unordered_set/unordered_map与map/set 关联式容器的性能测试

unordered系列关联式容器

在C++98中，STL提供了底层为红黑树结构的一系列关联式容器，在查询时效率可达到$lo{g}_{2}N$，即最差情 况下需要比较红黑树的高度次，当树中的节点非常多时，查询效率也不理想。最好的查询是，进行很少的比 较次数就能够将元素找到，
因此在C++11中，STL又提供了4个unordered系列的关联式容器，这四个容器与红黑树结构的关联式容器使用方式基本类似，只是其底层结构不同，即unordered_map和unordered_set，unordered_multimap和unordered_multiset

#define _CRT_SECURE_NO_WARNINGS 1
#include
#include
#include
#include
#include
#include
#include
#include
using namespace std;
  //无序+去重
void test_unordered_set()
{
	unordered_set<int> us;
	us.insert(4);
	us.insert(8);
	us.insert(4);
	us.insert(3);
	us.insert(4);
	us.insert(2);
	us.insert(5);
	us.insert(1);
	unordered_set<int>::iterator it = us.begin();
	while (it != us.end())
	{
		cout << *it << " ";
		it++;
	}
	cout << endl;
	for (auto e : us)
	{
		cout << e << " ";
	} cout << endl;
	auto ret = us.find(5);
	if (ret != us.end())
	{
		cout << "找到了";
	}
}
void test_unordered_map()
{
	unordered_map<string, string> um;
	um.insert(make_pair("sort", "排序"));
	um.insert(make_pair("string", "字符串"));
	um.insert(make_pair("left", "左边"));
	um.insert(make_pair("left", "剩余"));//这个插入失败，key不能重复
	um["insert"];
	um["insert"] = "插入";

	unordered_map<string, string>::iterator it = um.begin();
	while (it != um.end())
	{
		cout << it->first << ": " << it->second << endl;
		it++;
	}
	cout << endl;
	for (const auto e : um)
	{
		cout << e.first << ": " << e.second << endl;
	}
	cout << endl;

}
void test_op()//性能分析
{
	set<int> s;
	unordered_set<int> us;

	srand(time(0));
	vector<int> v;
	const size_t n = 10000000;
	v.reserve(n);
	for (size_t i = 0; i < 10000000; i++)
	{
		v.push_back(rand());
	}
	//insert测性能
	size_t begin1 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		s.insert(v[i]);
	}
	size_t end1 = clock();

	size_t begin2 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		us.insert(v[i]);
	}
	size_t end2 = clock();

	cout <<"set insert:" << end1 - begin1 << endl;
	cout << "unordered_set insert:" << end2 - begin2 << endl;

	//find测性能
	begin1 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		s.find(rand());
	}
	end1 = clock();

	begin2 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		us.find(rand());
	}
	end2 = clock();

	cout << "set find:" << end1 - begin1 << endl;
	cout << "unordered_set find: "<< end2 - begin2 << endl;

	//erase测性能
	begin1 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		s.erase(rand());
	}
	end1 = clock();

	begin2 = clock();
	for (size_t i = 0; i < v.size(); i++)
	{
		us.erase(rand());
	}
	end2 = clock();

	cout << "set erase:" << end1 - begin1 << endl;
	cout << "unordered_set erase: " << end2 - begin2 << endl;
}
int main()
{
	//test_unordered_set();
	//test_unordered_map();
	test_op();
	system("pause");
	return 0;
}

unordered_set删除数据的代价会大一点
其他都会比set好很多

所有unordered_set在性能效率方面会更胜一筹