Brainman
Time Limit: 1000MS | Memory Limit: 30000K | |
Total Submissions: 10575 | Accepted: 5489 |
Description
Raymond Babbitt drives his brother Charlie mad. Recently Raymond counted 246 toothpicks spilled all over the floor in an instant just by glancing at them. And he can even count Poker cards. Charlie would love to be able to do cool things like that, too. He wants to beat his brother in a similar task.
Problem
Here's what Charlie thinks of. Imagine you get a sequence of N numbers. The goal is to move the numbers around so that at the end the sequence is ordered. The only operation allowed is to swap two adjacent numbers. Let us try an example:
swap (2 8) 8 2 0 3
swap (2 0) 8 0 2 3
swap (2 3) 8 0 3 2
swap (8 0) 0 8 3 2
swap (8 3) 0 3 8 2
swap (8 2) 0 3 2 8
swap (3 2) 0 2 3 8
swap (3 8) 0 2 8 3
swap (8 3) 0 2 3 8
swap (8 0) 2 0 8 3
swap (2 0) 0 2 8 3
swap (8 3) 0 2 3 8
Input
For every scenario, you are given a line containing first the length N (1 <= N <= 1000) of the sequence,followed by the N elements of the sequence (each element is an integer in [-1000000, 1000000]). All numbers in this line are separated by single blanks.
Output
Sample Input
4 4 2 8 0 3 10 0 1 2 3 4 5 6 7 8 9 6 -42 23 6 28 -100 65537 5 0 0 0 0 0
Sample Output
Scenario #1: 3 Scenario #2: 0 Scenario #3: 5 Scenario #4: 0
Source
1 #include <iostream> 2 #include <stdio.h> 3 using namespace std; 4 const int N=200020; 5 int a[N],b[N]; 6 int main() 7 { 8 int n; 9 scanf("%d",&n); 10 for(int k=1;k<=n;k++) 11 { 12 int m; 13 scanf("%d",&m); 14 for(int i=1;i<=m;i++) 15 scanf("%d",&a[i]); 16 int ans=0; 17 for(int i=1;i<=m;i++) 18 for(int j=i+1;j<=m;j++) 19 if(a[i]>a[j]) 20 ans++; 21 printf("Scenario #%d:\n%d\n\n",k,ans); 22 } 23 return 0; 24 }
第二种归并排序, 对2个已经排好序的数列,进行再排序,只需要把2个数列,从头到尾,按顺序,谁小,谁就先进入tmp数组, 最后tmp数组一定排好序了,然后把TMP数组的元素复制回原数组中即可。
同理,如果我们知道2个子序列的逆序对数量,是否可以通过归并排序一样,求出整体的数量呢?显然是可以的。
这里有一个地方,当左边的数列的a[k]要进tmp数组了, 这个时候,如果右边的指针指向a+mid+p,就说明a[k]比a[mid+1]...a[mid + 2]..a[mid+3].....a[mid+p]都要大!【重要】
也就是说,对于a[k]而言,整个数列中, mid+ mid+2...mid+p都在k后面,同时a[k]比a[mid+1],a[mid+2]...a[mid+p]都要大。 那么显然是增加逆序对数量的。 通过整个方法,计算出整个逆序对的数量即可。
下面给出AC代码:
1 #include <iostream> 2 #include <cstdio> 3 #include <cstdlib> 4 using namespace std; 5 const int max_n = 1000 + 10; 6 7 int n, a[max_n]; 8 int tmp[max_n], ans; 9 10 void merge(int *a, int *tmp, int l, int mid, int r) 11 { 12 if (l >= r) return; 13 int i = l, j = mid + 1, k = 0; 14 int count = 0, flag = 0; 15 while (i <= mid && j <= r) 16 { 17 if (a[i] <= a[j]) 18 { 19 tmp[k ++] = a[i++]; 20 ans += j - mid - 1; 21 }else tmp[k ++ ] = a[j++]; 22 } 23 while (i <= mid) tmp[k ++] = a[i++], ans += r- mid; 24 while (j <= r) tmp[k ++] = a[j++]; 25 for (i = 0; i != k; ++ i) a[l + i] = tmp[i]; 26 } 27 28 void mergesort(int *a, int *tmp, int l, int r) 29 { 30 if (l >= r) return; 31 int mid = (l + r) / 2; 32 mergesort(a, tmp, l, mid); 33 mergesort(a, tmp , mid + 1, r); 34 merge(a, tmp, l, mid, r); 35 } 36 37 int main() 38 { 39 int tt; 40 scanf("%d", &tt); 41 for (int i = 1; i <= tt; ++ i) 42 { 43 cout<<"Scenario #"<<i<<":"<<endl; 44 scanf("%d", &n); 45 ans = 0; 46 for (int i = 0; i != n; ++ i) scanf("%d", &a[i]); 47 mergesort(a, tmp, 0, n - 1); 48 cout<<ans<<endl<<endl; 49 } 50 }
第三种线段树单点更新
1 #include <map> 2 #include <iostream> 3 #include <set> 4 #include <cstdio> 5 #include <cstdlib> 6 using namespace std; 7 8 const int max_n = 1000 + 10; 9 10 int n; 11 int a[max_n], count; 12 map<int, int>G; 13 map<int, int>::iterator it; 14 15 16 17 struct node 18 { 19 int cd, key; 20 int ls, rs; 21 int L, R; 22 node():L(0),R(0),ls(0),rs(0),cd(0),key(0){}; 23 void clear() 24 { 25 cd = key = 0; 26 } 27 }t[max_n * 3]; 28 int tail = 1; 29 30 31 void init() 32 { 33 for (int i = 0; i != max_n * 3; ++ i) t[i].clear(); 34 G.clear(); 35 scanf("%d", &n); 36 for (int i = 0; i != n; ++ i) 37 { 38 scanf("%d", &a[i]); 39 G[a[i]] = 0; 40 } 41 count = 0; 42 for (it = G.begin(); it != G.end(); ++ it) it -> second = ++ count; 43 } 44 45 46 47 void make_tree(int now, int LL, int RR) 48 { 49 t[now].L = LL; 50 t[now].R = RR; 51 if (LL == RR) return; 52 int mid = (LL + RR)/ 2; 53 make_tree(t[now].ls = ++ tail, LL, mid); 54 make_tree(t[now].rs = ++ tail, mid + 1, RR); 55 } 56 57 void tran(int now) 58 { 59 int left_son = t[now].ls, right_son = t[now].rs; 60 t[left_son].cd += t[now].cd; 61 t[right_son].cd += t[now].cd; 62 t[now].key += t[now].cd; 63 t[now].cd = 0; 64 } 65 66 void ins(int now, int LL, int RR) 67 { 68 69 tran(now); 70 if (t[now].L == LL && t[now].R == RR) 71 { 72 t[now].cd ++; 73 return; 74 } 75 t[now].key ++; 76 int mid = (t[now].L + t[now].R) / 2; 77 if (RR <= mid) {ins(t[now].ls, LL, RR); return;} 78 if (mid < LL) {ins(t[now].rs, LL, RR); return;} 79 ins(t[now].ls, LL, mid); 80 ins(t[now].rs, mid + 1, RR); 81 } 82 83 int find(int now, int LL, int RR)//因为题目的特殊性,只会找一个…… 84 { 85 tran(now); 86 if (t[now].L == LL && t[now].R == RR) return t[now].key; 87 int mid = (t[now].L + t[now].R) / 2; 88 if (RR <= mid) return find(t[now].ls, LL, RR); 89 if (mid < LL) return find(t[now].rs, LL, RR); 90 cout<<"wtf?"<<endl; 91 } 92 93 void doit() 94 { 95 int ans=0; 96 for (int i = 0; i != n; ++ i) 97 { 98 int num = G[a[i]]; 99 ans += find(1, num + 1, num + 1); 100 ins(1, 0, num); 101 } 102 cout<<ans<<endl; 103 } 104 105 int main() 106 { 107 int tt; 108 scanf("%d",&tt); 109 make_tree(1, 0, 1002); 110 for (int i = 1; i <= tt; ++ i) 111 { 112 cout<<"Scenario #"<<i<<":"<<endl; 113 init(); 114 doit(); 115 cout<<endl; 116 } 117 }
另外还有几种好办法,贴一下
第四种:树状数组
树状数组, 其实和线段树道理一样。 但是对于树状数组,我会单独开一张好好研究哒。 这里就贴一下速度,并没有比线段树快很多……也许我的写法不好?【如果对树状数组有疑惑,可以看我下一篇文章,我会带领你们好好学会树状数组这个神奇的东西~】
1 #include <cstdio> 2 #include <cstdlib> 3 #include <map> 4 #include <cstring> 5 using namespace std; 6 #define lowbit(k) ((k)&(-k)) 7 8 const int max_n = 1000 + 10; 9 int n, a[max_n], s[max_n]; 10 map<int, int>G; 11 map<int, int>::iterator it; 12 int count; 13 void init() 14 { 15 scanf("%d", &n); 16 G.clear(); 17 count = 1; 18 memset(s, 0, sizeof(s)); 19 for (int i = 0; i != n; ++ i) 20 { 21 scanf("%d", &a[i]); 22 G[a[i]] = 0; 23 } 24 for (it = G.begin(); it != G.end(); ++ it) it -> second = ++ count; 25 } 26 27 void ins(int k) 28 { 29 s[k] += 1; 30 while ((k += lowbit(k)) <= 1000) s[k] += 1; 31 } 32 33 34 int ask(int k)//1..k的和 35 { 36 int tot = s[k]; 37 while (k -= lowbit(k)) tot += s[k]; 38 return tot; 39 } 40 41 42 void doit() 43 { 44 int ans = 0; 45 for (int i = 0; i != n; ++ i) 46 { 47 int num = G[a[i]]; 48 ans += ask(count) - ask(num); 49 ins(num); 50 } 51 printf("%d\n",ans); 52 } 53 54 int main() 55 { 56 int tt; 57 scanf("%d", &tt); 58 for (int i = 1; i <= tt; ++ i) 59 { 60 printf("Scenario #%d:\n",i); 61 init(); 62 doit(); 63 printf("\n"); 64 } 65 }
第五种:平衡树
只要查找,当前在树中,有多少个数字比a[k]要大, 因为是按顺序插入的,所以这个数字的数量就是逆序对的个数
这里有一个小技巧,如果平衡树每次要删的话很麻烦,直接用写成struct,然后新的树就new,最后delete掉即可~
1 #include <iostream> 2 #include <cstdio> 3 #include <cstdlib> 4 using namespace std; 5 const int max_n = 1000 + 10; 6 7 int n; 8 const int maxint = 0x7fffffff; 9 10 struct node 11 { 12 node *c[2]; 13 int key; 14 int size; 15 node():key(0),size(0) 16 { 17 c[0] = c[1] = this; 18 } 19 node(int KEY_, node *a0, node *a1): 20 key(KEY_){c[0] =a0, c[1]=a1;} 21 node* rz(){return size = c[0]->size + c[1]->size + 1, this;} 22 }Tnull, *null=&Tnull; 23 24 struct splay 25 { 26 node *root; 27 splay() 28 { 29 root = (new node(*null)) -> rz(); 30 root -> key = maxint; 31 } 32 void zig(int d) 33 { 34 node *t = root -> c[d]; 35 root -> c[d] = null -> c[d]; 36 null -> c[d] = root; 37 root = t; 38 } 39 void zigzig(int d) 40 { 41 node *t = root -> c[d] -> c[d]; 42 root -> c[d] -> c[d] = null -> c[d]; 43 null -> c[d] = root -> c[d]; 44 root -> c[d] = null -> c[d] -> c[!d]; 45 null -> c[d] -> c[!d] = root -> rz(); 46 root = t; 47 } 48 49 void finish(int d) 50 { 51 node *t = null -> c[d], *p = root -> c[!d]; 52 while (t != null) 53 { 54 t = null -> c[d] -> c[d]; 55 null -> c[d] -> c[d] = p; 56 p = null -> c[d] -> rz(); 57 null -> c[d] = t; 58 } 59 root -> c[!d] = p; 60 } 61 void select(int k)//谁有k个儿子 62 { 63 int t; 64 while (1) 65 { 66 bool d = k > (t = root -> c[0] -> size); 67 if (k == t || root -> c[d] == null) break; 68 if (d) k -= t + 1; 69 bool dd = k > (t = root -> c[d] -> c[0] -> size); 70 if (k == t || root -> c[d] -> c[dd] == null){zig(d); break;} 71 if (dd) k -= t + 1; 72 d != dd ? zig(d), zig(dd) : zigzig(d); 73 } 74 finish(0), finish(1); 75 root -> rz(); 76 } 77 void search(int x) 78 { 79 while (1) 80 { 81 bool d = x > root -> key; 82 if (root -> c[d] == null) break; 83 bool dd = x > root -> c[d] -> key; 84 if (root -> c[d] -> c[dd] == null){zig(d); break;} 85 d != dd ? zig(d), zig(dd) : zigzig(dd); 86 } 87 finish(0), finish(1); 88 root -> rz(); 89 if (x > root -> key) select(root -> c[0] -> size + 1); 90 } 91 92 void ins(int x) 93 { 94 search(x); 95 node *oldroot = root; 96 root = new node(x, oldroot -> c[0],oldroot); 97 oldroot -> c[0] = null; 98 oldroot -> rz(); 99 root -> rz(); 100 } 101 int sel(int k){return select(k - 1), root -> key;} 102 int ran(int x){return search(x), root -> c[0] -> size + 1;} 103 }*sp; 104 105 106 int main() 107 { 108 int tt; 109 scanf("%d", &tt); 110 for (int i = 1; i <= tt; ++ i) 111 { 112 sp = new splay; 113 cout<<"Scenario #"<<i<<":"<<endl; 114 scanf("%d", &n); 115 int ans = 0; 116 int tmp; 117 for (int i = 0; i != n; ++ i) 118 { 119 scanf("%d", &tmp); 120 tmp = - tmp; 121 ans += sp -> ran(tmp) - 1; 122 //cout<<sp.ran(tmp) - 1<<endl; 123 sp -> ins(tmp); 124 } 125 delete sp; 126 cout<<ans<<endl<<endl; 127 } 128 }