            df_enhanced[f'{col}_Squared'] = df_enhanced[col] ** 2
+    # Return features in EXACT order the model was trained with
+    feature_order = [
+        'Engine rpm',
+        'Lub oil pressure',
+        'Fuel pressure',
+        'Coolant pressure',
+        'lub oil temp',
+        'Coolant temp',
+        'Lub_Stress_Index',
+        'Thermal_Efficiency',
+        'Power_Load_Index'
+    ]
     
-    return df_enhanced
+    return df_enhanced[feature_order]
 
 # ============================================
 # MAIN APP