一道并查集的应用，没什么好说的，具体见代码

#include<bits/stdc++.h>
using namespace std;
int father[10005];
struct node{
    int id,M;
    double AVG=0.0,AVA=0.0;
};
bool cmp(const node&n1,const node&n2){
    return n1.AVA!=n2.AVA?n1.AVA>n2.AVA:n1.id<n2.id;
}
int findFather(int a){
    if (father[a]==a)
        return a;
    int temp=findFather(father[a]);
    father[a]=temp;
    return temp;
}
void UnionSet(int a,int b){
    int ua=findFather(a),ub=findFather(b);
    if (ua!=ub)
    father[max(ua,ub)]=min(ua,ub);
}
int main(){
    for (int i = 0; i <= 10005; ++i) {
        father[i]=i;
    }
    int N;
    cin>>N;
    unordered_map<int,pair<double,double>> m;
    unordered_map<int,node> mp;
    while (N--) {
        int id, fatherId, momId;
        cin >> id >> fatherId >> momId;
        if (fatherId != -1)
            UnionSet(id, fatherId);
        if (momId != -1)
            UnionSet(id, momId);
        int k;
        cin >> k;
        while (k--) {
            int a;
            cin >> a;
            UnionSet(id, a);
        }
        cin >> m[id].first>>m[id].second;
    }
    for (int j = 0; j < 10005; ++j) {
        int temp=findFather(j);
        if (temp!=j)
            mp[temp].M++;
        if (m.find(j)!=m.end()){
            mp[temp].AVG+=m[j].first;
            mp[temp].AVA+=m[j].second;
        }
    }
    vector<node> v;
    for (auto it=mp.begin();it!=mp.end();it++) {
        (it->second).id=it->first;
        ++(it->second).M;
        (it->second).AVA/=(it->second).M;
        (it->second).AVG/=(it->second).M;
        v.push_back(it->second);
    }
    sort(v.begin(),v.end(),cmp);
    printf("%d\n",v.size());
    for (int i = 0; i < v.size(); ++i) {
        printf("%04d %d %.3lf %.3lf\n",v[i].id,v[i].M,v[i].AVG,v[i].AVA);
    }
    return 0;
}