We learn about groups, rings and fields in algebra - but floating point numbers (like double in many modern programming languages) do not form one of the above algebraic entities because they are not associative.
Is there an algebraic theory behind these numbers?