How to find absolute value in C?

Question

I need to lighten the code, and then eliminate the underlying if that check if the subtraction are <0 and if the condition is true I multiply * -1 in order to have subtraction >0 (absolute value).

I wonder if there is a function or some operator able to perform the calculation directly in absolute value?

int i, j, N, max, s, s1;
s = v[i] - v[j];
s1 = i - j;
if (s1 < 0){
    s1 *= -1;
}
if (s < 0){
    s *= -1;
}

I cannot find a modulo or a sum in that code. Please show the real code snippet you want to "lighten". — Gerhardh, Apr 12 '22 at 13:12
for integers [abs](https://en.cppreference.com/w/c/numeric/math/abs) for floats [fabs](https://en.cppreference.com/w/c/numeric/math/fabs) — Iłya Bursov, Apr 12 '22 at 13:14
It would be nice to see how your variables are declared and initialized. — hetepeperfan, Apr 12 '22 at 13:14
Just use the in-built functions. Or, if you're feeling brave (and `int` is 32 bits), then: `s = (s ^ (((int64_t)(s) >> 32) & 0x0FFFFFFFFll)) - (((int64_t)(s) >> 32) & 0x0FFFFFFFFll);`. Similarly for `s1`. — Adrian Mole, Apr 12 '22 at 13:26

score 1 · Answer 1 · answered Apr 12 '22 at 13:21

As the comments say:

s = abs(x - y)

The manpage for abs (and also labs, llabs): https://linux.die.net/man/3/abs

It looks like you're doing this in a for loop of some sort (i,j are hopefully loop indices, otherwise you need to name your variables more descriptively). In that case, you should be able to architect it such that j is never greater than i for a performance increase.

score 0 · Accepted Answer · answered Apr 12 '22 at 13:32

Absolute value

To avoid int overflow of abs(v[i] - v[j]) and its undefined behavior (UB) , take advantage that the unsigned range is very commonly twice the int positive range and perform unsigned math.

// int s;
unsigned s;
if (v[i] > v[j]) {
  s = 0u + v[i] - v[j];
} else {
  s = 0u + v[j] - v[i];
}

How to find absolute value in C?

2 Answers2