Prev: fast and accurate in mixed mode operations
Next: Software for plotting in Fortranxx?
From: glen herrmannsfeldt on 13 Sep 2009 02:59
Richard Maine <nospam(a)see.signature> wrote:
< Hifi-Comp <wenbinyu.heaven(a)gmail.com> wrote:

<> Relevant to this argument, how the complex number and its arithemetic
<> is implemented in Fortran, is it an overload from real numbers?

< No (at least if I have parsed the question correctly, which I'm not
< entirely sure of). Complex numbers are their own type with their own
< implementation. Some of the low-level stuff clearly shares
< implementation details, but that's a compiler implementation detail
< rather than something user visible. you don't somehow automatically get
< complex stuff just by implementing the corresponding real.

I suppose that could be what he meant, but that isn't what I thought
it meant.

There was for years the suggestion that complex arithmetic be
implemented in C++ through operator overloading. (Since it wasn't
a C++ native type, and supposedly a good demonstration of operator
overloading.) C99 now includes a complex type, I forget if the later
C++ version does.

-- glen

 | 
Pages: 1
Prev: fast and accurate in mixed mode operations
Next: Software for plotting in Fortranxx?