LU decomposition of complex dense matrices

Hi,

I'm interested in inverting small dense complex
matrices.  The mtl function, lu_factorize, works fine
for a real matrix.  I'd like to make it work for a
complex matrix.  Naively putting in a complex matrix
does not give the correct LU decomposition.  Delving a
little bit deeper, I found that the mtl max_index
function does not work for complex vectors.  For
example, the largest (magnitude) component in the
vector (1 + 0*i, 0 - 2*i) is the second (last) one. 
But max_index gives the first one.  The following code
demonstrates this:

// -*- c++ -*-
//
// $COPYRIGHT$
//
//===========================================================================
#include "C:/Include/mtl/workaround.h"
#include "C:/Include/mtl/utils.h" 
#include "C:/Include/mtl/linalg_vec.h"
#include "C:/Include/mtl/mtl.h"
using namespace mtl;
using namespace std;
//begin
typedef external_vec<complex <double> > Vec;
//end

/*
  Sample Output 

  [(1,0),(0,-2),]
	Largest element in the vector z is at location 1

 */

// How can we make it give the desired answer:
// "Largest element in the vector z is at location 2"



int main()
{
	const int N = 2;
	complex<double> I(0,1);
	complex<double> test[] = {1, -2.0*I};
	Vec z(test,N);
	int imax = max_index(z);
	print_vector(z);
	cout << "Largest element in the vector z is at
location "
         << imax << endl;
    return 0;
}  

How can I modify the code so that it works for complex
numbers?

Dave



=====
[email protected]
David Lidsky


__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com
_______________________________________________
This list is archived at http://www.osl.iu.edu/MailArchives/mtl-devel/