Update app.py

app.py

@@ -55,39 +55,51 @@ def predict_need_maintenance ():
 # Define an endpoint to predict sales for Super Kart
 @pred_mainteanance_api.post ('/v1/EngPredMaintenanceForBatch')
 def predict_need_maintenance_for_batch ():
-    # Get JSON data from the request
-    engine_sensor_inputs = request.get_json ()
-
+    
     # validate request (json)
     # if input is valid - return prediction
     # in case of error - return appropriate error
     try:
         # Get the uploaded CSV file from the request
-        file = request.files ['file']
+        file = request.files.get('file')
 
+        if file is None:
+            return jsonify({
+                "status": "error",
+                "error_type": "input_error",
+                "message": "File not provided"
+            }), 400
+            
         if file.filename == "":
             return jsonify({
                 "status": "error",
+                "error_type": "input_error",
                 "message": "No file selected"
             }), 400        
     
         # Read the file into a DataFrame
         input_df = pd.read_csv (file)
 
+        if input_df.empty:
+            return jsonify({
+                "status": "error",
+                "error_type": "input_error",
+                "message": "Uploaded file is empty"
+            }), 400
+
+        
         # Process the data to clean up and make it ready for prediction
         # mostly we will use the file with same format as given in problem statement for batch prediction
         
         # remove/drop engine condition column if present
         input_df.drop(columns=['Engine Condition'], inplace=True, errors='ignore')
 
-        # update column names to replace space with _ (underscore)
+        # update column names to replace spaces with underscore
         input_df.columns = input_df.columns.str.replace(' ', '_')
 
-        # get all integer columns
+        # Convert int → float
         int_columns = input_df.select_dtypes(include=['int64']).columns
-        
-        # convert integer columns to float
-        int_columns[int_columns] = int_columns[int_columns].astype('float64')
+        input_df[int_columns] = input_df[int_columns].astype('float64')
         
         # Validate entire batch
         validated_df = validate_and_prepare_input(input_df, model)