算法竞赛专题讲座结课作业

题目一

算法分类:并查集,DFS,Tarjan算法

原题:

How far away ?

Time Limit: 2000/1000 MS (Java/Others)

Memory Limit: 32768/32768 K (Java/Others)

 

Description

There are n houses in the village and some bidirectional roads connecting them. Every day peole always like to ask like this "How far is it if I want to go from house A to house B"? Usually it hard to answer. But luckily int this village the answer is always unique, since the roads are built in the way that there is a unique simple path("simple" means you can't visit a place twice) between every two houses. Yout task is to answer all these curious people.

Input

First line is a single integer T(T<=10), indicating the number of test cases.
  For each test case,in the first line there are two numbers n(2<=n<=40000) and m (1<=m<=200),the number of houses and the number of queries. The following n-1 lines each consisting three numbers i,j,k, separated bu a single space, meaning that there is a road connecting  house i and house j,with length k(0<k<=40000).The houses are labeled from 1 to n.
  Next m lines each has distinct integers i and j, you areato answer the distance between house i and house j.

Output

For each test case,output m lines. Each line represents the answer of the query. Output a bland line after each test case.
 

Sample Input

2 3 2 1 2 10 3 1 15 1 2 2 3 2 2 1 2 100 1 2 2 1

Sample Output

10 25 100 100
 
题目大意:
   现有n个村庄,村庄之间通过公路相连,并且任意两个村庄之间只有一条最简路径,也就是不重复经过其中的任意一条路,每条路有其对应的长度,现给出村庄个数,已经任意两两村庄之间存在的公路及公路的长度,需要你求出任意两个村庄之间的距离
 
题目实质:
    一个n个节点带边权值的无环图,求图中任意两个节点的最近公共祖先
 
题解:
   首先先来看一看我第一遍自己做的时候TLE的做法。当时因为从来没有接触过LCA的相关问题,所以采取了一种非常暴力的做法。
    确定了这条题目是一个n个节点的无环图之后,我的想法是确定任意一个点为根节点,然后通过并查集并且按秩合并,将每个节点所在的高度,都求出来。如确定1这个点为根节点,那么这个点的高度为0,接着所有与1直接相连的节点的高度为1,依次往下类推。开一个数组,把每一条边的长度都保存在高度的数值较大的那个点上。
    紧接着,需要查询两个节点之间的距离,如果两个节点在同一条路径上,那么只要将从一个节点到另一节点之间的所有边长相加即可。如果不在同一条路径上,那么只要将分别将两个节点到最近公共祖先的距离求出再相加即可。
    那么,首先判断两个点是否在同一个高度,如果不在的话,将高度较低的点向上移动直到高度相等为止,在过程中顺便加上经过边的边长。高度相等以后,判断两点是否重合,如果重合就结束,如果不重合,就将两个点每次同时向上移动一个高度,直到两个点重合为止。
TLE的代码:
  
 1 #include <bits/stdc++.h>
 2 using namespace std;
 3 const int maxn=1e5+5;
 4 int s[maxn],height[maxn],val[maxn];
 5 int n,m;
 6 inline int read(){
 7     register int x=0,f=1;char c=getchar();
 8     while(c<'0'||c>'9'){if(c=='-') f=-1;c=getchar();}
 9     while(c>='0'&&c<='9') x=(x<<3)+(x<<1)+(c^48),c=getchar();
10     return x*f;
11 }
12 int find_set(int x){                   
13     return x==s[x]?x:find_set(s[x]);
14 }
15 void init_set(){   
16    for(int i = 1; i <= n; i++){
17         s[i] = i;
18         height[i]=0;       
19    }
20 }
21 void union_set(int x1, int y1,int l){             //将联通的两个节点按秩合并 
22     int x = find_set(x1);
23     int y = find_set(y1);
24     if (height[x] == height[y]) {
25         height[x] = height[x] + 1; 
26         s[y1] = x1;
27         val[y1]=l;       
28     }
29     else{
30         if (height[x]<height[y]) s[x1]=y1,val[x1]=l;
31         else  s[y1] = x1,val[y1]=l;
32     }
33     if(s[y1]==x1)    {
34         int t=y1;
35         while(t!=s[t]){
36             height[s[t]]=max(height[s[t]],height[t]+1);
37             t=s[t];
38         }
39     }
40     else{
41         int t=x1;
42         while(t!=s[t]){
43             height[s[t]]=max(height[s[t]],height[t]+1);
44             t=s[t];
45         }
46     }
47 }
48 int main(){
49     int t;
50     t=read();
51     while(t--){
52     n=read();m=read();
53     n-=1;
54     init_set();
55     while(n--){
56         int x,y,v;
57         x=read();
58         y=read();
59         v=read();
60         union_set(x,y,v);
61     }
62     while(m--){
63         int x,y,k;
64         x=read();y=read();
65         long long cnt=0;
66         if(height[y]>height[x]){                    //将不在同一高度的两个节点移动至同一个高度 
67             while(height[x]!=height[y]){
68                 cnt+=val[x];
69                 x=s[x];
70             }
71         }
72         else if(height[y]<height[x]){
73             while(height[y]!=height[x]){
74                 cnt+=val[y];
75                 y=s[y];
76             }
77         }
78         else{}
79         if(x==y){                                        
80             printf("%lld\n",cnt);continue;            //如果两点已经重合,输出结果 
81         }
82         else{                                        //将两点同时上移,直至两点重合 
83             while(x!=y){
84                 cnt+=val[x]+val[y];
85                 x=s[x];
86                 y=s[y];
87             }
88         }
89         printf("%lld\n",cnt);
90         }
91     }
92     return 0;
93 }

      虽然这道题目只有4e4的数据,并且给了两秒的时间,但是这样暴力的做法还是超时了。自己做了不少改进还是没有任何效果。所以这道题目得推翻一开始的做法重新来。

   想了好久没有思路,后来通过几篇博客,了解到了Tarjan算法。

   先来说一说Tarjan算法:

      Tarjan算法实际上是一种离线算法。所谓离线,就是通过一次遍历一次性将所有询问结果进行计算并进行保存。最后不需要再回到图中计算直接输出结果。其时间复杂度为O(n+q);

  n为所有节点的总个数,q为所有询问的个数。

        Tarjin算法的基本流程大概是:

  1.任选一个节点为根节点,以这个节点为起点

 

  2.遍历该点u所有子节点v,并标记这些子节点v已被访问过。

  3.若是v还有子节点,返回2,否则下一步。

  4.将v的祖先记为u。

  5.寻找与当前点u有询问关系的点v。若是v已经被访问过了,则可以确认u和v的最近公共祖先为v被合并到的父亲节点。如果v 没有被访问,则继续下一步操作

 

  其实画一个图自己来手动模拟会显得更加直观

  现在有这样一个图:                                                                                                                           

 

   

初始化:    

fa[1]=1,vis[1]=false              现在有两组点需要询问,要求这两个点的最近公共祖先

fa[2]=2,vis[2]=false                分别为点2和点4,点8与点5

fa[3]=3,vis[3]=false

fa[4]=4,vis[4]=false

fa[5]=5,vis[5]=false

fa[6]=6,vis[6]=false

fa[7]=7,vis[7]=false

fa[8]=8,vis[8]=false

 
   下面开始模拟的过程,首先以点1为根节点,发现点1有两个子节点,点2和点3,先对点2进行操作,
  发现点2有两个子节点,点4和点5。发现点4有子节点点8,再对点8进行操作,点8没有子节点,查看询问中
  与8有关系的点为点5,但是vis[5]=false,点5还未搜索到,于是跳过,vis[8]=true。将点8合并到它的父节点点4上
 
        vis[8]=false --> vis[5]=false  -->vis[8]=true -->fa[8]=4;
        接着再对8的父节点4进行操作,查看询问中与4有关系的点为点2,但是vis[2]=false,点2还未搜索到,于是跳过,
  vis[4]=true。将点4合并到它的父节点点2上
   vis[4]=false  -->vis[2]=false  -->vis[4]=true  -->fa[4]=2;
        接下来,再对点4的父节点2进行操作,查看询问中与点2 有关系的节点为点4,并且vis[4]=true,所以节点2和
  节点4的最近公共祖先为find(4)=2,答案为2,并且更新vis[2]=true;
  int find(int i) {return fa[i]==i?i:find(fa[i])}
  vis[2]=false -->vis[4]=true  --> ans=find(4)  -->vis[2]=true;
 
 
 
   
   下面继续搜点2的另一个子节点点5,发现点5没有子节点,于是查看询问中与点5有关系的节点为点8,发现vis[8]=true,
  于是答案为find(8)=2,于是点5和点8的最近公共祖先为点2。更新vis(5)=true;
   fa[8]=4  -->fa[4]=2 -->fa[2]=2  -->ans=2
    vis[5]=false  -->vis[8]=true  -->ans=find(8)=2 -->vis[5]=true;
 
依次类推,我们可以通过一次DFS,按照分支的顺序依次得到所有询问的答案。知道搜索完根节点点1的所有分支之后,我们便可以退出DFS。
 
下面上代码:
 1 #include <bits/stdc++.h>
 2 using namespace std;
 3 const int N=40000+5;
 4 struct Edge{
 5     int cnt,x[N],y[N],z[N],nxt[N],fst[N];
 6     void set(){
 7         cnt=0;
 8         memset(x,0,sizeof x);
 9         memset(y,0,sizeof y);
10         memset(z,0,sizeof z);
11         memset(nxt,0,sizeof nxt);
12         memset(fst,0,sizeof fst);
13     }
14     void add(int a,int b,int c){
15         x[++cnt]=a;
16         y[cnt]=b;
17         z[cnt]=c;
18         nxt[cnt]=fst[a];
19         fst[a]=cnt;
20     }
21 }e,q;
22 int T,n,m,from,to,dist,in[N],rt,dis[N],fa[N],ans[N];
23 bool vis[N];
24 void dfs(int rt){
25     for (int i=e.fst[rt];i;i=e.nxt[i]){
26         dis[e.y[i]]=dis[rt]+e.z[i];
27         dfs(e.y[i]);
28     }
29 }
30 int getf(int k){
31     return fa[k]==k?k:fa[k]=getf(fa[k]);
32 }
33 void LCA(int rt){
34     for (int i=e.fst[rt];i;i=e.nxt[i]){
35         LCA(e.y[i]);
36         fa[getf(e.y[i])]=rt;
37     }
38     vis[rt]=1;
39     for (int i=q.fst[rt];i;i=q.nxt[i])
40         if (vis[q.y[i]]&&!ans[q.z[i]])
41             ans[q.z[i]]=dis[q.y[i]]+dis[rt]-2*dis[getf(q.y[i])];
42 }
43 int main(){
44     scanf("%d",&T);
45     while (T--){
46         q.set(),e.set();
47         memset(in,0,sizeof in);
48         memset(vis,0,sizeof vis);
49         memset(ans,0,sizeof ans);
50         scanf("%d%d",&n,&m);
51         for (int i=1;i<n;i++)
52             scanf("%d%d%d",&from,&to,&dist),e.add(from,to,dist),in[to]++;
53         for (int i=1;i<=m;i++)
54             scanf("%d%d",&from,&to),q.add(from,to,i),q.add(to,from,i);
55         rt=0;
56         for (int i=1;i<=n&&rt==0;i++)
57             if (in[i]==0)
58                 rt=i;
59         dis[rt]=0;
60         dfs(rt);
61         for (int i=1;i<=n;i++)
62             fa[i]=i;
63         LCA(rt);
64         for (int i=1;i<=m;i++)
65             printf("%d\n",ans[i]);
66     }
67     return 0;
68 }

 

 

 

题目二

算法分类:递推DP,组合数

原题:HDU4489

 

The King’s Ups and Downs

Time Limit: 2000/1000 MS (Java/Others)

Memory Limit: 32768/32768 K (Java/Others)

 

Description

The king has guards of all   different heights. Rather than line them up in increasing or  decreasing height order, he wants to line them up so each guard   is either shorter than the guards next to him or taller than the guards next to him (so the heights go up and  down along the line). For example, seven guards of heights 160, 162, 164, 166, 168, 170   and   172 cm. could be arranged as:


or perhaps:


The king wants to know how many guards he needs so he can have a   different up and down order at each changing of the guard for rest of his reign. To be able to do this, he needs to know for a given number of guards, n, how many  different up and  down orders there are:

For example, if there are four guards: 1, 2, 3,4 can be arrange   as:

1324, 2143, 3142, 2314, 3412, 4231, 4132, 2413, 3241, 1423

For this problem, you will write a program that takes as input a positive integer n, the number of guards and returns the number of up and down orders for n guards of   differing heights.

 

Input

The first line of input contains a single integer P, (1 <= P <= 1000), which is the number of  data sets that follow.  Each  data set consists of single line of input containing two integers.  The first integer, D is the   data set number.  The second  integer, n (1 <= n <= 20), is the number of guards of   differing heights.

 

Output

For each   data set there is one line of output.  It contains the   data set number (D) followed  by a single space, followed by the number of up and down orders for the n guards.

 

Sample Input

4
1 1
2 3
3 4
4 20

 

Sample Output

1 1
2 4
3 10
4 740742376475050
 
题目大意:
  国王想要给他的士兵排队,现在有n个士兵,身高各不相同,国王想要使他的士兵按照“高矮高矮高矮......”或者“矮高矮高矮高......”的顺序排成一队,现在想要求出对于n个士兵,存在多少种不同的排列方法。n最大为20。
 
题解:
  这道题目n最大只有20,但当n达到20的时候,答案已经是一个很大很大的数了,显然是一个指数级的增长,暴力的做法肯定跑不出来。想到使用动态规划。
   现在我们先给n个士兵按照身高从矮到高排序,按照从矮到高的顺序依次插入每个士兵。对于每次插入的士兵i,他的身高一定高于已经在队列中的i-1个士兵中的任意一个。对于
士兵i放入的每一个位置,在他前面的士兵一定是以“高矮”结尾,而在他后面的士兵则一定是“矮高”开头,所以如果将士兵i放在第j个位置,那么对于的排列方式便为前面的j个士兵以“高矮”结尾的方法数,乘以后面的(i-1-j)个士兵以“矮高”开头的方法数,由于j个士兵到底是哪j个士兵并没有确定,所以还要乘上从所有的i-1个士兵中选取j个士兵的方法数Ci-1j,这样将j从0到i-1的所有情况相加,便可以得到排列i个士兵的总方法数。
 
  对于组合数的计算,我们只需要用一个最基础的递推式即可,即Cij=Ci-1j-1+ci-1j。代码如下。
1 c[1][0]=c[1][1]=1;
2     for(int i=2;i<=20;i++){
3         c[i][0]=c[i][i]=1;
4         for(int j=1;j<i;j++){
5             c[i][j]=c[i-1][j-1]+c[i-1][j];
6         }
7     }

 

  我们可以用两个dp数组,dp1和dp2分别表示以“高矮”结尾的方法数和以“矮高”开头的方法数。由于对于n个士兵排列的总方法数,以“高矮”和“矮高”开头的方法数各占总数的一半,同理以“高矮”结尾的方法数也一样。所以dp1[n]和dp2[n]都等于ans[n]的一半。

  递推方程为:

    ans[i]=dp1[j]*dp2[i-1-j]*c[i-1][j]  (0<=j<i)

  当然如果只用一个dp数组也是一样,

    dp[i]=(dp[j]/2)*(dp[i-1-j]/2)*c[i-1][j]  (0<=j<i)

  这里n最大只有20,所以直接一次性打表把所有答案存下来,避免后面重复计算。注意答案的数值可能很大,所以要用long long。

上代码:

 

 1 #include <bits/stdc++.h>
 2 using namespace std;
 3 typedef long long ll;
 4 ll c[22][22];
 5 ll dp1[25],dp2[25];
 6 #define scan(i) scanf("%d",&i)
 7 int main(){
 8     int N;scan(N);
 9     c[1][0]=c[1][1]=1;
10     for(int i=2;i<=20;i++){
11         c[i][0]=c[i][i]=1;
12         for(int j=1;j<i;j++){
13             c[i][j]=c[i-1][j-1]+c[i-1][j];
14         }
15     }
16     dp1[1]=dp1[0]=dp2[0]=dp2[1]=1;
17     for(int i=2;i<=20;i++){
18         ll t=0;
19         for(int j=0;j<i;j++){
20             t+=dp1[j]*dp2[i-1-j]*c[i-1][j];
21         }
22         dp1[i]=dp2[i]=t/2;
23     }
24     while(N--){
25         int x,y;
26         scanf("%d%d",&x,&y);
27         if(y==1){
28             printf("%d 1\n",x);
29             continue;
30         }
31         ll ans=dp1[y]*2;
32         printf("%d %lld\n",x,ans);
33     }
34     return 0; 
35 }

 

 

 

题目三

题目来源:CF#552(div3)

G题:数论

G. Minimum Possible LCM

time limit per test
4 seconds
memory limit per test
1024 megabytes
input
standard input
output
standard output

 

You are given an array aa consisting of nn integers a1,a2,,ana1,a2,…,an .

Your problem is to find such pair of indices i,ji,j (1i<jn1≤i<j≤n ) that lcm(ai,aj)lcm(ai,aj) is minimum possible.

lcm(x,y)lcm(x,y) is the least common multiple of xx and yy (minimum positive number such that both xx and yy are divisors of this number).

 

Input

The first line of the input contains one integer nn (2n1062≤n≤106 ) — the number of elements in aa .

The second line of the input contains nn integers a1,a2,,ana1,a2,…,an (1ai1071≤ai≤107 ), where aiai is the ii -th element of aa .

 

Output

Print two integers ii and jj (1i<jn1≤i<j≤n ) such that the value of lcm(ai,aj)lcm(ai,aj) is minimum among all valid pairs i,ji,j . If there are multiple answers, you can print any.

 

Examples
Input
 
5
2 4 8 3 6
Output
 
1 2
Input
 
5
5 2 11 3 7
Output
 
2 4
Input
 
6
2 5 10 1 10 2
Output
 
1 4

 

题意:

    给出n个数,要求输出最小公倍数最大的两个数的下标

知识补充:GCD和LCM:

    GCD即最大公约数,求最大公约数有多种方法,常见的有质因数分解法、短除法、辗转相除法、更相减损法。

     algorithm库里有函数__gcd()

     欧几里得(辗转相除法):

    

 1 int gcd(int a,int b)
 2 
 3 {
 4 
 5     if(b==0) return a;
 6 
 7     return gcd(b,a%b);
 8 
 9 }
10 
11 
12 
13 
14 int gcd(int x,int y){return y==0?x:GCD(x%y)}

 

 

    更相减损术:

 1 while(!(a%2) && !(b%2))
 2 {
 3         a = a/2;
 4         b = b/2;
 5 }
 6 while(a != b)
 7 {
 8         if(a>b){
 9             a = a-b;
10         }else{
11             b = b-a;
12         }
13 }

 

    stein算法:Stein算法只有整数的移位和加减法。原理如下:

  • 若a和b都是偶数,则记录下公约数2,然后都除2(即右移1位);
  • 若其中一个数是偶数,则偶数除2,因为此时2不可能是这两个数的公约数了
  • 若两个都是奇数,则a = |a-b|,b = min(a,b),因为若d是a和b的公约数,那么d也是|a-b|和min(a,b)的公约数。
 1 int SteinGCD(int a, int b) {
 2     if (a < b) { int t = a; a = b; b = t; }
 3     if (b == 0) return a;
 4     if ((a & 1) == 0 && (b & 1) == 0)
 5         return SteinGCD(a >> 1, b >> 1) << 1;
 6     else if ((a & 1) == 0 && (b & 1) != 0)
 7         return SteinGCD(a >> 1, b);
 8     else if ((a & 1) != 0 && (b & 1) == 0)
 9         return SteinGCD(a, b >> 1);
10     else
11         return SteinGCD(a - b, b);
12 }

 

    LCM即最小公倍数

    有如下公式 LCM(a,b)*GCD(a,b)=a*b;通过分解分解质因数可以很容易的证明。

 1 int GCD(int x,int y){
 2 
 3     return !y?x:GCD(y,x%y);
 4 
 5 }
 6 
 7 int LCM(int x,int y){
 8 
 9     return x*y/GCD(x,y);
10 
11 }

 

题解:   

    首先,将每个出现的位置记在pos[]数组里,考虑到,存在lcm(x,x)==x的情形,在读入的时候特判一下再考虑到,lcm(x,y)==x*y/gcd(x,y)枚举i==gcd(x,y),对于每个i,找到其倍数里出现的最小的两个x、y,这个是埃筛的操作,j+=i的时候会遍历到i的倍数,取到两个就break就好,毕竟更大的x、y,对于固定下来的gcd来讲,答案不会更优。
 1 #include <bits/stdc++.h>
 2 using namespace std;
 3 typedef long long ll;
 4 const int maxn=1e7+10;
 5 int n,v;
 6 int pos[maxn];
 7 ll ans;
 8 int ai,aj; 
 9 int main(){
10     scanf("%d",&n);
11     ans=1e18;
12     for(int i=1;i<=n;i++){
13         scanf("%d",&v);
14         if(pos[v]&&ans>v)
15         ans=v,ai=pos[v],aj=i;
16         pos[v]=i;
17     }
18     for(ll i=1;i<maxn;i++)
19     {
20         ll first=-1;
21         for(ll j=i;j<maxn;j+=i)
22         {
23             if(pos[j])
24             {
25                 if(first==-1)first=j;
26                 else 
27                 {
28                    if(ans>first*j/i)
29                    ans=first*j/i,ai=pos[first],aj=pos[j];
30                    break;
31                 }
32             }
33         }
34     }
35     if(ai>aj){
36         int temp=ai;
37         ai=aj;
38         aj=temp;
39     }
40     printf("%d %d\n",ai,aj);
41     return 0;
42 }

 

E题:

    

E. Two Teams
time limit per test
2 seconds
memory limit per test
256 megabytes
input
standard input
output
standard output

There are nn students standing in a row. Two coaches are forming two teams — the first coach chooses the first team and the second coach chooses the second team.

The ii -th student has integer programming skill aiai . All programming skills are distinct and between 11 and nn , inclusive.

Firstly, the first coach will choose the student with maximum programming skill among all students not taken into any team, and kk closest students to the left of him and kk closest students to the right of him (if there are less than kk students to the left or to the right, all of them will be chosen). All students that are chosen leave the row and join the first team. Secondly, the second coach will make the same move (but all students chosen by him join the second team). Then again the first coach will make such move, and so on. This repeats until the row becomes empty (i. e. the process ends when each student becomes to some team).

Your problem is to determine which students will be taken into the first team and which students will be taken into the second team.

Input

The first line of the input contains two integers nn and kk (1kn21051≤k≤n≤2⋅105 ) — the number of students and the value determining the range of chosen students during each move, respectively.

The second line of the input contains nn integers a1,a2,,ana1,a2,…,an (1ain1≤ai≤n ), where aiai is the programming skill of the ii -th student. It is guaranteed that all programming skills are distinct.

Output

Print a string of nn characters; ii -th character should be 1 if ii -th student joins the first team, or 2 otherwise.

Examples
Input
5 2
2 4 5 3 1
Output
11111
Input
5 1
2 1 3 5 4
Output
22111
Input
Copy
7 1
7 2 1 3 5 4 6
Output
1121122
Input
5 1
2 4 5 3 1
Output
21112
Note

In the first example the first coach chooses the student on a position 33 , and the row becomes empty (all students join the first team).

In the second example the first coach chooses the student on position 44 , and the row becomes [2,1][2,1] (students with programming skills [3,4,5][3,4,5] join the first team). Then the second coach chooses the student on position 11 , and the row becomes empty (and students with programming skills [1,2][1,2] join the second team).

In the third example the first coach chooses the student on position 11 , and the row becomes [1,3,5,4,6][1,3,5,4,6] (students with programming skills [2,7][2,7] join the first team). Then the second coach chooses the student on position 55 , and the row becomes [1,3,5][1,3,5] (students with programming skills [4,6][4,6] join the second team). Then the first coach chooses the student on position 33 , and the row becomes [1][1] (students with programming skills [3,5][3,5] join the first team). And then the second coach chooses the remaining student (and the student with programming skill 11 joins the second team).

In the fourth example the first coach chooses the student on position 33 , and the row becomes [2,1][2,1] (students with programming skills [3,4,5][3,4,5] join the first team). Then the second coach chooses the student on position 11 , and the row becomes empty (and students with programming skills [1,2][1,2] join the second team).

 

题意:

     现有n个人排成一队,两个教练轮流从队中挑选最高的人以及最高的人左边k个人和右边k个人,现在需要你求出每个人分别被哪一个教练挑选走。

题解:

     这条题目其实就是一条模拟题,唯一要说的就是如何快速找到当前队中存在的最大值,以及最大值所对应的位置。

    用一个check数组保存每一个值的状态,这样在寻找最大值的时候,可以从n开始,如果这个值已经被用过了,就减一再判断,直到找到当前未用的最大值。然后用position数组保存每一个值的下标,方便索引。

代码:

 1 #include <bits/stdc++.h>
 2 using namespace std;
 3 const int maxn=2e5+5;
 4 int n,k,x;
 5 struct node{
 6     int val;
 7     int id;
 8     node *pre;
 9     node *next;
10 };
11 node p[maxn];
12 int ans[maxn];
13 int position[maxn];          
14 bool check[maxn];            
15 inline void build(){
16     for(int i=1;i<=n;i++){
17         scanf("%d",&x);
18         p[i].val=x;
19         p[i].id=i;
20         p[i].pre=&p[i-1];p[i].next=&p[i+1];
21         if(i==1)p[i].pre=NULL;
22         if(i==n)p[i].next=NULL;
23         position[x]=i;
24     }
25     return;
26 }
27 int main(){
28     scanf("%d%d",&n,&k);
29     build();
30     int ma=n;
31     int an=1;
32     while(ma!=0){
33         check[ma]=true;
34         ans[position[ma]]=an;
35         int po=position[ma];
36         int k1=k;int k2=k;
37         int p1=po,p2=po;
38         while(k1--){
39             if(p[po].next==NULL)break;
40             int v=p[po].next->val; 
41             check[v]=true;
42             ans[position[v]]=an;
43             p1=p[po].next->id;
44             p[po].next=p[po].next->next;
45         }
46         while(k2--){
47             if(p[po].pre==NULL)break;
48             int v=p[po].pre->val; 
49             check[v]=true;
50             ans[position[v]]=an;
51             p2=p[po].pre->id;
52             p[po].pre=p[po].pre->pre;    
53         }
54         if(p[p1].next!=NULL)p1=p[p1].next->id;
55         else p1=n+1;
56         if(p[p2].pre!=NULL)p2=p[p2].pre->id;
57         else p2=0;
58         if(p1<=n&&p2>=1){
59             p[p1].pre=&p[p2];
60             p[p2].next=&p[p1];
61         }
62         if(p1<=n&&p2<1)p[p1].pre=NULL;
63         if(p1>n&&p2>=1)p[p2].next=NULL;
64         if(an==1)an=2;else an=1;
65         while(check[ma])ma--;
66     }
67     for(int i=1;i<=n;i++)
68         printf("%d",ans[i]);
69     return 0;
70 }

 

 
posted @ 2019-08-01 07:37  dongdong222  阅读(267)  评论(0编辑  收藏  举报