Path: news.mathworks.com!not-for-mail
From: "Petros" <painter@geo.auth.gr>
Newsgroups: comp.soft-sys.matlab
Subject: bug with graphallshortestpaths?
Date: Thu, 19 Feb 2009 09:58:01 +0000 (UTC)
Organization: Aristotle University of Thessaloniki, Greece
Lines: 34
Message-ID: <gnjaf9$7ce$1@fred.mathworks.com>
Reply-To: "Petros" <painter@geo.auth.gr>
NNTP-Posting-Host: webapp-02-blr.mathworks.com
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 8bit
X-Trace: fred.mathworks.com 1235037481 7566 172.30.248.37 (19 Feb 2009 09:58:01 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Thu, 19 Feb 2009 09:58:01 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 465221
Xref: news.mathworks.com comp.soft-sys.matlab:519379


I think there is a bug with graphallshortestpaths function. Here is an example with the problem.
clear
load A.mat
% A.mat incudes only one sparce double 357x357 matrix called "A"   


[DIST1] = graphallshortestpaths(A,'directed',false);
% DIST1 includes unreached points (indicated with "Inf" )
%this is the wrong behavior 


A2=A./1000;
% I produce A2 with lower weith values.

[DIST2] = graphallshortestpaths(A2,'directed',false);
% DIST2 does not include unreached points (indicated with "Inf"), just 
%because the A matrix has lower weight values 
%this is the normal behavior 


Any help is welcome 

below I post a link someone can download the example above as good as the A.mat in order to try it.

http://users.auth.gr/~painter/bug/bug_example.zip




MATLAB/PLATFORM INFO
Release: R2008b
Product: Bioinformatics Toolbox
Product Version: 3.2
Operating System: Windows Vista