How to round float to up and down value c++ with own function?

Question

double round(double a)
{
    double b, c, f, g;
    float d[2];
    c = modf(a, &b);
    if (a > 0) {
        f = a - c;
        g = a - c + 1;
        d[0] = f;
        d[1] = g;
        return d[0], d[1];
    }
    else {
        f = a - c;
        g = a - c - 1;
        d[0] = f;
        d[1] = g;
        return d[0], d[1];
    }
}

I need to get 2 numbers it the end(for ex: if I have num 12.34, I want to get 12 and 13)This is my function of rounding for pos and neg numbers. But it returns only 1 value(( So I'm stack...pls, help how to return 2 values?

or `std::floor(d)` and `std::floor(d) +1`. depends if you want `{12, 12}` or `{12, 13}` for `12.0`. — Jarod42, Feb 21 '19 at 15:09
Read attentively - OWN function, I don't need to use build-in functions — Max, Feb 21 '19 at 15:22
If `a` is an integer, for instance, `a=12.0`, then which values do you need? `{12,12}`, `{11,12}` or `{12,13}`? — Hiroki, Feb 21 '19 at 15:29
https://stackoverflow.com/questions/321068/returning-multiple-values-from-a-c-function — Hanjoung Lee, Feb 21 '19 at 15:41
The expected behavior in the corner cases is key to provide a solution other than a [trivial one](https://wandbox.org/permlink/ChydsmoQ9hlG9Ini). — Bob__, Feb 21 '19 at 15:42
Possible duplicate of [Returning multiple values from a C++ function](https://stackoverflow.com/questions/321068/returning-multiple-values-from-a-c-function) — rsjaffe, Feb 21 '19 at 15:46
@MaximEgorushkin My mistake. Above comments clearly show that one needs some kind of check for fraction. — user14063792468, Feb 21 '19 at 16:42
What's wrong with [the standard rounding functions](https://en.cppreference.com/w/cpp/numeric/math/round) including [std::floor](https://en.cppreference.com/w/cpp/numeric/math/floor) & [std::ceil](https://en.cppreference.com/w/cpp/numeric/math/ceil)? Why re-invent the wheel by rolling your own function? — Jesper Juhl, Feb 21 '19 at 16:48
You don't really need to return multiple values. Just return the smaller of the two values. The other value will always be one more than the returned value. — Pete Becker, Feb 21 '19 at 17:25

score 2 · Answer 1 · answered Feb 21 '19 at 15:53

You cannot return two things in the return, so return d[0],d[1] compiles but doesn't work as you expect. You can use two reference parameters in the function prototype to return. Something like void round(double a, double* result1, double* result2). Into the function, set d[0] to *result1 and d[1] to *result2.

Another thing: Are you sure the line g = a - c - 1; when a is negative is correct? I think you need to do g = a + c - 1;, because a is negative.

score 0 · Accepted Answer · answered Feb 21 '19 at 16:56

#include "pch.h"
#include <iostream>
#include <array>
using namespace std;

auto rounding(double x)
{
    int part = static_cast<int>(x);
    if (x < 0.0)
    {
        return array<int, 2> {
            part - 1, part
        };
    }
    else
    {
        return array<int, 2> {
            part, part + 1
        };
    }
}

int main()
{
    double x;
    cout << "Please, enter a float number to round: ";
    cin >> x;

    auto r1 = rounding(x);
    if (x > 0) {
        cout << "A lower value: " << r1[0] << endl << "A bigger value: " << r1[1];
    }
    else {
        cout << "A bigger value: " << r1[0] << endl << "A lower value: " << r1[1];
    }
}

How to round float to up and down value c++ with own function?

2 Answers2