[ROOT] Re: [CINT] If does not work but switch does

From: Damir Buskulic (buskulic@lapp.in2p3.fr)
Date: Fri Apr 11 2003 - 10:51:10 MEST


Hi,

This kind of work may be interesting for some people, even if not 
exactly the way you use it. Is there a web page describing the project 
and where one can look at the code ?

Cheers

Damir

Le vendredi, 11 avr 2003, à 09:54 Europe/Zurich, chenu@tiscali.fr a 
écrit :

> Hello
>
>  I use ROOT to develop my algorithms.
>  I have developped a C++ wrapper class to perform Matrix operations 
> over complex elements (C).
>  This wraper relies on BLAS, LAPACK, fftw and gnuplot (or plplot).
>  The goal is to have MATLAB like syntax and environment (using CINT) 
> with performance of compiled code. It works quite fine: computation 
> (A*B), overlaoding (A=B), sub-matrices (A(1:2:10)=B),...
>  To acheive high performances I also use the ATLAS library.
>  This wrapper library is compiled and loaded in root.
>
>  It generally works fine but I sometimes find some weird things.
>
>  The last one is the following.
>  I have a loop (for(i=...)) in which a test is made (if (..))
>  within this test, cint tels me I have a bad lvalue.
>  The same code compiles fine.
>  When I change the if by switch, it works.
>  Any Idea?
>
>  I use the PRO version 3.04/02 (compiled on my target)over Mandrake 9.0
>  I noticed the same things on previous versions.
>
>
>  Thanks
>  Marc
>
>  P.S. are some of you interested by this library and MATLAB like 
> framework?
>
>
>  ************ VITE, C'EST LE MOMENT DE VOUS ABONNER A L'ADSL !!! 
> **********
>  Du 3 avril au 5 mai, série limitée ADSL 128K, cliquez vite ici : 
> http://register.tiscali.fr/adsl/
>  Offre promotionnelle soumise à conditions, réservée aux nouveaux 
> abonnés ADSL Tiscali.
>
>
>
====================================
Damir Buskulic,  Universite de Savoie/LAPP
Chemin de Bellevue, B.P. 110, F-74941 Annecy-le-Vieux Cedex, FRANCE
Tel : +33 (0)450091600
e-mail: buskulic@lapp.in2p3.fr
====================================



This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:10 MET