Also to have an equivalent floating-point interface that can actually generate both 1.0 and 0.0

On Wed, Mar 3, 2021 at 3:02 PM Edward Catmur <ecatmur@googlemail.com> wrote:
It seems to me that what you really want is a uniform_int_distribution that accepts a half-open interval. I assume the reason the Standard specifies a closed interval is to allow a distribution over the whole of an integer type?

On Wed, Mar 3, 2021 at 9:59 AM Gašper Ažman via Std-Proposals <std-proposals@lists.isocpp.org> wrote:
Not sure if I understand, because it seems so obvious otherwise: you just want a function that returns size()-1?

On Wed, Mar 3, 2021 at 7:56 AM Kurt Raburn via Std-Proposals <std-proposals@lists.isocpp.org> wrote:
hello,

i'm not sure if this was my fault, but this is the issue i was having.

std::uniform_int_distribution<int> dist(0, vec.size()); // runtime error

std::uniform_int_distribution<int> dist(0, vec.size() - 1); // ok

after some thought, i wrote a function that returns the index count.

template<typename T>
int indices(const std::vector<T>& vec)
{
    int cnt = 0;

    for (T i : vec) {
        ++cnt;
    }

    return --cnt;
}

then, i modified the size() function, and it returned the same result.

_NODISCARD size_type indices() const noexcept
        {   
        return (static_cast<size_type>(this->_Mylast() - this->_Myfirst() - 1));
        }

usage:

std::uniform_int_distribution<int> dist(0, vec.indices());

forgive me if i overlooked a library function that would've solved this.





--
Std-Proposals mailing list
Std-Proposals@lists.isocpp.org
https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
--
Std-Proposals mailing list
Std-Proposals@lists.isocpp.org
https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals