OpenACC Parallel problem

When programming with OpenACC recently, if I have a 3 * 1 matrix and a 1 * 4 matrix, I would like to use OpenACC for multiplication, but it seems that my usage is not very correct

 cufftComplex a[3];
 cufftComplex b[4];
 a[0].x=1;
 a[1].x=2;
 a[2].x=3;
 a[0].y=1;
 a[1].y=2;
 a[2].y=4;
 
 b[0].x=3;
 b[1].x=2;
 b[2].x=3;
 b[3].x=4;
 b[0].y=4;
 b[1].y=2;
 b[2].y=3;
 b[3].y=4;

cufftComplex c[3][4];
 #pragma acc data copyin(a[0:3],b[0:4])\
 copy(c[0:3][0:4])
 {
 #pragma acc loop 
 for(int i=0;i<3;i++)
 for(int j=0;j<4;j++)
 {
 c[i][j]=a[i]*b[j];
 }
 }

Although it can run, it is obvious that his results are not correct. I think it is a problem with the introduction.The result is this

0+0j	0+0j	0+0j	0+0j	
-1.41339e-21+0j	2.10704e+15+-3.14499e+19j	2.36943e-38+2.36943e-38j	0+0j	
-1.46479e-21+0j	0+0j	8.26032e+11+2.38368e+11j	3.39615e+38+3.39615e+38j

Hi 1446561717,

I am sorry, but I don’t think this forum category is the correct one for this question.

But I see you already found one where you might get some help.
(Legacy PGI Compilers - NVIDIA Developer Forums)

Thanks!