Add built-in Double operations to vectorisation