function SIGN(double x,double y) in c++

Question

What does function SIGN(double x,double y) do in C++ language and how to translate it to c#

its not a sign(double) whic returns sign of number this one takes two arguments and I can't find it nowhere on the net — Милош Чолић, Jun 21 '12 at 21:58
@thb It's lacking important information. Since it's not a standard function, the question can so far only be answered "Whatever the coder made it do, and for translating it, first find out what it does". Or with wild guesses. — Daniel Fischer, Jun 21 '12 at 22:15

hatchet - done with SOverflow · Answer 1 · 2012-06-21T22:22:37.973

2

A function like that exists in Fortan http://gcc.gnu.org/onlinedocs/gfortran/SIGN.html

SIGN(A,B) returns the value of A with the sign of B.

If B >= 0 then the result is ABS(A), else it is -ABS(A).

Possibly someone implemented the same function in C++ (but something custom, not a standard part of C++ or the standard libraries).

A c# version of that would be

  A = Math.Abs(A);
  if (B<0.0) A = -A;

edited Jun 21 '12 at 22:22

answered Jun 21 '12 at 22:04

hatchet - done with SOverflow

15,919
5
47
57

If `A = B = -1` then `A * Math.Sign(B)` will not have the same sign as `B`. – aioobe Jun 21 '12 at 22:07
thanks, in a hurry, will edit. – hatchet - done with SOverflow Jun 21 '12 at 22:08
1

`Math.Abs(A) * Math.Sign(B)` should do. – aioobe Jun 21 '12 at 22:09
thank you that is what was the problem – Милош Чолић Jun 21 '12 at 22:12
Math.Sign won't work because it returns 0 if B == 0. – hatchet - done with SOverflow Jun 21 '12 at 22:13
It's a good idea to either clearly document the preconditions or otherwise throw an exception if `B >= 0` and `A = MIN_VAL` though. Just looked it up: `Math.Abs` in C# does already throw an exception (good thing, java got that wrong), so the code is wrong for one corner case but handles the other fine. – Voo Jun 21 '12 at 22:23
@Voo - I can see that with integral types, but I don't think that would be an issue for double type inputs since Double.MaxValue = Abs(Double.MinValue). – hatchet - done with SOverflow Jun 21 '12 at 22:29
@hatchet Ah I see - I thought we were talking about ints here (didn't notice the `< 0.0` check). Yep works fine for doubles. – Voo Jun 21 '12 at 22:31

score 1 · Answer 2 · answered Jun 21 '12 at 22:02

In the C++ code you are translating to C#, look for a macro definition or inline function definition for SIGN. The definition should spell out exactly how you should implement it in C#.

If I had to guess from the name, SIGN(x,y) returns true if x and y have the same sign, and false otherwise.

score 0 · Answer 3 · answered Jun 21 '12 at 22:06

The C++ standard library has no SIGN() function. Your question does not give enough information for me to tell you for certain what this SIGN() is, but C++ succeeds the older programming language C, and my informed hunch is that SIGN() is a preprocessor macro defined in the old C style.

Search your code for a line that begins #define SIGN, possibly with more than one space between the #define and the SIGN. If you find the line, it is likely to tell you what your SIGN() is and does.

function SIGN(double x,double y) in c++

3 Answers3