矩阵求逆的并行算法.doc

.矩阵求逆的并行算法：#include stdio.h#include stdlib.h#include mpi.h #define intsize sizeof(int)#define floatsize sizeof(float)#define A(x,y) Ax*N+y#define Q(x,y) Qx*N+y#define a(x,y) ax*N+y#define f(x) fxfloat *A,*Q;float *a,*f;int N,v,m;int p;int myid;MPI_Status status;FILE *dataFile;double starttime,endtime,time1; void readData() int i,j; starttime = MPI_Wtime(); dataFile = fopen(dataIn.txt,r); fscanf(dataFile,%d,&N); A = (float *)malloc(floatsize*N*N); for(i = 0; i N; i+) for(j = 0; j N; j+) fscanf(dataFile,%f,&A(i,j); fclose(dataFile); printf(Input of file dataIn.txtn); printf(%dn,N); for(i = 0; i N; i+) for(j = 0; j N; j+) printf(%ft,A(i,j); printf(n); Q = (float *)malloc(floatsize*N*N); void printResult() int i,j; printf(nOutput of Matrix Qn); for(i = 0; i N; i+) for(j = 0; j N; j+) printf(%ft,Q(i,j); printf(n); endtime = MPI_Wtime(); printf(n); printf(Whole running time = %f secondsn,endtime-starttime); printf(Distribute data time = %f secondsn,time1-starttime); printf(Parallel compute time = %f secondsn,endtime-time1); dataFile=fopen(dataOut.txt, w); fprintf(dataFile, Input of file dataIn.txtn); fprintf(dataFile, %dn,N); for(i=0;iN;i+) for(j=0; jN; j+) fprintf(dataFile, %ft, A(i,j); fprintf(dataFile, n); fprintf(dataFile, nOutput of Matrix As inversionn); for(i=0; iN; i+) for(j=0; jN; j+) fprintf(dataFile, %ft, Q(i,j); fprintf(dataFile, n); /* 0号进程将时间统计写入目标文件 */ fprintf(dataFile, n); fprintf(dataFile, Whole running time = %f secondsn,endtime-starttime); fprintf(dataFile, Distribute data time = %f secondsn,time1-starttime); fprintf(dataFile, Parallel compute time = %f secondsn,endtime-time1); fprintf(dataFile, Parallel Process number = %dn, p); fclose(dataFile); void broadcast(int i,int j,int v) int k; if(myid = j) a(i,v) = (float)(1 / a(i,v); for(k = 0; k N; k+) if(k != v) a(i,k) = a(i,k) * a(i,v); for(k = 0; k N; k+) f(k) = a(i,k); MPI_Bcast(&f(0),N,MPI_FLOAT,myid,MPI_COMM_WORLD); else MPI_Bcast(f,N,MPI_FLOAT,j,MPI_COMM_WORLD); void transform(int i,int j,int v) int k,w; if(myid != j) for(k = 0; k m; k+) for(w = 0; w N; w+) if(w != v) a(k,w) = a(k,w) - f(w) * a(k,v); for(k = 0; k m; k+) a(k,v) = -f(v) * a(k,v); if(myid = j) for(k = 0; k m; k+) if(k != i) for(w = 0; w N; w+) if(w != v) a(k,w) = a(k,w) - f(w) * a(k,v); for(k = 0; k m; k+) if(k != i) a(k,v) = -f(v) * a(k,v); int main(int argc,char *argv) int i,j,k,w,group_size; MPI_Init(&argc,&argv); MPI_Comm_size(MPI_COMM_WORLD,&group_size); MPI_Comm_rank(MPI_COMM_WORLD,&myid); p = group_size; if(myid = 0) readData(); m = N/p; if(myid = 0) for(i = 1; i p; i+) MPI_Send(&N,1,MPI_INT,i,i,MPI_COMM_WORLD); MPI_Send(&m,1,MPI_INT,i,i,MPI_COMM_WORLD); if(myid != 0) MPI_Recv(&N,1,MPI_INT,0,myid,MPI_COMM_WORLD,&status); MPI_Recv(&m,1,MPI_INT,0,myid,MPI_COMM_WORLD,&status); a = (float *)malloc(floatsize*m*N); f = (float *)malloc(floatsize*N); if(a = NULL | f = NULL) printf(Allocate space for a or f fail!); if(myid = 0) for(i = 0; i m; i+) for(j = 0; j N; j+) a(i,j) = A(p*i,j); for(k = 1; k p; k+) for(i = 0; i 0)&(myid p) MPI_Recv(&a(0,0),m*N,MPI_FLOAT,0,myid,MPI_COMM_WORLD,&status); if(myid p) for(i = 0; i m; i+) for(j = 0; j 0)&(myid p) MPI_Send(a,m*N,MPI_FLOAT,0,myid,MPI_COMM_WORLD); if(myid = 0) for(i = 0; i m; i+) for(j = 0; j N; j+) Q(p*i,j)=a(i,j); for(k = 1;