python无序链表删除重复项的方法
人气:0题目描述:
给定一个没有排序的链表,去掉重复项,并保留原顺序 如: 1->3->1->5->5->7,去掉重复项后变为:1->3->5->7
方法:
- 顺序删除
- 递归删除
1.顺序删除
由于这种方法采用双重循环对链表进行遍历,因此,时间复杂度为O(n**2)
在遍历链表的过程中,使用了常数个额外的指针变量来保存当前遍历的结点,前驱结点和被删除的结点,所以空间复杂度为O(1)
#!/usr/bin/env python3 # -*- coding: utf-8 -*- # @Time : 2020/1/15 20:55 # @Author : buu # @Software: PyCharm # @Blog :https://blog.csdn.net/weixin_44321080 class LNode: def __init__(self, x): self.data = x self.next = None def removeDup(head): """ 对带头结点的无序单链表删除重复的结点 顺序删除:通过双重循环直接在链表上进行删除操作 即,外层循环用一个指针从第一个结点开始遍历整个链表,内层循环从外层指针指向的下一个结点开始, 遍历其余结点,将与外层循环遍历到的的指针所指的结点的数据域相同的结点删除 :param head: 头指针 :return: """ if head is None or head.next is None: return outerCur = head.next innerCur = None innerPre = None while outerCur is not None: innerCur = outerCur.next innerPre = outerCur while innerCur is not None: if outerCur.data == innerCur.data: innerPre.next = innerCur.next innerCur = innerCur.next else: innerPre = innerCur innerCur = innerCur.next outerCur = outerCur.next if __name__ == '__main__': i = 1 head = LNode(6) tmp = None cur = head while i < 7: if i % 2 == 0: tmp = LNode(i + 1) elif i % 3 == 0: tmp = LNode(i - 2) else: tmp = LNode(i) cur.next = tmp cur = tmp i += 1 print("before removeDup:") cur = head.next while cur is not None: print(cur.data, end=' ') cur = cur.next removeDup(head) print("\nafter removeDup:") cur = head.next while cur is not None: print(cur.data, end=' ') cur = cur.next
结果:
2.递归
此方法与方法一类似,从本质上而言,由于这种方法需要对链表进行双重遍历,所以时间复杂度为O(n**2)
由于递归法会增加许多额外的函数调用,所以从理论上讲,该方法效率比方法一低
#!/usr/bin/env python3 # -*- coding: utf-8 -*- # @Time : 2020/1/15 21:30 # @Author : buu # @Software: PyCharm # @Blog :https://blog.csdn.net/weixin_44321080 class LNode: def __init__(self, x): self.data = x self.next = None def removeDupRecursion(head): """ 递归法:将问题逐步分解为小问题,即,对于结点cur,首先递归地删除以cur.next为首 的子链表中重复的结点;接着删除以cur为首的链表中的重复结点, :param head: :return: """ if head.next is None: return head pointer = None cur = head head.next = removeDupRecursion(head.next) pointer = head.next while pointer is not None: if head.data == pointer.data: cur.next = pointer.next pointer = cur.next else: pointer = pointer.next cur = cur.next return head def removeDup(head): """ 对带头结点的单链表删除重复结点 :param head: 链表头结点 :return: """ if head is None: return head.next = removeDupRecursion(head.next) if __name__ == '__main__': i = 1 head = LNode(6) tmp = None cur = head while i < 7: if i % 2 == 0: tmp = LNode(i + 1) elif i % 3 == 0: tmp = LNode(i - 2) else: tmp = LNode(i) cur.next = tmp cur = tmp i += 1 print("before recursive removeDup:") cur = head.next while cur is not None: print(cur.data, end=' ') cur = cur.next removeDup(head) print("\nafter recurseve removeDup:") cur = head.next while cur is not None: print(cur.data, end=' ') cur = cur.next
结果:
引申:从有序链表中删除重复项
上述介绍的方法也适用于链表有序的情况,但是由于上述方法没有充分利用到链表有序这个条件,因此,算法的性能肯定不是最优的。本题中,由于链表具有有序性,因此不需要对链表进行两次遍历。所以有如下思路:
用cur指向链表的第一个结点,此时需要分为以下两种情况讨论:
- 如果cur.data == cur.next.data,则删除cur.next结点;
- 如果cur.data != cur.next.data,则cur=cur.next,继续遍历其余结点;
总结
以上所述是小编给大家介绍的python无序链表删除重复项的方法,希望对大家有所帮助,如果大家有任何疑问请给我留言,小编会及时回复大家的。在此也非常感谢大家对网站的支持!
如果你觉得本文对你有帮助,欢迎转载,烦请注明出处,谢谢!
您可能感兴趣的文章:
加载全部内容