Submission #7328671

---

Source Code Expand

#pragma GCC optimize ("Ofast")
#include<bits/stdc++.h>
using namespace std;
void *wmem;
template<class T> inline void walloc1d(T **arr, int x, void **mem = &wmem){
  static int skip[16]={0, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1};
  (*mem) = (void*)( ((char*)(*mem)) + skip[((unsigned long long)(*mem)) & 15] );
  (*arr)=(T*)(*mem);
  (*mem)=((*arr)+x);
}
void sortF_L(int N, unsigned A[], void *mem = wmem){
  int bt, i, m, *sz;
  unsigned *arr, c;
  if(N < 256){
    sort(A, A+N);
    return;
  }
  bt = sizeof(unsigned) * 8;
  walloc1d(&arr, N, &mem);
  walloc1d(&sz, N, &mem);
  for(m=0;m<bt;m+=8){
    for(i=0;i<(257);i++){
      sz[i] = 0;
    }
    for(i=0;i<(N);i++){
      sz[ 1+((A[i]>>m)&255u) ]++;
    }
    for(i=(1);i<(257);i++){
      sz[i] += sz[i-1];
    }
    for(i=0;i<(N);i++){
      c = ((A[i]>>m)&255u);
      arr[sz[c]++] = A[i];
    }
    swap(A, arr);
  }
}
void sortF_L(int N, int A[], void *mem = wmem){
  int *arr, i, x, y, z;
  unsigned *send;
  if(N < 256){
    sort(A, A+N);
    return;
  }
  send = (unsigned*)A;
  sortF_L(N, send, mem);
  if(A[0] < 0 || A[N-1] >= 0){
    return;
  }
  x = 0;
  y = N;
  while(x < y){
    z = (x+y) / 2;
    if(A[z] < 0){
      y = z;
    }
    else{
      x = z+1;
    }
  }
  walloc1d(&arr, N, &mem);
  z = 0;
  for(i=(x);i<(N);i++){
    arr[z++] = A[i];
  }
  for(i=0;i<(x);i++){
    arr[z++] = A[i];
  }
  for(i=0;i<(N);i++){
    A[i] = arr[i];
  }
}
void sortF_L(int N, unsigned long long A[], void *mem = wmem){
  int bt, i, m, *sz;
  unsigned c;
  unsigned long long *arr;
  if(N < 512){
    sort(A, A+N);
    return;
  }
  bt = sizeof(unsigned long long) * 8;
  walloc1d(&arr, N, &mem);
  walloc1d(&sz, N, &mem);
  for(m=0;m<bt;m+=8){
    for(i=0;i<(257);i++){
      sz[i] = 0;
    }
    for(i=0;i<(N);i++){
      sz[ 1+((A[i]>>m)&255u) ]++;
    }
    for(i=(1);i<(257);i++){
      sz[i] += sz[i-1];
    }
    for(i=0;i<(N);i++){
      c = ((A[i]>>m)&255u);
      arr[sz[c]++] = A[i];
    }
    swap(A, arr);
  }
}
void sortF_L(int N, long long A[], void *mem = wmem){
  int i, x, y, z;
  long long *arr;
  unsigned long long *send;
  if(N < 512){
    sort(A, A+N);
    return;
  }
  send = (unsigned long long*)A;
  sortF_L(N, send, mem);
  if(A[0] < 0 || A[N-1] >= 0){
    return;
  }
  x = 0;
  y = N;
  while(x < y){
    z = (x+y) / 2;
    if(A[z] < 0){
      y = z;
    }
    else{
      x = z+1;
    }
  }
  walloc1d(&arr, N, &mem);
  z = 0;
  for(i=(x);i<(N);i++){
    arr[z++] = A[i];
  }
  for(i=0;i<(x);i++){
    arr[z++] = A[i];
  }
  for(i=0;i<(N);i++){
    A[i] = arr[i];
  }
}
inline void rd(int &x){
  int k, m=0;
  x=0;
  for(;;){
    k = getchar_unlocked();
    if(k=='-'){
      m=1;
      break;
    }
    if('0'<=k&&k<='9'){
      x=k-'0';
      break;
    }
  }
  for(;;){
    k = getchar_unlocked();
    if(k<'0'||k>'9'){
      break;
    }
    x=x*10+k-'0';
  }
  if(m){
    x=-x;
  }
}
inline void wt_L(char a){
  putchar_unlocked(a);
}
inline void wt_L(double x){
  printf("%.15f",x);
}
char memarr[96000000];
int N;
int K;
int A[100000];
int main(){
  double res=0;
  int i;
  wmem = memarr;
  rd(N);
  rd(K);
  {
    int Lj4PdHRW;
    for(Lj4PdHRW=0;Lj4PdHRW<(N);Lj4PdHRW++){
      rd(A[Lj4PdHRW]);
    }
  }
  sortF_L(N,A);
  for(i=0;i<(K);i++){
    res += (double)(A[i] + A[N-K+i]) / A[N-K+i];
  }
  wt_L(res);
  wt_L('\n');
  return 0;
}
// cLay varsion 20190902-1

// --- original code ---
// int N, K, A[1d5];
// {
//   double res = 0;
//   
//   rd(N,K,A(N));
//   sortF(N,A);
// 
//   rep(i,K) res += (double)(A[i] + A[N-K+i]) / A[N-K+i];
//   wt(res);
// }

Submission Info

Judge Result