Re: [ublas] compile error on gesv.hpp on 64Bit system

Subject: Re: [ublas] compile error on gesv.hpp on 64Bit system
From: Kraus Philipp (philipp.kraus_at_[hidden])
Date: 2010-07-25 12:15:05

Hi Rutger,

Am 25.07.2010 um 16:15 schrieb Rutger ter Borg:

> Kraus Philipp wrote:
>
> [snip]
>
>> Do I have forgot anything? I hope anyone can help me and send me a  
>> tip
>> or an idea.
>> Thanks a lot
>>
>
> Do you have an example of your code?  Are you using fortran_int_t?

This is a code snip that I call:

  ublas::matrix<T, ublas::column_major> l_matrix(p_matrix);

  ublas::matrix<T, ublas::column_major> l_eigvec(l_matrix.size1(),  
l_matrix.size2());
  ublas::vector<T> l_eigval(l_matrix.size1());

  ublas::vector<T> l_tmp1(l_eigval.size());
  ublas::matrix<T, ublas::column_major>  
l_tmp2(l_matrix.size1(),l_matrix.size2());
  ublas::matrix<T, ublas::column_major>  
l_tmp3(l_matrix.size1(),l_matrix.size2());

  linalg::geev( 'N', 'V', l_matrix, l_eigval,  l_tmp1,l_tmp2,   
l_eigvec, linalg::optimal_workspace() );

with this namespaces
namespace ublas     = boost::numeric::ublas;
namespace blas      = boost::numeric::bindings::blas;
namespace linalg    = boost::numeric::bindings::lapack;

Is this enought information otherwise I would create tomorrow a full  
program.
The code excerpt ist part of a template class.

Thx

Phil