c++ amp - vector multiplication problem - optimization with big arrays (bottom post)

hi,
i was wandering if you could help me with my C++ amp code. I want to parallel the multiplication of vector and matrix. Here is my code

#include <amp.h>

using namespace std;
using namespace concurrency;

int _tmain(int argc, _TCHAR* argv[])

int n;
    cin>> n;
double **A = new double*[n];
    for(int i = 0; i < n; ++i) {
        A[i] = new double[n];
    }

for (int i = 0; i < n ; ++ i)
        for (int j = 0; j < n; ++ j){

            A[i][j]=rand()%100;
            A[j][i]=A[i][j];

        }

for( int i=0; i < n; i++)
            x[i]=0;
        double *ri = new double[n];
        for( int i=0; i < n; i++)
            ri[i]=0;
        double *ri1 = new double[n];

array_view<const double, 2> a(n, n, A);
array_view<const double, 1> X(n, x);
array_view<double, 1> RI(n, ri);
    RI.discard_data();
parallel_for_each(
    RI.extent,
    [=] (index<1> idx) restrict(amp) {
        int y = idx[0];

    double result = 0.0f;
            for (int k=0; k<n; k++) {
                result += a(y, k) * X(k); }
            RI[idx] = result;
});
RI.synchronize();


This, however, does raise the errors C2228 and C2338, which I've not been able to resolve so far.

Thank you in advance!
Regards,
Bartosz