题解

#include <cstdio>
#include <vector>
#include <cmath>
using namespace std;

const int N = 5e5 + 10;
const double EPS = 1e-8;
vector<pair<int, double>> G[N];
double q[N], f[N], g[N], ans;

void dfs1(int u, int fa) {
f[u] = 1.0 - q[u];
for (auto &e : G[u]) {
int v = e.first;
double p = e.second;
if (v == fa) continue;
dfs1(v, u);
f[u] *= (1.0 - p + p * f[v]);
}
}

void dfs2(int u, int fa) {
ans += 1.0 - g[u];
for (auto &e : G[u]) {
int v = e.first;
double p = e.second;
if (v == fa) continue;
double deno = 1.0 - p + p * f[v];
if (fabs(deno) < EPS) {
g[v] = f[v] * (1.0 - p);
} else {
double T = g[u] / deno;
g[v] = f[v] * (1.0 - p + p * T);
}
dfs2(v, u);
}
}

int main() {
int n;
scanf("%d", &n);
for (int i = 1; i < n; ++i) {
int a, b, p_val;
scanf("%d%d%d", &a, &b, &p_val);
double prob = p_val / 100.0;
G[a].emplace_back(b, prob);
G[b].emplace_back(a, prob);
}
for (int i = 1; i <= n; ++i) {
int q_val;
scanf("%d", &q_val);
q[i] = q_val / 100.0;
}
dfs1(1, -1);
g[1] = f[1];
dfs2(1, -1);
printf("%.6lf\n", ans);
return 0;
}