parallel_sort¶
[algorithms.parallel_sort]
Function template that sorts a sequence.
// Defined in header <tbb/parallel_sort.h>
namespace tbb {
template<typename RandomAccessIterator>
void parallel_sort( RandomAccessIterator begin, RandomAccessIterator end );
template<typename RandomAccessIterator, typename Compare>
void parallel_sort( RandomAccessIterator begin, RandomAccessIterator end, const Compare& comp );
template<typename Container>
void parallel_sort( Container& c );
template<typename Container>
void parallel_sort( Container& c, const Compare& comp );
} // namespace tbb
Requirements:
The
RandomAccessIterator
type must meet the Random Access Iterators requirements from [random.access.iterators] and Swappable requirements from the [swappable.requirements] ISO C++ Standard section.The
Compare
type must meet the Compare type requirements from the [alg.sorting] ISO C++ Standard section.The
Container
type must meet the ContainerBasedSequence requirements which iterators must meet the Random Access Iterators requirements from [random.access.iterators] and Swappable requirements from the [swappable.requirements] ISO C++ Standard section.
Sorts a sequence or a container. The sort is neither stable nor deterministic: relative ordering of elements with equal keys is not preserved and not guaranteed to repeat if the same sequence is sorted again.
A call parallel_sort( begin, end, comp )
sorts the sequence [begin, end) using the argument
comp
to determine relative orderings. If comp( x, y )
returns true
, x appears before
y in the sorted sequence.
A call parallel_sort( begin, end )
is equivalent to parallel_sort( begin, end, comp )
, where comp
uses operator< to determine relative orderings.
A call parallel_sort( c, comp )
is equivalent to parallel_sort( std::begin(c), std::end(c), comp )
.
A call parallel_sort( c )
is equivalent to parallel_sort( c, comp )
, where comp uses operator<
to determine relative orderings.
Complexity
parallel_sort
is a comparison sort with an average time complexity of O(N×log(N)), where N is
the number of elements in the sequence. parallel_sort
may be executed concurrently to improve execution time.