<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Recent changes to bugs</title><link>https://sourceforge.net/p/ibex-lib/bugs/</link><description>Recent changes to bugs</description><atom:link href="https://sourceforge.net/p/ibex-lib/bugs/feed.rss" rel="self"/><language>en</language><lastBuildDate>Tue, 30 Aug 2011 14:54:27 -0000</lastBuildDate><atom:link href="https://sourceforge.net/p/ibex-lib/bugs/feed.rss" rel="self" type="application/rss+xml"/><item><title>empty box exception  </title><link>https://sourceforge.net/p/ibex-lib/bugs/11/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Sortie en emptybox exception   dans le gradient appelé par x_newton&lt;/p&gt;
&lt;p&gt;ligne de commande &lt;br /&gt;
minimizer1 ../benchs-optim/pollut.bch 1.e-9 1.e-8 roundrobin hc4 0.1 0.9 0.1 10 1 600&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Bertrand Neveu</dc:creator><pubDate>Tue, 30 Aug 2011 14:54:27 -0000</pubDate><guid>https://sourceforge.net4545b7b02710fbf8e3e40b8a66d4a51bc1a5bad6</guid></item><item><title>segmentation fault dans constructeur InnerBox</title><link>https://sourceforge.net/p/ibex-lib/bugs/10/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;bug rare ;  se produit en particulier dans le cas suivant  resolution de ex6_2_11.bch  avec Mohc et ACID&lt;/p&gt;
&lt;p&gt;avec la commande suivante &lt;br /&gt;
minimizer1bug ex6_2_11.bch 1.e-9 1.e-8 smearsumrel 3bcidmohc 0.1 0.9 0.1 10 1 3600&lt;/p&gt;
&lt;p&gt;j'attache les fichiers pour reproduire le bug :  ex6_2_11.bch IbexACID.cpp IbexACID.h makefile minimizer1bug.cpp&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Bertrand Neveu</dc:creator><pubDate>Mon, 22 Aug 2011 14:02:04 -0000</pubDate><guid>https://sourceforge.netea13f31bf3341fd7e6884735e396c161b42fa378</guid></item><item><title>impossibilité traiter optimisation sans contraintes</title><link>https://sourceforge.net/p/ibex-lib/bugs/9/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Quand il n' y a pas de contraintes dans l'optimisation, &lt;br /&gt;
sortie de Ibex avec message d'erreur &lt;br /&gt;
Empty composite contractor&lt;/p&gt;
&lt;p&gt;contournement provisoire : ajout d'une conrtrainte bidon   de type x&amp;gt;= xmin&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Mon, 09 May 2011 08:12:38 -0000</pubDate><guid>https://sourceforge.neta6908ca20549aa05f84457901edd04ddde5a1015</guid></item><item><title>Reference to static cell default buffer</title><link>https://sourceforge.net/p/ibex-lib/bugs/8/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Several pavers share the same reference to a cell buffer when the latter is created by default&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Sat, 07 May 2011 13:49:58 -0000</pubDate><guid>https://sourceforge.net4840d43290415621e10c10b8967f0a150582c26a</guid></item><item><title>Empty INTERVAL is [0,0] instead of [-oo,-oo]</title><link>https://sourceforge.net/p/ibex-lib/bugs/7/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Voir programme ci-joint.&lt;/p&gt;
&lt;h1 id="resultat"&gt;Résultat&lt;/h1&gt;
&lt;p&gt;Non-empty interval :     [4,5]    0&lt;br /&gt;
Interval has been set_emptied :     [0,0]    0&lt;br /&gt;
Manually empty interval :     [-inf,-inf]    1&lt;br /&gt;
Interval has been set_emptied :     [0,0]    0&lt;br /&gt;
[0,0] &lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Tue, 12 Apr 2011 15:25:26 -0000</pubDate><guid>https://sourceforge.net324bc4e26d8f204e386063700f1990a668f64645</guid></item><item><title>Bias failure with constant expression involving Power</title><link>https://sourceforge.net/p/ibex-lib/bugs/6/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;The following Quimper program will provoke a run-time exception:&lt;br /&gt;
"BIAS: Power: Base is negative"&lt;/p&gt;
&lt;p&gt;Constants&lt;br /&gt;
a in [0,4];&lt;br /&gt;
b = 4;&lt;/p&gt;
&lt;p&gt;Variables&lt;br /&gt;
l in [0,5];&lt;/p&gt;
&lt;p&gt;Constraints&lt;br /&gt;
(a-b)^2 = 0;&lt;br /&gt;
end&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Anonymous</dc:creator><pubDate>Fri, 08 Apr 2011 16:12:44 -0000</pubDate><guid>https://sourceforge.net8b61ac18ed37328672ae80dcc959bc20f3a3a95a</guid></item><item><title>elimination de solution optimal par "contract with loup"</title><link>https://sourceforge.net/p/ibex-lib/bugs/5/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;Le procedure suivante dans la class  IbexOptimizer.cpp peut eliminer l'optimum: &lt;br /&gt;
/*======================== contract with loup ========================*/&lt;br /&gt;
INTERVAL&amp;amp; y=space.box(y_num+1); // the epr y_num is the objective&lt;br /&gt;
REAL ymax=loup;&lt;/p&gt;
&lt;p&gt;if ((fabs (loup) &amp;lt; 1) || (goal_ceil &amp;gt; 1))  // goal_ceil is a relative precision  and should be less than 1&lt;br /&gt;
ymax=loup-goal_ceil;&lt;br /&gt;
else&lt;br /&gt;
ymax=loup*(1-goal_ceil);&lt;br /&gt;
if (Inf (y) &amp;gt; ymax) {&lt;br /&gt;
delete c; // if this procedure was in a contractor ---&amp;gt; throw EmptyBoxException();&lt;br /&gt;
return;&lt;br /&gt;
}&lt;br /&gt;
if (Sup (y) &amp;gt; ymax) y = INTERVAL(Inf(y),ymax); &lt;br /&gt;
/*====================================================================*/&lt;/p&gt;
&lt;p&gt;Alors, si tout les boite son elimine et on arrive pas a la precisión, ça veut dire que l'optimum est trouve dans:&lt;br /&gt;
[loup-goal_ceil, loup]  ou  [loup*(1-goal_ceil), loup]&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Ignacio Araya</dc:creator><pubDate>Thu, 20 Jan 2011 14:12:29 -0000</pubDate><guid>https://sourceforge.net1c5e1a771dda72391751d78c74ff921783ece483</guid></item><item><title>Sequence and solver</title><link>https://sourceforge.net/p/ibex-lib/bugs/4/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;When a Paver is used in solver mode and one of the contractor is a Sequence, the boxes corresponding to the Sequence are not stored...&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Marie Pelleau</dc:creator><pubDate>Wed, 19 Jan 2011 19:07:31 -0000</pubDate><guid>https://sourceforge.net772b6de5b51d33fefbbf549a3ee9703cbeee2a35</guid></item><item><title>M_INF</title><link>https://sourceforge.net/p/ibex-lib/bugs/3/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;IbexEvaluator.cpp lignes 170-173:&lt;/p&gt;
&lt;p&gt;case M_INF    : M[ic--] = Inf(M[info[i--]]);                 break;&lt;br /&gt;
case M_MID    : M[ic--] = Mid(M[info[i--]]);                 break;&lt;br /&gt;
case M_SUP    : M[ic--] = Sup(M[info[i--]]);                 break;&lt;/p&gt;
&lt;p&gt;ic ou mc ??????&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Gilles Chabert</dc:creator><pubDate>Wed, 19 Jan 2011 16:48:21 -0000</pubDate><guid>https://sourceforge.net1ee88cc17b870c48061528a84dadc8bcb8e83e6a</guid></item><item><title>is_inside on a loup candidate</title><link>https://sourceforge.net/p/ibex-lib/bugs/2/</link><description>&lt;div class="markdown_content"&gt;&lt;p&gt;In candidate_check, the is_inside test is called on the entire box and not on the loup.&lt;/p&gt;&lt;/div&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Bertrand Neveu</dc:creator><pubDate>Tue, 18 Jan 2011 14:52:49 -0000</pubDate><guid>https://sourceforge.net28bbb1eda6d7613572cf9f630029c9f17e1d2730</guid></item></channel></rss>