Hello Karl,
My mistake is from a misunderstanding of the documentation. I read that:
Conjugate Gradient (CG) symmetric positive definite
Stabilized Bi-CG (BiCGStab) non-symmetric
Generalized Minimum Residual (GMRES) general
Like "General A" NxM, my is for any kind of square-matrices.
I have already try least square via QR, SVD... But the result are "too
sharp" (the result are visual position on 3D space), I would like a more
smooth method, that why I try to use a method based on gradient or
iterative method.
Did you know any method with a "relaxation-step" inside or based on
gradient I could use with rectangular matrices?
Thanks
Soufiane KHIAT
Software Engineer
2013/12/4 Karl Rupp <r...@iue.tuwien.ac.at>
> Hi Soufiane,
>
> from your description it seems to me that you want to solve a
> least-squares problem, while the GMRES implementation in ViennaCL is for
> square systems. I suggest you use a QR-factorization as outlined in
> examples/tutorial/least-squares.cpp.
>
> If you have any good pointers for GMRES applied to least-squares problems,
> feel free to share so that we can consider extending our GMRES
> implementation suitably.
>
> Best regards,
> Karli
>
>
>
>
> On 12/03/2013 10:54 PM, Soufiane Khiat wrote:
>
>> Hello,
>>
>> I am trying to solve Ax=b
>> With a Dense and Rectangular (14000 Rows and 140 Cols).
>> I use as presented on the sample/documentation:
>>
>> viennacl::vector< Scalar > _b( m_RowsCount );
>> viennacl::vector< Scalar > _x( m_ColumnsCount );
>>
>> viennacl::fast_copy( b.Datas, b.Datas + m_RowsCount, _b.begin() );
>> _x = viennacl::linalg::solve( _A, _b, viennacl::linalg::gmres_tag() );
>> viennacl::fast_copy( _x.begin(), _x.end(), x.Datas );
>>
>> x, b is a container of contigüe datas (float*).
>> But during the solve I have an assert:
>> On Matrix_Operation.hpp:
>> assert( (viennacl::traits::size2(mat) == viennacl::traits::size(vec)) &&
>> bool("Size check failed at v1 = prod(A, v2): size2(A) != size(v2)"));
>>
>> On:
>> template <typename NumericT, typename F>
>> void prod_impl(const matrix_base<NumericT, F> & mat,
>> const vector_base<NumericT> & vec,
>> vector_base<NumericT> & result)
>>
>> But what is strange for me is in the call stack I pass through another
>> assert without anyu trouble:
>> template <typename NumericT, typename F>
>> vector<NumericT>
>> operator-=(vector_base<NumericT> & v1,
>> const viennacl::vector_expression< const matrix_base<NumericT, F>, const
>> vector_base<NumericT>, viennacl::op_prod> & proxy)
>> {
>> assert(viennacl::traits::size1(proxy.lhs()) == v1.size() &&
>> bool("Size check failed for v1 -= A * v2: size1(A) != size(v1)"));
>>
>> vector<NumericT> result(viennacl::traits::size1(proxy.lhs()));
>> viennacl::linalg::prod_impl(proxy.lhs(), proxy.rhs(), result);
>> v1 -= result;
>> return v1;
>> }
>>
>> ...
>>
>> I initialize _A like that:
>> viennacl::matrix< Scalar > _A( m_RowsCount, m_ColumnsCount );
>>
>> for ( u32 i = 0; i < m_ColumnsCount; ++i )
>> {
>> for ( u32 j = 0; j < m_RowsCount; ++j )
>> {
>> _A( j, i ) = A[ i ][ j ];
>> }
>> }
>>
>> I am a new user with ViennaCL. If you see some mistake can told me where.
>>
>> In my opinion my system is correctly balanced:
>> A : 14000x140
>> b : 14000x1
>> x : 140x1
>>
>> Cheer,
>>
>> Soufiane
>> Software Engineer
>>
>>
>> ------------------------------------------------------------
>> ------------------
>> Sponsored by Intel(R) XDK
>> Develop, test and display web and hybrid apps with a single code base.
>> Download it for free now!
>> http://pubads.g.doubleclick.net/gampad/clk?id=111408631&;
>> iu=/4140/ostg.clktrk
>>
>>
>>
>> _______________________________________________
>> ViennaCL-devel mailing list
>> ViennaCL-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/viennacl-devel
>>
>>
>
------------------------------------------------------------------------------
Sponsored by Intel(R) XDK
Develop, test and display web and hybrid apps with a single code base.
Download it for free now!
http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk
_______________________________________________
ViennaCL-devel mailing list
ViennaCL-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/viennacl-devel
