求解一下这个ci zhe ge k f和d办法课程哦从这里开始看大家反馈日常n热闹
双循环遍历时间复杂度是O(N²)。
可以先排序——用归并排序时间复杂度是O(Nlog(N))——再剔除,因为排序后相同的都相邻,一次循环就可以全部剔除,所以剔除的时间复杂度是O(N)。于是总体时间复杂度为O(Nlog(N))。
#include<stdio.h>
#include<malloc.h>
#include<stdlib.h>
typedef struct student_test {
int val;
struct student_test*next;
} node;
//类似于strcmp
int cmp_test(node*a, node*b) {
return a->val - b->val;
}
node* next_n(node*p, int n) {
while (p && n--) {
p = p->next;
}
return p;
}
//两个有序链表合并
void lmerge(node **h, node **t, node *h1, node *e1, node *h2, node *e2, int (*cmp)(node*, node*)) {
node h3, *p, *q;
*t = &h3;
for (p = h1, q = h2; p != e1 && q != e2;) {
if (cmp(p, q)<=0) {
(*t)->next = p, *t = p, p = p->next;
} else {
(*t)->next = q, *t = q, q = q->next;
}
}
while (p != e1) (*t)->next = p, *t = p, p = p->next;
while (q != e2) (*t)->next = q, *t = q, q = q->next;
(*t)->next = NULL;
(*h) = h3.next;
}
//链表归并排序,时间复杂度为O(nlog(n)),空间复杂度为O(1)
void lsort(node **head, int (*cmp)(node*, node*)) {
node *t, *l, *m, *r, hh;
int w;
for (w = 1; next_n(*head, w); w <<= 1) {
t = &hh;
for (l = *head; l; l = r) {
m = next_n(l, w);
r = next_n(m, w);
if (!m)
while (l)t->next = l, t = l, l = l->next;
else
lmerge(&t->next, &t, l, m, m, r, cmp);
}
*head = hh.next;
//t->next = NULL;
}
}
int main() {
node head={0,NULL},*p,*q;
int n,i,count=0;
scanf("%d",&n);
p=&head;
for(i=0;i<n;i++){
p->next=(node*)malloc(sizeof(node));
p=p->next;
scanf("%d",&p->val);
}
p->next=NULL;
lsort(&head.next,cmp_test);
printf("被删除的元素:");
for(p=head.next;p!=NULL;p=q){//看起来像双循环,实际上只遍历了一次
for(q=p->next;q!=NULL && p->val==q->val;q=p->next){
printf("%d ",q->val);
p->next=q->next;
free(q);
count++;
}
}
printf("\n删除个数:%d",count);
for(p=head.next;p!=NULL;){
q=p->next;
free(p);
p=q;
}
return 0;
}
双循环遍历吧