int main()
{
int n;
int *array;
int i,j,k,p;
double st,et;
FILE *insertomp;
double res_time[40];
p=0;
for(j=0;j<20;j++)
{
n=pow(2,j);
array = (int *)malloc(n*sizeof(int));
for (i=0; i<n; i++)
array[i] = i;
for (i=0; i<n; i++) {
int num1 = rand() % n;
int num2 = rand() % n;
int t = array[num1];
array[num1] = array[num2];
array[num2] = t;
}
st=timerval();
for(k=0;k<1000;k++)
{
merge(array,n);
}
et=timerval();
res_time[p]=n;
res_time[p+1] = et-st;
p=p+2;
}
insertomp= fopen("insertomp.txt","w+");
for(j=0;j<40;j=j+2)
{
fprintf(insertomp,"\n n=%f time=%f seconds ",res_time[j],res_time[j+1]);
}
/* Freeing Allocated Memory */
free(array);
fclose(insertomp);
return 0;
}
int main(int argc, char*argv[]) {
int p, q, m, n, i,j,k;
double c, d,row, column;
long long int N = atoi(argv[1]);
long long int M = atoi(argv[2]);
double st, et;
st = timerval ();
k = 0;
for (i=0; i<M; i++)
{
printf("\n");
for (j=0; j<N; j++)
{
//for (k=0; k<4; k++)
{
k++;
in[i][j] = k;
printf("\t in[%d][%d] = %d", i,j,k);
}
}
}
#pragma omp parallel //for schedule(static) num_threads(8)
{
#pragma omp for
// row operation
for(p=0; p < N; p++) {
for(n=0; n<N; n++){
c = (PI/N)*p*(n+0.5);
if(p == 0)
row = cos(c)*1/sqrt(N);
else
row = cos(c)*sqrt(2)/sqrt(N);
out1[p][n] = row;// cos term
printf("sum calculated by thread %d= %f for q = %d.\n",omp_get_thread_num(),out1[p][n],p);
}
}
}
#pragma omp parallel //for schedule(static) num_threads(8)
{
#pragma omp for
//column operation
for(q=0; q < M; q++) {
for(m=0; m<N; m++){
d = (PI/M)*q*(m+0.5);
if(q == 0)
column = cos(d)*1/sqrt(M);
else
column = cos(d)*sqrt(2)/sqrt(M);
out2[q][m] = column;//cos term
printf("sum calculated by thread %d= %f for q = %d.\n",omp_get_thread_num(),out2[q][m],q);
}
}
}
#pragma omp parallel //for schedule(static) num_threads(8)
{
#pragma omp for
for(p=0; p < N; p++)
{
printf("\n");
for(q=0; q < M; q++)
{
for(n=0; n<N; n++)
for(m=0;m<M; m++)
out[p][q]+= in[n][m]*out1[p][n]*out2[q][m];
//printf("[%d][%d]= %f \t",p,q, out[p][q]);
}}
}
et = timerval ();
printf ("\n\nIt took TIME of %f seconds \n", et - st);
for(p=0; p < N; p++)
{
printf("\n");
for(q=0; q < M; q++)
{
printf("[%d][%d]= %f \t",p,q, out[p][q]);
}}
}
int main(int argc, char *argv[])
{
long long int N = atoi(argv[1]);
long long int M = atoi(argv[2]);
unsigned int i,k;
int nr,nl;
nr = 16;
nl = 2;
float *nsec;
nsec = (float*) malloc(sizeof(float)*nr*nl);
for(i=0;i<32;i++){
nsec[i] = 0.5;
}
float c = 6.0;
int dnr,dnl;
dnr =16;
dnl=2;
float *dsec;
dsec = (float*) malloc(sizeof(float)*dnr*dnl);
for(i=0;i<32;i++){
dsec[i] = 2.0;
}
for(i=0;i<M;i++)
for(k=0;k<N;k++)
{
X[i][k] = 10.0;
}
float *u; // intermediate data
u = (float*) calloc(dnr*2,sizeof(float));
float unew;
int j;
float y = 0;
double st, et;
st = timerval ();
int stats[240];
for(i = 0; i < omp_get_max_threads(); i++)
stats[i] = 0;
int total_stats = 0;
for(i=0;i<M;i++)
{
for(k=0;k<N;k++)
{
y = c * X[i][k];
#pragma omp parallel for
for(j=0;j<dnr;j++)
{
unew = X[i][k]- u[j*2]*dsec[j*2] - u[j*2+1]*dsec[j*2+1];
u[2*j+1]=u[2*j];
u[2*j]=unew;
y = y + u[j*2]*nsec[j*2] + u[j*2+1]*nsec[j*2+1];
stats[omp_get_thread_num()] = stats[omp_get_thread_num()] + 4;
}
Y[i][k] = y;
}
}
et = timerval ();
printf ("\n\nIt took TIME of %f seconds \n", et - st);
for(i=0;i<M;i++)
{
printf("\n");
for(k=0;k<N;k++)
{
printf("%lf ", Y[i][k]);
}
}
for(i = 0; i < omp_get_max_threads(); i++)
{
printf("Stats[%d] = %d",i, stats[i]);
total_stats = total_stats + stats[i];
}
printf("\nOperations = %d, time = %f, ops = %lf ", total_stats, et-st, total_stats/(et-st));
// release
//free(X);
//free(Y);
free(nsec);
free(dsec);
free(u);
return 0;
}
