冒泡排序的优化

待我称王封你为后i 2022-03-08 00:42 204阅读 0赞

public class BubbleSort {
    
    	public static void main(String[] args) throws Exception {
    		int []a = {5,4,6,7,9,8,1,2,3};
    		
    		for(int i:bubbleSort(a)){
    			System.out.print(i+" ");
    		}
    		System.out.println();
    		for(int i:bubbleSort1(a)){
    			System.out.print(i+" ");
    		}
    		System.out.println();
    		for(int i:bubbleSort2(a)){
    			System.out.print(i+" ");
    		}
    	}
    		//初始冒泡排序
    		public static int[] bubbleSort(int[] before){
    			long s1 = System.nanoTime();
    			for(int j = 0;j<before.length-1;j++){
    				for(int i=0;i<before.length-j-1;i++){
    					if(before[i]>before[i+1]){
    					int temp = before[i];
    					before[i] = before[i+1];
    					before[i+1] = temp;
    					}
    				}
    			}
    			long e1 = System.nanoTime();
    			System.out.println("未优化："+(e1-s1));
    			return before;
    		}
    	
    	//优化一：类似123456 98 这类前面已经有序，第一趟已经符合，然后第二趟发现没有改变就说明已经排好序，
    	public static int[] bubbleSort1(int[] before){
    		long s1 = System.nanoTime();
    		boolean change = true;
    		for(int j = 0;j<before.length-1 && change;j++){
    			for(int i=0;i<before.length-j-1;i++){
    				change = false;
    				if(before[i]>before[i+1]){
    				int temp = before[i];
    				before[i] = before[i+1];
    				before[i+1] = temp;
    				change = true;
    				}
    			}
    		}
    		long e1 = System.nanoTime();
    		System.out.println("优化一："+(e1-s1));
    		return before;
    	}
    	
    	//优化二：类似231 456789这类后面已经有序，每趟最后一次交换的位置后面都是有序的
    		public static int[] bubbleSort2(int[] before){
    			long s2 = System.nanoTime();
    			boolean change = true;
    			int k = before.length-1;
    			int judge = 0;
    			for(int j = 0;j<before.length-1 && change;j++){
    				for(int i=0;i<k;i++){
    					change = false;
    					if(before[i]>before[i+1]){
    					int temp = before[i];
    					before[i] = before[i+1];
    					before[i+1] = temp;
    					change = true;
    					judge = i;//一会得到的judge就是后面都大于前面的
    					}
    				}
    				k = judge;
    			}
    			long e2 = System.nanoTime();
    			
    			System.out.println("优化二："+(e2-s2));
    			return before;
    		}
    
    }

## 以下是运行结果，不同排序所需要时间（其中时间以纳秒计算） ##

![在这里插入图片描述][20190313202854981.png]

[20190313202854981.png]: /images/20220308/d7d583536d444443a698acbf343d95c6.png