问题
The function std::shuffle has been introduced in C++11:
template< class RandomIt, class URNG >
void shuffle( RandomIt first, RandomIt last, URNG&& g );
and it has the same signature as one of the overloads of std::random_shuffle which was also introduced in C++11:
template< class RandomIt, class RandomFunc >
void random_shuffle( RandomIt first, RandomIt last, RandomFunc&& r );
The difference is in the third parameter where:
URNG must meet the requirements of UniformRandomNumberGenerator
Is this all? Is the difference just that shuffle performs an extra compile time check? Is the behavior otherwise the same?
回答1:
If you read the documentation at cppreference.com closely, you will find that the RandomFunc passed to random_shuffle has a different interface. It is invoked as r(n). This existed before C++11.
std::shuffle uses a standardized way of getting random numbers and invokes g(). This standardized random number generators where introduced with C++11 together with std::shuffle.
回答2:
std::random_shuffle uses std::rand() function to randomize the items, while the std::shuffle uses urng which is a better random generator, though with the particular overload of std::random_shuffle, you can get the same behavior (as with the std::shuffle) but that requires you to do some work to pass the third argument.
Watch this talk by Stephan T. Lavavej, in which he explains why std::rand is a bad function and what C++ programmers should use instead:
- rand() Considered Harmful
The gist is, std::shuffle is an improvement over std::random_shuffle, and C++ programmers should prefer using the former.
来源：https://stackoverflow.com/questions/19219726/what-is-the-difference-between-shuffle-and-random-shuffle-c