count_sort计数排序OpenMP的并行化

版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/a19990412/article/details/85162006

简述

计数排序，就是统计某个数值在所有的数字中所应该存在的位置，然后，放到一个确定的位置上。非常简单的排序算法。

程序

• 会读取data.txt中的文件
• 数据的样子

10 
0 2 3 1 4 8 6 7 5 9

• 效果

PS D:\C++\VS\repo\OpenMP-TEST\Debug> ./OpenMP-TEST
0 1 2 3 4 5 6 7 8 9

• 代码

#include <iostream>
#include <omp.h>
#include <fstream>
using namespace std;
#pragma warning(disable : 4996)
int main(int argc, char **argv) {
	if (argc < 1) return 0;
	ifstream cin("./data.txt");
	int i, j, count, n;
	cin >> n;
	int *a = new int[n];
	int *temp = new int[n];
	
	//int thread_count = strtol(argv[1], NULL, 10);
	for (i = 0; i < n; ++i)  cin >> a[i];

#pragma omp parallel for private(i, j, count) shared(n, a, temp)
	for (i = 0; i < n; ++i) {
		count = 0;
		for (j = 0; j < n; ++j) {
			if (a[j] < a[i]) count++;
			else if (a[j] == a[i] && j < i) count++;
		}
		temp[count] = a[i];
	}

#pragma omp parallel for 
	for (i = 0; i < n; ++i) a[i] = temp[i];

	for (i = 0; i < n; ++i) cout << a[i]<< " ";
	cout << endl;

	delete[]temp;
	delete[]a;
}

和串行的计数排序进行对比

• 不要输出排序结果，因为数据量太大了。
• 10000个数据的排序效果

PS D:\C++\VS\repo\OpenMP-TEST\Debug> ./OpenMP-TEST r
Total time Serial: 1.256s
Total time parallel: 0.141s

• 接近9倍的速度（因为我在笔记本上有8个核）
• 后面多了一个r命令，其实是免得在vs编译的时候被调用。随便写什么都是ok的。

#include <iostream>
#include <omp.h>
#include <fstream>
#include <ctime>
using namespace std;
#pragma warning(disable : 4996)
int main(int argc, char **argv) {
	if (argc < 2) return 0;
	ifstream cin("./data.txt");
	int i, j, count, n;
	cin >> n;
	int *a = new int[n];
	int *b = new int[n];
	int *temp = new int[n];
	
	//int thread_count = strtol(argv[1], NULL, 10);
	for (i = 0; i < n; ++i) cin >> b[i];
	for (i = 0; i < n; ++i) a[i] = b[i];
	
	clock_t starttime, endtime;
	starttime = clock();
	for (i = 0; i < n; ++i) {
		count = 0;
		for (j = 0; j < n; ++j) {
			if (b[j] < b[i]) count++;
			else if (b[j] == b[i] && j < i) count++;
		}
		temp[count] = b[i];
	}
	memcpy(b, temp, n * sizeof(int));
	endtime = clock();

	// for (i = 0; i < n; ++i) cout << b[i] << " ";
	// cout << endl;
	cout << "Total time Serial: " << (double)(endtime - starttime) / CLOCKS_PER_SEC << "s" << endl;

	starttime = clock();
#pragma omp parallel for private(i, j, count) shared(n, a, temp)
	for (i = 0; i < n; ++i) {
		count = 0;
		for (j = 0; j < n; ++j) {
			if (a[j] < a[i]) count++;
			else if (a[j] == a[i] && j < i) count++;
		}
		temp[count] = a[i];
	}

#pragma omp parallel for 
	for (i = 0; i < n; ++i) a[i] = temp[i];

	endtime = clock();

	//for (i = 0; i < n; ++i) cout << a[i]<< " ";
	//cout << endl;
	cout << "Total time parallel: " << (double)(endtime - starttime) / CLOCKS_PER_SEC << "s" << endl;
	delete[]temp;
	delete[]a;
	delete[]b;
}

和qsort的串行版也做对比

#include <iostream>
#include <omp.h>
#include <fstream>
#include <ctime>
using namespace std;
#pragma warning(disable : 4996)
int cmp(const void * a, const void *b) //from small to big
{
	return *(int *)a - *(int *)b;
}

int main(int argc, char **argv) {
	if (argc < 2) return 0;
	ifstream cin("./data.txt");
	int i, j, count, n;
	cin >> n;
	int *a = new int[n];
	int *b = new int[n];
	int *c = new int[n];
	int *temp = new int[n];
	
	//int thread_count = strtol(argv[1], NULL, 10);
	for (i = 0; i < n; ++i) cin >> b[i];
	for (i = 0; i < n; ++i) { a[i] = b[i]; c[i] = b[i]; }
	

	clock_t starttime, endtime;
	starttime = clock();
	qsort(c, n, sizeof(int), cmp);
	endtime = clock();
	cout << "Total time Serial-qsort: " << (double)(endtime - starttime) / CLOCKS_PER_SEC << "s" << endl;

	starttime = clock();
	for (i = 0; i < n; ++i) {
		count = 0;
		for (j = 0; j < n; ++j) {
			if (b[j] < b[i]) count++;
			else if (b[j] == b[i] && j < i) count++;
		}
		temp[count] = b[i];
	}
	memcpy(b, temp, n * sizeof(int));
	endtime = clock();
	cout << "Total time Serial: " << (double)(endtime - starttime) / CLOCKS_PER_SEC << "s" << endl;

	starttime = clock();
#pragma omp parallel for private(i, j, count) shared(n, a, temp)
	for (i = 0; i < n; ++i) {
		count = 0;
		for (j = 0; j < n; ++j) {
			if (a[j] < a[i]) count++;
			else if (a[j] == a[i] && j < i) count++;
		}
		temp[count] = a[i];
	}

#pragma omp parallel for 
	for (i = 0; i < n; ++i) a[i] = temp[i];

	endtime = clock();

	//for (i = 0; i < n; ++i) cout << a[i]<< " ";
	//cout << endl;
	cout << "Total time parallel: " << (double)(endtime - starttime) / CLOCKS_PER_SEC << "s" << endl;
	delete[]temp;
	delete[]a;
}