Representation and similarity: Suarez on necessary and sufficient conditions of scientific representation