[Scilab-users] Scilab6 - ATOMS linalg_dgesv

[Claus Futtrup]

Hi Samuel, et al.

Now that you mention it, yes, I tried X = A\B, but for my specific 
problem that failed in Scilab 5.5 - I just tested and it works fine in 
Scilab 6. :-) Problem solved, no need for the toolbox.

/Claus

On 20-06-2016 18:18, Samuel Gougeon wrote:
> Hello Claus,
>
> The description header of this toolbox says:
> /"These algorithms are most of the time already provided by Scilab, but 
> are available here for comparison purpose."/
>
> Indeed, in Scilab
> A*X = B
> is solved with*X = A\B*
> and that's it.
>
> Example:
> A= rand(4,4);
> Xo = rand(4,1);
> B = A*Xo;
> X = A\B;
> X-Xo
>
> --> A= rand(4,4);
> --> Xo = rand(4,1);
> --> B = A*Xo;
> --> X = A\B;
> --> X-Xo
>  ans  =
>    0.
>    0.
>    2.220D-16
>   -3.053D-16
>
> Samuel
>
> Le 20/06/2016 17:53, Claus Futtrup a écrit :
>>
>> Hi there
>>
>> I've downloaded Scilab 6 beta 2 (Windows 7, 64 bit, Intel MKL) for 
>> testing.
>>
>> Today I use a Scilab ATOMS module named linalg ... 
>> https://atoms.scilab.org/toolboxes/linalg/0.3.2/
>>
>> In my case it's good for solving a linear algebra problem (A*X = B) 
>> based on gauss elimination. It uses an input vector and a matrix - it 
>> looks like this:
>>
>> x = linalg_dgesv(mat,x);
>>
>> The big question is - will I also need DGESV from the Linalg ATOMS 
>> toolbox with Scilab6? (Or will Scilab6 do better than Scilab5 and 
>> solve this problem without a toolbox - how? - Searching Scilab6 help 
>> for DGESV didn't give any hits).
>>
>> More info:
>>
>> https://software.intel.com/sites/products/documentation/doclib/mkl_sa/11/mkl_lapack_examples/dgesv.htm
>>
>> Best regards,
>>
>> Claus
>>
>
>
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20160620/2b341c3b/attachment.htm>