The overloads with float is unnecessary because the float class implements the SupportsRound protocol.