IQR Loss of Precision

[a-penton]

Some values are losing precision in the IQR problem (2), likely because of the float return type. Example:

#include <iostream>
#include <iomanip>

int main()
{
    float n = 12777776.50;
    std::cout << std::fixed << std::setprecision(2) << n;  // 12777776.00
    return 0;
}