between floating point format, binary coded decimal (BCD), or exponential format which one  you would recommend for processing efficiency and explain why