STL 源码剖析 算法 stl_algo.h -- nth_element

本文为senlie原创，转载请保留此地址：http://blog.csdn.net/zhengsenlie

nth_element
------------------------------------------------------------------------------


描述：重新排序，使得[nth,last)内没有任何一个元素小于[first,nth)内的元素，
但对于[first,nth)和[nth,last)两个子区间内的元素次序则无任何保证。
思路：
1.以 median-of-3-partition 将整个序列分割为更小的左、右子序列
2.如果 nth 迭代器落于左序列，就再对左子序列进行分割，否则就再对右子序列进行分割
3.直到分割后的子序列长大于3，对最后这个待分割的子序列做 Insertion Sort
图6-17
复杂度：O(n)
源码：

template <class RandomAccessIterator>
inline void nth_element(RandomAccessIterator first, RandomAccessIterator nth,
                        RandomAccessIterator last) {
  __nth_element(first, nth, last, value_type(first));
}


template <class RandomAccessIterator, class T>
void __nth_element(RandomAccessIterator first, RandomAccessIterator nth,
                   RandomAccessIterator last, T*) {
  while (last - first > 3) {
    //采用 median-of-3-partition 。参数：(first,last,pivot)
	//返回一个迭代器，指向分割后的右段第一个元素
	RandomAccessIterator cut = __unguarded_partition
      (first, last, T(__median(*first, *(first + (last - first)/2),
                               *(last - 1))));
    if (cut <= nth) //如果  nth 落于右段，再对右段实施分割
      first = cut;
    else  //如果 nth 落于左段，对左段实施分割
      last = cut;
  }
  __insertion_sort(first, last); //对分割后的子序列做 Insertion Sort
}


template <class RandomAccessIterator, class T>
RandomAccessIterator __unguarded_partition(RandomAccessIterator first, 
                                           RandomAccessIterator last, 
                                           T pivot) {
  while (true) {
    while (*first < pivot) ++first;
    --last;
    while (pivot < *last) --last;
    if (!(first < last)) return first;
    iter_swap(first, last);
    ++first;
  }
}    

int A[] = {7, 2, 6, 11, 9, 3, 12, 10, 8, 4, 1, 5};
const int N = sizeof(A) / sizeof(int);


nth_element(A, A + 6, A + N);
copy(A, A + N, ostream_iterator<int>(cout, " "));
// The printed result is "5 2 6 1 4 3 7 8 9 10 11 12".

STL 源码剖析 算法 stl_algo.h -- nth_element,布布扣,bubuko.com

STL 源码剖析 算法 stl_algo.h -- nth_element

原文：http://blog.csdn.net/zhengsenlie/article/details/37995027