Also, when you run it on its own have you had a chance to review some sample data to ensure the structure and values all look appropriate?