#pragma GCC optimize ("Ofast")
#include<bits/stdc++.h>
using namespace std;
void *wmem;
char memarr[96000000];
template<class S, class T> inline S max_L(S a,T b){
  return a>=b?a:b;
}
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);
}
inline int my_getchar_unlocked(){
  static char buf[1048576];
  static int s = 1048576;
  static int e = 1048576;
  if(s == e && e == 1048576){
    e = fread_unlocked(buf, 1, 1048576, stdin);
    s = 0;
  }
  if(s == e){
    return EOF;
  }
  return buf[s++];
}
inline void rd(int &x){
  int k;
  int m=0;
  x=0;
  for(;;){
    k = my_getchar_unlocked();
    if(k=='-'){
      m=1;
      break;
    }
    if('0'<=k&&k<='9'){
      x=k-'0';
      break;
    }
  }
  for(;;){
    k = my_getchar_unlocked();
    if(k<'0'||k>'9'){
      break;
    }
    x=x*10+k-'0';
  }
  if(m){
    x=-x;
  }
}
struct MY_WRITER{
  char buf[1048576];
  int s;
  int e;
  MY_WRITER(){
    s = 0;
    e = 1048576;
  }
  ~MY_WRITER(){
    if(s){
      fwrite_unlocked(buf, 1, s, stdout);
    }
  }
}
;
MY_WRITER MY_WRITER_VAR;
void my_putchar_unlocked(int a){
  if(MY_WRITER_VAR.s == MY_WRITER_VAR.e){
    fwrite_unlocked(MY_WRITER_VAR.buf, 1, MY_WRITER_VAR.s, stdout);
    MY_WRITER_VAR.s = 0;
  }
  MY_WRITER_VAR.buf[MY_WRITER_VAR.s++] = a;
}
inline void wt_L(char a){
  my_putchar_unlocked(a);
}
inline void wt_L(int x){
  int s=0;
  int m=0;
  char f[10];
  if(x<0){
    m=1;
    x=-x;
  }
  while(x){
    f[s++]=x%10;
    x/=10;
  }
  if(!s){
    f[s++]=0;
  }
  if(m){
    my_putchar_unlocked('-');
  }
  while(s--){
    my_putchar_unlocked(f[s]+'0');
  }
}
int WRITER_DOUBLE_DIGIT = 15;
inline int writerDigit_double(){
  return WRITER_DOUBLE_DIGIT;
}
inline void writerDigit_double(int d){
  WRITER_DOUBLE_DIGIT = d;
}
inline void wt_L(double x){
  const int d = WRITER_DOUBLE_DIGIT;
  int k;
  int r;
  double v;
  if(x!=x || (x==x+1 && x==2*x)){
    my_putchar_unlocked('E');
    my_putchar_unlocked('r');
    my_putchar_unlocked('r');
    return;
  }
  if(x < 0){
    my_putchar_unlocked('-');
    x = -x;
  }
  x += 0.5 * pow(0.1, d);
  r = 0;
  v = 1;
  while(x >= 10*v){
    v *= 10;
    r++;
  }
  while(r >= 0){
    r--;
    k = floor(x / v);
    if(k >= 10){
      k = 9;
    }
    if(k <= -1){
      k = 0;
    }
    x -= k * v;
    v *= 0.1;
    my_putchar_unlocked(k + '0');
  }
  if(d > 0){
    my_putchar_unlocked('.');
    v = 1;
    for(r=(0);r<(d);r++){
      v *= 0.1;
      k = floor(x / v);
      if(k >= 10){
        k = 9;
      }
      if(k <= -1){
        k = 0;
      }
      x -= k * v;
      my_putchar_unlocked(k + '0');
    }
  }
}
template<class T> inline T pow2_L(T a){
  return a*a;
}
template <class T> struct DijkstraHeap{
  int *hp;
  int *place;
  int size;
  char *visited;
  T *val;
  void malloc(int N){
    hp = (int*)std::malloc(N*sizeof(int));
    place = (int*)std::malloc(N*sizeof(int));
    visited = (char*)std::malloc(N*sizeof(char));
    val = (T*)std::malloc(N*sizeof(T));
  }
  void free(){
    std::free(hp);
    std::free(place);
    std::free(visited);
    std::free(val);
  }
  void walloc(int N, void **mem=&wmem){
    walloc1d(&hp, N, mem);
    walloc1d(&place, N, mem);
    walloc1d(&visited, N, mem);
    walloc1d(&val, N, mem);
  }
  void init(int N){
    int i;
    size = 0;
    for(i=(0);i<(N);i++){
      place[i]=-1;
    }
    for(i=(0);i<(N);i++){
      visited[i]=0;
    }
  }
  void up(int n){
    int m;
    while(n){
      m=(n-1)/2;
      if(val[hp[m]]<=val[hp[n]]){
        break;
      }
      swap(hp[m],hp[n]);
      swap(place[hp[m]],place[hp[n]]);
      n=m;
    }
  }
  void down(int n){
    int m;
    for(;;){
      m=2*n+1;
      if(m>=size){
        break;
      }
      if(m+1<size&&val[hp[m]]>val[hp[m+1]]){
        m++;
      }
      if(val[hp[m]]>=val[hp[n]]){
        break;
      }
      swap(hp[m],hp[n]);
      swap(place[hp[m]],place[hp[n]]);
      n=m;
    }
  }
  void change(int n, T v){
    if(visited[n]||(place[n]>=0&&val[n]<=v)){
      return;
    }
    val[n]=v;
    if(place[n]==-1){
      place[n]=size;
      hp[size++]=n;
      up(place[n]);
    }
    else{
      up(place[n]);
    }
  }
  int pop(void){
    int res=hp[0];
    place[res]=-1;
    size--;
    if(size){
      hp[0]=hp[size];
      place[hp[0]]=0;
      down(0);
    }
    visited[res]=1;
    return res;
  }
}
;
struct graph{
  int N;
  int *es;
  int **edge;
  void setEdge(int N__, int M, int A[], int B[], void **mem = &wmem){
    int i;
    N = N__;
    walloc1d(&es, N, mem);
    walloc1d(&edge, N, mem);
    for(i=(0);i<(N);i++){
      es[i] = 0;
    }
    for(i=(0);i<(M);i++){
      es[A[i]]++;
      es[B[i]]++;
    }
    for(i=(0);i<(N);i++){
      walloc1d(&edge[i], es[i], mem);
    }
    for(i=(0);i<(N);i++){
      es[i] = 0;
    }
    for(i=(0);i<(M);i++){
      edge[A[i]][es[A[i]]++] = B[i];
      edge[B[i]][es[B[i]]++] = A[i];
    }
  }
}
;
template<class T> struct wgraph{
  int N;
  int *es;
  int **edge;
  T **cost;
  graph g;
  void setEdge(int N__, int M, int A[], int B[], T C[], void **mem = &wmem){
    int i;
    N = N__;
    walloc1d(&es, N, mem);
    for(i=(0);i<(N);i++){
      es[i] = 0;
    }
    for(i=(0);i<(M);i++){
      es[A[i]]++;
      es[B[i]]++;
    }
    walloc1d(&edge, N, mem);
    for(i=(0);i<(N);i++){
      walloc1d(&edge[i], es[i], mem);
    }
    walloc1d(&cost, N, mem);
    for(i=(0);i<(N);i++){
      walloc1d(&cost[i], es[i], mem);
    }
    for(i=(0);i<(N);i++){
      es[i] = 0;
    }
    for(i=(0);i<(M);i++){
      edge[A[i]][es[A[i]]] = B[i];
      edge[B[i]][es[B[i]]] = A[i];
      cost[A[i]][es[A[i]]++] = C[i];
      cost[B[i]][es[B[i]]++] = C[i];
    }
    g.N = N;
    g.es = es;
    g.edge = edge;
  }
}
;
int N;
int M;
int K;
int S;
int T;
int X[200000];
int Y[200000];
int A[200000];
int B[200000];
double c[200000];
double dis[200000];
wgraph<double> g;
int sz;
int ok;
int done;
double cost[10];
vector<int> path[10];
vector<int> cpath;
priority_queue<pair<double,vector<int>>> q;
DijkstraHeap<double> hp;
int bk[2000];
vector<int> rt;
int main(){
  wmem = memarr;
  int i;
  int j;
  int k;
  int b;
  int cr;
  rd(N);
  rd(M);
  rd(K);
  rd(S);S += (-1);
  rd(T);T += (-1);
  {
    int Lj4PdHRW;
    for(Lj4PdHRW=(0);Lj4PdHRW<(N);Lj4PdHRW++){
      rd(X[Lj4PdHRW]);
      rd(Y[Lj4PdHRW]);
    }
  }
  {
    int e98WHCEY;
    for(e98WHCEY=(0);e98WHCEY<(M);e98WHCEY++){
      rd(A[e98WHCEY]);A[e98WHCEY] += (-1);
      rd(B[e98WHCEY]);B[e98WHCEY] += (-1);
    }
  }
  for(i=(0);i<(M);i++){
    c[i] = sqrt((pow2_L((X[A[i]]-X[B[i]])))+(pow2_L((Y[A[i]]-Y[B[i]]))));
  }
  g.setEdge(N,M,A,B,c);
  hp.malloc(N);
  for(;;){
    if(q.size()==0 && ok){
      break;
    }
    if(q.size()){
      cpath = q.top().second;
      for(i=(0);i<(ok);i++){
        if(cpath == path[i]){
          q.pop();
          goto WYIGIcGE;
        }
      }
      cost[ok] = -q.top().first;
      path[ok] = cpath;
      ok++;
      if(ok==K){
        break;
      }
    }
    for(b=(0);b<(max_L(1, (int)cpath.size()-1));b++){
      hp.init(N);
      hp.change(S,0);
      cr = 0;
      bk[S] = -1;
      while(hp.size){
        i = hp.pop();
        for(j=(0);j<(g.es[i]);j++){
          k = g.edge[i][j];
          if(cpath.size() && cr==b){
            int x;
            if(k==cpath[cr+1]){
              continue;
            }
            for(x=(0);x<(ok-1);x++){
              int y;
              for(y=(0);y<(b+1);y++){
                if(path[x][y] != cpath[y]){
                  goto jbtyPBGc;
                }
              }
              if(k==path[x][cr+1]){
                goto YREPHmFM;
              }
              jbtyPBGc:;
            }
          }
          if(cpath.size() && cr<b && k != cpath[cr+1]){
            continue;
          }
          if(hp.visited[k]){
            continue;
          }
          if(hp.place[k]==-1 || hp.val[k] > hp.val[i]+g.cost[i][j]){
            bk[k] = i;
            hp.change(k,hp.val[i]+g.cost[i][j]);
          }
          YREPHmFM:;
        }
        cr++;
      }
      if(hp.visited[T]==0){
        continue;
      }
      rt.clear();
      i = T;
      while(i >= 0){
        rt.push_back(i);
        i = bk[i];
      }
      reverse(rt.begin(), rt.end());
      q.push(make_pair(-hp.val[T], rt));
    }
    WYIGIcGE:;
  }
  for(i=(0);i<(K);i++){
    if(i < ok){
      wt_L(cost[i]);
      wt_L('\n');
    }
    else{
      wt_L(-1);
      wt_L('\n');
    }
  }
  return 0;
}
// cLay varsion 20200509-1

// --- original code ---
// int N, M, K, S, T, X[2d5], Y[2d5], A[2d5], B[2d5];
// double c[2d5], dis[2d5];
// wgraph<double> g;
// 
// int sz, ok, done;
// double cost[10];
// vector<int> path[10], cpath;
// 
// priority_queue<pair<double,vector<int>>> q;
// 
// DijkstraHeap<double> hp;
// int bk[2000];
// vector<int> rt;
// 
// {
//   int i, j, k, b, cr;
// 
//   rd(N,M,K,S--,T--,(X,Y)(N),(A--,B--)(M));
//   rep(i,M) c[i] = sqrt( (X[A[i]]-X[B[i]])**2 + (Y[A[i]]-Y[B[i]])**2 );
//   g.setEdge(N,M,A,B,c);
// 
//   hp.malloc(N);
// 
//   for(;;){
//     if(q.size()==0 && ok) break;
//     if(q.size()){
//       cpath = q.top().second;
//       rep(i,ok) if(cpath == path[i]){
//         q.pop();
//         break_continue;
//       }
//       cost[ok] = -q.top().first;
//       path[ok] = cpath;
// //      wt(cost[ok],":",cpath(cpath.size()));
//       ok++;
//       if(ok==K) break;
//     }
//     rep(b,max(1,(int)cpath.size()-1)){
//       hp.init(N);
//       hp.change(S,0);
//       cr = 0;
//       bk[S] = -1;
//       while(hp.size){
//         i = hp.pop();
//         rep(j,g.es[i]){
//           k = g.edge[i][j];
//           if(cpath.size() && cr==b){
//             if(k==cpath[cr+1]) continue;
//             rep(x,ok-1){
//               rep(y,b+1) if(path[x][y] != cpath[y]) break_continue;
//               if(k==path[x][cr+1]) break_continue;
//             }
//           }
//           if(cpath.size() && cr<b && k != cpath[cr+1]) continue;
//           if(hp.visited[k]) continue;
//           if(hp.place[k]==-1 || hp.val[k] > hp.val[i]+g.cost[i][j]){
//             bk[k] = i;
//             hp.change(k,hp.val[i]+g.cost[i][j]);
//           }
//         }
//         cr++;
//       }
//       if(hp.visited[T]==0) continue;
//       rt.clear();
//       i = T;
//       while(i >= 0){
//         rt.push_back(i);
//         i = bk[i];
//       }
//       reverse(rt.begin(), rt.end());
//       q.push(make_pair(-hp.val[T], rt));
// //      wt(b,":",cpath(cpath.size()),":",rt(rt.size()),":",hp.val[T]);
//     }
//   }
// 
//   rep(i,K) wt(if[i < ok, cost[i], -1]);
// }