题目链接:BZOJ - 1875
题目分析:
这道题如果去掉“不会立刻沿着刚刚走来的路走回”的限制,直接用邻接矩阵跑矩阵乘法就可以了。然而现在加了这个限制,建图的方式就要做一些改变。如果我们把每一条边看做点建矩阵,那么每次从一条边出发都只会到其他的边,不能仍然在这条边上“停留”,所以这就可以满足题目的限制。将每条边拆成两条单向边,比如一条编号为 4,一条编号为 5。那么 4^1=5, 5^1=4。这样只要不从第 i 条边走到 i 或 i^1 就可以了。初始的矩阵中以 A 为起点的边到达的方案数为 1 ,其余为 0。最后将终点为 B 的边的方案数累加即为答案。、
这种将边与点灵活转化的思想十分巧妙,应注意。
代码如下:
#include <iostream> #include <cstdio> #include <cstdlib> #include <cstring> #include <cmath> #include <algorithm> using namespace std; const int MaxN = 20 + 5, MaxM = 120 + 5, Mod = 45989; int n, m, t, A, B, TopA, TopB, Index, RT, Ans, a, b; int EA[MaxM], EB[MaxM]; struct Edge { int u, v; Edge() {} Edge(int a, int b) { u = a; v = b; } } E[MaxM]; struct Matrix { int x, y, Num[MaxM][MaxM]; void SetXY(int xx, int yy) { x = xx; y = yy; } void Clear(int nn) { for (int i = 0; i < x; ++i) { for (int j = 0; j < y; ++j) { Num[i][j] = nn; } } } } M0, MZ; Matrix Mul(Matrix A, Matrix B) { Matrix ret; ret.SetXY(A.x, B.y); ret.Clear(0); for (int i = 0; i < ret.x; ++i) { for (int j = 0; j < ret.y; ++j) { for (int k = 0; k < A.y; ++k) { ret.Num[i][j] += A.Num[i][k] * B.Num[k][j]; ret.Num[i][j] %= Mod; } } } return ret; } Matrix Pow(Matrix A, int b) { Matrix ret, f; f = A; ret.SetXY(f.x, f.y); for (int i = 0; i <= ret.x; ++i) ret.Num[i][i] = 1; while (b) { if (b & 1) ret = Mul(ret, f); b >>= 1; f = Mul(f, f); } return ret; } int main() { scanf("%d%d%d%d%d", &n, &m, &t, &A, &B); Index = -1; for (int i = 1; i <= m; ++i) { scanf("%d%d", &a, &b); E[++Index] = Edge(a, b); E[++Index] = Edge(b, a); } MZ.SetXY(m * 2, m * 2); MZ.Clear(0); TopA = TopB = 0; for (int i = 0; i <= Index; ++i) { if (E[i].u == A) EA[++TopA] = i; if (E[i].v == B) EB[++TopB] = i; for (int j = 0; j <= Index; ++j) { if (i != j && i != (j ^ 1) && E[i].v == E[j].u) MZ.Num[i][j] = 1; } } M0.SetXY(1, m * 2); M0.Clear(0); for (int i = 1; i <= TopA; ++i) M0.Num[0][EA[i]] = 1; MZ = Pow(MZ, t - 1); M0 = Mul(M0, MZ); Ans = 0; for (int i = 1; i <= TopB; ++i) { Ans += M0.Num[0][EB[i]]; Ans %= Mod; } printf("%d ", Ans); return 0; }