km 算法模版
详见分析:http://blog.csdn.net/sixdaycoder/article/details/47720471
#include <stdio.h> #include <string.h> #define M 310 #define inf 0x3f3f3f3f int n,nx,ny; int link[M],lx[M],ly[M],slack[M]; //lx,ly为顶标,nx,ny分别为x点集y点集的个数 int visx[M],visy[M],w[M][M]; int DFS(int x) { visx[x] = 1; for (int y = 1;y <= ny;y ++) { if (visy[y]) continue; int t = lx[x] + ly[y] - w[x][y]; if (t == 0) // { visy[y] = 1; if (link[y] == -1||DFS(link[y])) { link[y] = x; return 1; } } else if (slack[y] > t) //不在相等子图中slack 取最小的 slack[y] = t; } return 0; } int KM() { int i,j; memset (link,-1,sizeof(link)); memset (ly,0,sizeof(ly)); for (i = 1;i <= nx;i ++) //lx初始化为与它关联边中最大的 for (j = 1,lx[i] = -inf;j <= ny;j ++) if (w[i][j] > lx[i]) lx[i] = w[i][j]; for (int x = 1;x <= nx;x ++) { for (i = 1;i <= ny;i ++) slack[i] = inf; while (1) { memset (visx,0,sizeof(visx)); memset (visy,0,sizeof(visy)); if (DFS(x)) //若成功(找到了增广轨),则该点增广完成,进入下一个点的增广 break; //若失败(没有找到增广轨),则需要改变一些点的标号,使得图中可行边的数量增加。 //方法为:将所有在增广轨中(就是在增广过程中遍历到)的X方点的标号全部减去一个常数d, //所有在增广轨中的Y方点的标号全部加上一个常数d int d = inf; for (i = 1;i <= ny;i ++) if (!visy[i]&&d > slack[i]) d = slack[i]; for (i = 1;i <= nx;i ++) if (visx[i]) lx[i] -= d; for (i = 1;i <= ny;i ++) //修改顶标后,要把所有不在交错树中的Y顶点的slack值都减去d if (visy[i]) ly[i] += d; else slack[i] -= d; } } int res = 0; for (i = 1;i <= ny;i ++) if (link[i] > -1) res += w[link[i]][i]; return res; } int main () { int i,j; while (scanf ("%d",&n)!=EOF) { nx = ny = n; // memset (w,0,sizeof(w)); for (i = 1;i <= n;i ++) for (j = 1;j <= n;j ++) scanf ("%d",&w[i][j]); int ans = KM(); printf ("%d\n",ans); } return 0; }