[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: arithmetic issues



Alan Watson <a.watson@xxxxxxxxxxxxxxxx> writes:

> Or quotient-and-remainder. Isn't "+" sufficiently overloaded as it is
> without having it stand for "and" :-)

quotient&remainder? Unfortunately quotient,remainder is not a valid
identifier.

>> Mathematically, mixed exactness complex numbers makes no sense.
>> Twisting the whole numeric tower around this artifice is wrong.
>
> Maybe, but a cheap way to get an inexact imaginary is an number with
> an exact zero for its real part and an inexact real for its imaginary
> part.

If I understood what William Kahan tried to say in one of his
articles, it is that it's sometimes essential to distinguish between
the real part being 0, 0.0, and -0.0, where having only 0.0 and -0.0
is insufficient. He called the first variant "imaginary type", as in
C99, and claimed that Java made a mistake by providing only a complex
type with two floating point parts.

I might be wrong however because I did not understand the technical
reasoning behind that. I think it had something to do with choosing
the right sides near branch cuts, or maybe with 0 not flipping to the
other side on negation, in contrast to 0.0 and -0.0.

-- 
   __("<         Marcin Kowalczyk
   \__/       qrczak@xxxxxxxxxx
    ^^     http://qrnik.knm.org.pl/~qrczak/