Mod(negative, positive) gives wrong result in my opinion

[smorgasbord]

http://en.wikipedia.org/wiki/Modular_arithmetic
as per the link above

MsgBox % mod(2,5) = MsgBox % mod(-3,5)

but they aren't

Simple explanation:
if we draw the number line the multiples of 5 are ......-15, -10, -5, 0, 5, 10, 15
so that Mod(-3,5) should return 2 because -5 + 2 = -3

So i wish MsgBox % mod(-3,5) resulted in 2 and not -3
edited and added:
Mod is something related to remainder and remainder can never be negative, i stumbled upon this while trying to learn by helping @dmg in his post on the other forum

[gregster]

remainder can never be negative

Not true.
But it can get a bit tricky, if dividend OR divisor are negative. There are different ways how programming languages define modulo operations and deal with it: " in C++ a negative number will be returned if the first argument is negative, and in Python a negative number will be returned if the second argument is negative" (http://en.wikipedia.org/wiki/Modular_ar ... Remainders)
See also http://en.wikipedia.org/wiki/Modulo_operation

In AHK Mod(-3,5) is -3 because AHK docs say "The sign of the result is always the same as the sign of the first parameter"
Therefore -3 : 5 = 0, rest -3, check: ( 5 * 0 ) -3 = 0-3 = -3
And Mod(3, -5) hence leads to 3: (-5) = 0, rest 3, -> -5*0 + 3 = 0+3 = 3

In another programming language, it could be like you expected (second parameter dtermines the sign of the result)
-3 : 5 = -1, rest 2 and (5 * (-1)) + 2 = -5 + 2 = -3

It's also true. Some programming languages have different functions for these two cases, some restrict themselves to one case. It depends on how you define modulo operations.

[smorgasbord]

@gregstar
I just realised that whatever is given in maths does not mean the same would be implemented in all computer languages exactly.
learning stuff on the go

[gregster]

Very true. Especially, if you try to port a program to another language, it can get really ugly, if you are not aware of that....

[smorgasbord]

All i know is a little bit like 2-3% of autohotkey

[gregster]

Same here

[smorgasbord]

@gregstar
http://mathforum.org/library/drmath/view/52343.html

some heavy debate seems to be there on the same issue

may be they made a Rem() function

[gregster]

So Lotus 1-2-3 and Excel used different Modulos? I never noticed... For some time, I had to use them both parallely, back in the nineties. But probably I didn't use many modulo operations back then, anyway...
Thanks for the link to that background info (not especially about Lotus and Excel, but about how these differences originated) and this quote

As _C: A Reference Manual_, by Harbison and Steele, says, "For maximum portability, programs should therefore avoid depending on the behavior of the remainder operator when applied to negative integral operands."

Very true!

[GameNtt]

I might be quite a bit late to the party (just a bit off of a decade), but I was doing something related to Day of the Week calculation and this was the reason for my interest in this topic. I whipped up a function to resolve this issue but beware, this does call the native Mod() function.

Code: Select all

cMod(dividend, divisor)
{
    if(dividend > 0)
    {
        Return Mod(dividend, divisor)
    }
    else if(dividend < 0)
    {
        While(dividend < 0)
        {
            dividend += divisor
        }
        return Mod(dividend, divisor)
    }
    else
    {
        Return 0
    }
}