Propagation via Guassian elimination?

[Torsten Anders]

Dear Eric,

unfortunately, I can't give you any specific answer on algorithms such  
as Guassian elimination. Nevertheless, here are a few pointers you  
might be interested in.


You may be interested in the XRI (eXtended Real Interval) extension to  
Mozart.

On 16.02.2005, at 15:38, Gustavo Gutierrez Sabogal wrote:
> Hi Carlos,
>
>
>> Hello. I'm new in mozart.
>> I'm trying to solve a problem which implies real numbers. When I  
>> import
>> the RI module as the documentation says, I get an error of non  
>> existence
>> of the module. I have searched the module into the mozart directory
>> installation without success. Can it be downloaded?. From where?. Any
>> advice about using that module?.
>> Thanks.
>
> The real interval module from mozart contribs is not working as far as
> i know. It works fine with mozart 1.2.5 and linux. Portability of  
> programs
> using this module is not possible. Here in Colombia we have developed a
> new more portable real interval constraint system for mozart called XRI
> which stands for eXtended Real Interval. This module has been ported to
> other platforms different from linux: MacOs X (powerpc) and Windows. If
> you wish to try it, go to http://home.gna.org/xrilpoz/
>
> Please, don't hesitate to ask for help or to share your experience for
> feedback.
>
> All the best,
> Gustavo Gutierrez


The book by Christian:

@Book{Schulte:Book:2002,
   Author    = "Christian Schulte",
   Title     = "Programming Constraint Services",
   Publisher = "Springer-Verlag",
   Year      = 2002,
   Address   = "Berlin, Germany",
   Series    = "Lecture Notes in Artificial Intelligence",
   Volume    = 2302,
   OPTURL       =  
"http://link.springer.de/link/service/series/0558/tocs/t2302.htm",
   Abstract  = "
   Constraint programming is an approach to modeling and solving  
combinatorial
   problems that has proven succesful in many applications. Building on
   techniques developed in AI, logic programming and operations research,
   constraint programming is based on an abstraction that decomposes the
   problem solver into a reusable constraint engine and a declarative
   program modeling the problem.

   This book is concerned with the architecture and implementation of  
constraint
   engines. The author's main contribution is that constraint services,
   such as search and combinators, are made programmable; this is  
achieved
   by devising computation spaces as simple abstraction for programming
   constraint services at a high level. State-of-the-art and novel search
   strategies, such as visual interactive search and parallel search
   are covered.

   The book is indispensible reading for anyone seriously interested in
   constraint technology."
}


Furthermore, there is the C++ lib Gecode which promises to make the  
definition of new domains more easy.

GECODE:  www.gecode.org


On 27.10.2005, at 12:44, Christian Schulte wrote:
>> Are there perhaps any plans to integrate Gecode in Mozart as well?
>
> I don't know whether there are any plans. Maybe the only thing that I  
> can
> (have to) say is that it is easy enough to do. We are currently  
> focusing on
> having a release out in November and then have a good Java interface  
> out
> soon after that.


Best,
Torsten


On 10.01.2006, at 20:17, mozart at randomhacks.net wrote:
> I have a really tangential question, and I hope it's not too far off
> topic. :-/
>
> I've recently run into some constraint problems.  For example, I need  
> to
> solve systems of linear equations over real numbers, and perhaps
> generalize this in various directions.  I understand how to propagate
> these constraints using well-known algorithms.  But I'm looking for a
> language which makes this style of programming natural.
>
> Everything I've read suggests that CLP is a promising technique:  
> There's
> a whole family of research languages from CLP(R) onward which tackle
> constraint problems in elegant ways.
>
> But here's where I get stuck: The most popular constraint  
> languages--Oz,
> Alice, GNU Prolog, etc.--focus heavily on finite domains.  In some
> cases, a real-interval library is available.  What I want, though, is
> strong propagators: Guassian elimination, the simplex method, and other
> tools of that sort.  And I don't understand how to make the jump from  
> finite
> domains to these tools.
>
> I've read CTM chapter 12 (what a cool book!), and I understand how a
> language like CLP(R) fits together.  I even see some analogies between
> CLP(R)'s deferred constraints and a dataflow model.  But I don't know
> how to reimplement these tools in Mozart.
>
> So, my questions:
>
>   1) Can propagation algorithms such as Guassian elimination be
>      represented elegantly using Mozart?
>   2) If so, what papers, books or code do I need to read to get  
> unstuck?
>
> Thank you for any advice you can provide!
>
> Sincerely,
> Eric
>
> _______________________________________________________________________ 
> __________
> mozart-users mailing list                                
> mozart-users at ps.uni-sb.de
> http://www.mozart-oz.org/mailman/listinfo/mozart-users
>
>
--
Torsten Anders
Sonic Arts Research Centre
Queen's University Belfast (UK)
www.torsten-anders.de