Boost C++ Libraries: Ticket #480: dijkstra_shortest_path doesn't scan the whole component https://svn.boost.org/trac10/ticket/480 <pre class="wiki">Disjkstra Shortest path stops in the middle of the graph. I think this is because of a bug in relaxed_heap. This bug is present in boost 1.33 and boost cvs HEAD. It is however not present in debian boost-graph-dev version 1.32.0-6. I attach a source code which demonstrates the bug. It produces a dot file with the nodes labeled with distances from the source and the arcs labeled with their weight. The bug is present if some nodes have a 2147483647 label. </pre> en-us Boost C++ Libraries /htdocs/site/boost.png https://svn.boost.org/trac10/ticket/480 Trac 1.4.3 Douglas Gregor Thu, 08 Sep 2005 19:16:26 GMT <link>https://svn.boost.org/trac10/ticket/480#comment:1 </link> <guid isPermaLink="false">https://svn.boost.org/trac10/ticket/480#comment:1</guid> <description> <pre class="wiki">Logged In: YES user_id=249098 I was able to reproduce this problem on x86 Linux but not on PowerPC MacOS X. This is extremely important for 1.33.1. </pre> </description> <category>Ticket</category> </item> <item> <dc:creator>Douglas Gregor</dc:creator> <pubDate>Fri, 09 Sep 2005 22:20:41 GMT</pubDate> <title>status changed https://svn.boost.org/trac10/ticket/480#comment:2 https://svn.boost.org/trac10/ticket/480#comment:2 <ul> <li><strong>status</strong> <span class="trac-field-old">assigned</span> → <span class="trac-field-new">closed</span> </li> </ul> <pre class="wiki">Logged In: YES user_id=249098 The problem was due to rounding issues in the computation of the base-2 logarithm of the size_type. I've replaced the calls to std::log(double) with a hand-rolled integer log2. The test passes for me on x86 Linux and PPC Mac OS X. The fixes are available on RC_1_33_0 and CVS HEAD. </pre> Ticket Douglas Gregor Fri, 09 Sep 2005 22:22:02 GMT <link>https://svn.boost.org/trac10/ticket/480#comment:3 </link> <guid isPermaLink="false">https://svn.boost.org/trac10/ticket/480#comment:3</guid> <description> <pre class="wiki">Logged In: YES user_id=249098 The problem was due to rounding issues in the computation of the base-2 logarithm of the size_type. I've replaced the calls to std::log(double) with a hand-rolled integer log2. The test passes for me on x86 Linux and PPC Mac OS X. The fixes are available on RC_1_33_0 and CVS HEAD. </pre> </description> <category>Ticket</category> </item> </channel> </rss>