A correction to Bahr’s phase deviation” method for tensor decomposition