- Boost-users - lists.boost.org

[BGL] premature termination in dijkstra using visitor
by Ralf Goertz 07 Oct '09

07 Oct '09

Hi, I have some very large undirected graphs (half a million vertices). Edges have weight 0 or 1. For each vertex I need to find all vertices * that are in sum 2 edges away * whose distance is 0 Both tasks can be done with breadth-first searches but I don't know in advance how many edges in depth I have to traverse. That's why I thought of Dijkstra's algorithm. But in my case this is overkill as I am not interested in the exact distance of a vertex from a given vertex but only in if it is not farther than 0 or 2 away. Is it possible to use a visitor to alter the colormap to mark vertices as final that are too far even thought they are not final yet? Ralf

5 12

[proto] Question on grammar and operator overloading
by Hal Finkel 06 Oct '09

06 Oct '09

Hello, I'm writing a complex number class using boost::proto by extending proto::terminal< boost::array<_, 2> >. I've attached a copy of my current code, which compiles, runs, and (at least with both g++ and icc on my machine) demonstrates a significant speed improvement compared to std::complex. However, there is (at least) one thing which don't work as it should: In the grammar I have to specify lazy_complex< double > instead of lazy_complex< _ >, otherwise the compiler cannot find proto's operator overloads, why? Note: This code will not compile under 1.37, but will compile under 1.40, because of some issue in 1.37 which manifests itself as a problem with the overloading of the '<<' operator. I've not tested the releases in between 1.37 and 1.40. General suggestions are also welcome. I'd like to get this cleaned up for general redistribution. Thank you in advance, Hal

3 7

Newbie: Boost Graph example request
by Brian Troutwine 06 Oct '09

06 Oct '09

Hello all. I'm new to using the Boost Graph library and am bewildered by its (seemingly essential) complexity. A solution to the simplistic problem I pose below would help provide a workable context around which to wrap my brain and would be most appreciated. -- Let G be the undirected graph given by the following adjacency list: 0 -> 2, 3 1 -> 2, 4 2 -> 3, 4 3 -> 1 4 -> 3 5 -> 0 6 -> 7 7 -> 8, 9 9 -> 10 Using, say, a breadth first search, for any two random nodes in the graph print the shortest path between them. For example, the shortest path between 5 and 3 is: 5 -> 0 -> 3 My apologies if a similar question has been asked before. A search through the mailing list archive did not return promising results. Also, I have, incidentally, ordered The Boost Graph Library text and am very much looking forward to its arrival. Thank you. Brian Troutwine

2 2

Shared memory volatility
by manish4gupta 06 Oct '09

06 Oct '09

Is shared memory created by Interprocess communication gets deleted when system is switched off? -- View this message in context: http://www.nabble.com/Shared-memory-volatility-tp25764705p25764705.html Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

[iterator_transform]'s reference
by er 06 Oct '09

06 Oct '09

Hi All, It appears that transform_iterator's reference is F::result_type, not result_of<F(Ref)> where Ref = iterator_reference<It>::type. Did I misinterpret the documentation?

1 0

dynamic objects (was Re: PBGL / Dynamic Property Maps complexity)
by Mathieu Malaterre 05 Oct '09

05 Oct '09

Nick, [reviving an old thread] On Sat, Sep 12, 2009 at 9:39 PM, Nick Edmonds <ngedmond(a)cs.indiana.edu> wrote: <...> > If your graphs aren't static... PBGL doesn't do well with dynamic graphs at > the moment, they'll *sort of* work... *sometimes* :) What do you mean by that ? What does not work ? I am currently trying to understand whether or not I can use dynamic objects stored in bundled properties, or will there be an issue in PBGL (when the object will be copied over to another machine) ? Sorry for evasive question but I could not find any example of BGL or PBGL that did not use POD struct for storing the properties. Thanks, -- Mathieu

2 1

compatibility of boost 1.39
by manish4gupta 05 Oct '09

05 Oct '09

Is boost library 1.39 compatible on ia64 machine(itanium) with compiler gcc 3.4.6 on a Redhat Enterprise Linux 4 Update 5? -- View this message in context: http://www.nabble.com/compatibility-of-boost-1.39-tp25696866p25696866.html Sent from the Boost - Users mailing list archive at Nabble.com.

2 12

Re: [Boost-users] Boost unit_test.cpp problem when try to compile simple program
by Robert Matusewicz 04 Oct '09

04 Oct '09

Gennadiy Rozental writes: >> You need different include than. Can't have it both ways. Either you include >> library implementation or you link with it. >> >> Use boost/test/unit_test.hpp " Just to close the topic. We can use boost/test/included/unit_test.hpp with Boost DLLs. We have to define BOOST_TEST_SOURCE. RM -- My public key can be found at https://keyserver2.pgp.com

2 1

modify_key() for a composite_key of a multi_index_container that contains a shared_ptr
by Zsolt Vajna 04 Oct '09

04 Oct '09

I'd like to modify_key() for a composite_key of a multi_index_container that contains a shared_ptr. I found a thread in this mailing list dealing with a similar kind of situation, not totally the same though: "[Boost-users] [multi_index] modify problem with composite_key" My example: This compiles fine: struct entry { entry(int a, int b):m_x(a), m_y(b){} int m_x; int m_y; }; typedef multi_index_container< entry, indexed_by< ordered_unique< composite_key< entry, member<entry,int,&entry::m_x>, member<entry,int,&entry::m_y>>>>> C; void updateX(C::iterator iter,int x, C& m_c) { m_c.get<0>().modify( iter, (&ll::_1)->*&entry::m_x=x); } My first problem is if I change modify to modify_key I get the following error: operator_lambda_func_base.hpp(237) : error C3892: 'boost::lambda::detail::select' : you cannot assign to a variable that is const My second problem is that instead of typedef multi_index_container< entry, indexed_by< ordered_unique< composite_key< entry, member<entry,int,&entry::m_x>, member<entry,int,&entry::m_y>>>>> C; i want to have a typedef multi_index_container< shared_ptr<entry>, indexed_by< ordered_unique< composite_key< entry, member<entry,int,&entry::m_x>, member<entry,int,&entry::m_y>>>>> C; Note the shared_ptr added. In this later case I get another error: boost\lambda\detail\member_ptr.hpp(512) : error C2440: 'newline' : cannot convert from 'const rt0 ' to 'entry *const ' I assume I need to change something around the _1 bit, but have no idea what. Thanks

2 1

cmake can't find Boost 1.40.0
by Timothy Chan 04 Oct '09

04 Oct '09

Hi all, I have a newbie question here. I am installing a software called gazebo which uses cmake. The software requres boost 1.35+. I have 1.32 on the system, and 1.40 installed locally. Cmake seems to only be able to find 1.32 and is not able to find my local installation of 1.40. I have boost installed in: /home/timchan/boost_1_40_0 In my .bashrc file I put: export LD_LIBRARY_PATH=/home/timchan/boost_1_40_0/libs:$LD_LIBRARY_PATH export CPATH=/home/timchan/boost_1_40_0/boost:$CPATH When I run cmake, I get the following error: CMake Error at /usr/share/cmake/Modules/FindBoost.cmake:868 (message): Unable to find the requested Boost libraries. Boost version: 1.32.0 Boost include path: /usr/include Detected version of Boost is too old. Requested version was 1.35 (or newer). Any help would be greatly appreciated!!

3 2