citation needed :)

LGTM

Lmao

): will be counted

why

But it is good for a laugh, plus taught me __builtin_popcount

@dust55 The const to function hack does not work on all platforms

What is the first number and what is the second number? Complex numbers are unordered. All solutions coincide with the the reference solution but the order of numbers can be switched because a + b = b + a and a * b = b * a.

One possible way to compare floating-point numbers is the following:

bool cmp(double a, double b, double rel_tol, double abs_tol) {
  return fabs(a - b) <= fmax(rel_tol * fmax(fabs(a), fabs(b)), abs_tol);
}

(This code is based on the math.isclose function from Python.)

The values of rel_tol and abs_tol should be carefully selected. Usually, abs_tol is very small. Something like 1e-15. rel_tol could be between 1e-3 and 1e-9.

It's stated that each of sol[{0 | 1 | 2 | 3}] is either a "real part [of a number]" or a "coefficient of the imaginary part", so these four seem to be all real; so there are no complex numbers anywhere in sol[0] + sol[1] + sol[2] + sol[3] ≅ sum and (sol[0] + sol[1]) * (sol[2] + sol[3]) ≅ product are all real.

(it is not clear which number is first and which is second)

It is specified in the description:

sol[0]: the real part of the first number

sol[1]: the coefficient of the imaginary part of the first number (expressed in terms of √-1, more commonly known as i)

sol[2]: the real part of the second number, must not be greater than sol[0]

sol[3]: the coefficient of the imaginary part of the second number, must not be greater than sol[1]

relative errors

Sorry, I'm not quite sure - does that mean percentage difference?

Sure, will fix this issue and the other one later today

Do not test that the returned values are close to the expected values (it is not clear which number is first and which is second). Suppose that the user function returns 2 complex numbers a and b. Then the tests should check that a + b is close to sum and a * b is close to product. I also recommend to use relative errors when checking that two floating-point numbers are close.

Do not ask to return a dynamically allocated array with a fixed number of elements. It is better to return a struct with 4 fields (but for this task it is much better to use complex numbers directly):

struct answer {
  double complex a;
  double complex b;
};

Does it work now? I've done that type-cast, and reduced the range of sum and product in the random tests.

It does not fix anything. You need to cast sum and product to double.