General Matrix inversion by MKL in C++.

Here is a example of complex matrix inversion code by using MKL.

#include
#include
#define MKL_Complex16 std::complex<double>
#include "mkl.h"
using namespace std;

void inverse(complex<double>* A, int N)
{
   int *IPIV = new int[N+1];
   int LWORK = N*N;
   complex<double> *WORK = new complex<double>[LWORK];
   int INFO;
   zgetrf_(&N,&N,A,&N,IPIV,&INFO);
   zgetri_(&N,A,&N,IPIV,WORK,&LWORK,&INFO);

   delete IPIV;
   delete WORK;
}
int main()
{
   int n=2;
   complex<double> * A = new complex<double> [n*n];
   A[0]=complex<double>(1.,2.);
   A[1]=complex<double>(3.,4.);
   A[2]=complex<double>(5.,6.);
   A[3]=complex<double>(7.,8.);
   inverse(A,n);
   for(int i=0;i<4;i++)
       cout<
   return 0;
}

Output is:

(-0.5,0.4375)
(0.25,-0.1875)
(0.375,-0.3125)
(-0.125,0.0625)

For comparing same code in numpy :

In [1]: from numpy.linalg import inv

In [2]: x=np.array([[1+2j,3+4j],[5+6j,7+8j]], dtype=complex)

In [3]: inv(x)
Out[3]:  
array([[-0.500+0.4375j,  0.250-0.1875j],
      [ 0.375-0.3125j, -0.125+0.0625j]])