- Boost-users - lists.preview.boost.org

[MSM] Use events twice
by Sebastian Redl 08 Apr '11

08 Apr '11

Hi, I'm using MSM to implement a client for a simple network application. Basically, the server sends messages and the client changes state and responds to them as appropriate. One of these messages is THROW. THROW gets two arguments: an identifier (0-3) and a value. My client is supposed to sum up the values for consecutive THROW events with the same identifier, and on a change of identifier, store the sum somewhere and start over for the new identifier. So I have three states (id0, id1, id2 - there's some special handling, which is why they're separate states) and an event Throw(id, value). The transitions for idN are: Throw(my id, value) -> no switch, state.sum += value Throw(other id, value) -> switch to state of other id, reprocess event, storing the sum is handled by the exit action In other words, I want the next state to immediately handle the event as well, because there's a value there I need to store. The question is, how do I do this? Here's a sequence of events and actions that could occur: E1: [init] Throw(0, 4) -> [id0: ?] Entry [id0: ?] -> [id0: 0] [id0: 0] reprocess E1 -> [id0: 4] E2: [id0: 4] Throw(0, 8) -> [id0: 12] E3: [id0: 12] Throw(1, 7) -> [id1: ?] Exit [id0: 12] -> save 12 Entry [id1: ?] -> [id1: 0] [id1: 0] reprocess E3 -> [id1: 7] I hope I was clear enough. Sebastian

1 0

[BGL] Writing an edge_iterator for add_edges in CSR graph
by Johan Oudinet 08 Apr '11

08 Apr '11

Hi, I'm trying to write an edge_iterator in order to load a CSR graph from a GraphViz by using the add_edges method. So far, I successfully write a edge_iterator so you can iterate on it and dereference it to get either a pair of vertex_descriptor or a edge label. However, if it try to use this iterator as arguments for the method add_edges of CSR graph, I got cryptic errors and I don't know how to solve them :-/ You can find the code that produce these errors here: http://tinyurl.com/csr-add-edges-error I will appreciate any help to understand what is going wrong or to write an edge_iterator that actually works with add_edges. Thanks. Regards, P.S.: I use Boost graph library 1.42 installed from Ubuntu 10.10 repositories. -- Johan () ascii ribbon campaign - against html e-mail /\ www.asciiribbon.org - against proprietary attachments

2 3

[MSM] problem compiling a functor-guard using a flag inside.
by Steckmann, Sven 08 Apr '11

08 Apr '11

Dear list, while developing I want to use functors for a guard. The state machine compiles well until I added a function call to the FSM which has a template parameter. Then I always got the following error message while compiling it with gcc 4.4.1. SubtractionControllerNew.cpp: In member function 'bool<unnamed>::player_::isCDLoaded::operator()(const EVT&, FSM&, SourceState&, TargetState&)': SubtractionControllerNew.cpp:175: error: expected primary-expression before '>' token SubtractionControllerNew.cpp:175: error: expected primary-expression before ')' token Does anybody know why this happens and how to circumvent it? Kind regards, Sven ---- Example from flags. I have modified the CDLoaded example to present the bug with the following lines: struct isCDLoaded { template <class EVT,class FSM,class SourceState,class TargetState> bool operator()(EVT const& ,FSM& fsm,SourceState& ,TargetState& ) { //return fsm.is_contained(); // this works well return fsm.is_flag_active<CDLoaded>(); // Line 175 } }; Row < Stopped , open_close , Open , none , isCDLoaded >, // Copyright 2010 Christophe Henry // henry UNDERSCORE christophe AT hotmail DOT com // This is an extended version of the state machine available in the boost::mpl library // Distributed under the same license as the original. // Copyright for the original version: // Copyright 2005 David Abrahams and Aleksey Gurtovoy. Distributed // under the Boost Software License, Version 1.0. (See accompanying // file LICENSE_1_0.txt or copy at // http://www.boost.org/LICENSE_1_0.txt) #include <iostream> // back-end #include <boost/msm/back/state_machine.hpp> //front-end #include <boost/msm/front/state_machine_def.hpp> #include <boost/msm/front/functor_row.hpp> #include <boost/msm/front/euml/common.hpp> namespace msm = boost::msm; namespace mpl = boost::mpl; using namespace msm::front; namespace // Concrete FSM implementation { // events struct play {}; struct end_pause {}; struct stop {}; struct pause {}; struct open_close {}; struct NextSong {}; struct PreviousSong {}; // Flags. Allow information about a property of the current state struct PlayingPaused{}; struct CDLoaded {}; struct FirstSongPlaying {}; // A "complicated" event type that carries some data. struct cd_detected { cd_detected(std::string name) : name(name) {} std::string name; }; // front-end: define the FSM structure struct player_ : public msm::front::state_machine_def<player_> { // The list of FSM states struct Empty : public msm::front::state<> { // every (optional) entry/exit methods get the event passed. template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "entering: Empty" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "leaving: Empty" << std::endl;} }; struct Open : public msm::front::state<> { typedef mpl::vector1<CDLoaded> flag_list; template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "entering: Open" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "leaving: Open" << std::endl;} }; struct Stopped : public msm::front::state<> { // when stopped, the CD is loaded typedef mpl::vector1<CDLoaded> flag_list; template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "entering: Stopped" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "leaving: Stopped" << std::endl;} }; // the player state machine contains a state which is himself a state machine // as you see, no need to declare it anywhere so Playing can be developed separately // by another team in another module. For simplicity I just declare it inside player struct Playing_ : public msm::front::state_machine_def<Playing_> { // when playing, the CD is loaded and we are in either pause or playing (duh) typedef mpl::vector2<PlayingPaused,CDLoaded> flag_list; template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "entering: Playing" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "leaving: Playing" << std::endl;} // The list of FSM states struct Song1 : public msm::front::state<> { typedef mpl::vector1<FirstSongPlaying> flag_list; template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "starting: First song" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "finishing: First Song" << std::endl;} }; struct Song2 : public msm::front::state<> { template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "starting: Second song" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "finishing: Second Song" << std::endl;} }; struct Song3 : public msm::front::state<> { template <class Event,class FSM> void on_entry(Event const&,FSM& ) {std::cout << "starting: Third song" << std::endl;} template <class Event,class FSM> void on_exit(Event const&,FSM& ) {std::cout << "finishing: Third Song" << std::endl;} }; // the initial state. Must be defined typedef Song1 initial_state; // transition actions void start_next_song(NextSong const&) { std::cout << "Playing::start_next_song\n"; } void start_prev_song(PreviousSong const&) { std::cout << "Playing::start_prev_song\n"; } // guard conditions typedef Playing_ pl; // makes transition table cleaner // Transition table for Playing struct transition_table : mpl::vector4< // Start Event Next Action Guard // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Song1 , NextSong , Song2 , &pl::start_next_song >, a_row < Song2 , PreviousSong, Song1 , &pl::start_prev_song >, a_row < Song2 , NextSong , Song3 , &pl::start_next_song >, a_row < Song3 , PreviousSong, Song2 , &pl::start_prev_song > // +---------+-------------+---------+---------------------+--------------- -------+ > {}; // Replaces the default no-transition response. template <class FSM,class Event> void no_transition(Event const& e, FSM&,int state) { std::cout << "no transition from state " << state << " on event " << typeid(e).name() << std::endl; } }; // back-end typedef msm::back::state_machine<Playing_> Playing; // state not defining any entry or exit struct Paused : public msm::front::state<> { typedef mpl::vector2<PlayingPaused,CDLoaded> flag_list; }; // the initial state of the player SM. Must be defined typedef Empty initial_state; // transition actions void start_playback(play const&) { std::cout << "player::start_playback\n"; } void open_drawer(open_close const&) { std::cout << "player::open_drawer\n"; } void close_drawer(open_close const&) { std::cout << "player::close_drawer\n"; } void store_cd_info(cd_detected const& ) {std::cout << "player::store_cd_info\n";} void stop_playback(stop const&) { std::cout << "player::stop_playback\n"; } void pause_playback(pause const&) { std::cout << "player::pause_playback\n"; } void resume_playback(end_pause const&) { std::cout << "player::resume_playback\n"; } void stop_and_open(open_close const&) { std::cout << "player::stop_and_open\n"; } void stopped_again(stop const&){std::cout << "player::stopped_again\n";} // guard conditions struct isCDLoaded { template <class EVT,class FSM,class SourceState,class TargetState> bool operator()(EVT const& ,FSM& fsm,SourceState& ,TargetState& ) { //return fsm.is_contained(); return fsm.is_flag_active<CDLoaded>(); } }; typedef player_ p; // makes transition table cleaner // Transition table for player struct transition_table : mpl::vector< // Start Event Next Action Guard // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Stopped , play , Playing , &p::start_playback >, Row < Stopped , open_close , Open , none , isCDLoaded >, a_row < Stopped , stop , Stopped , &p::stopped_again >, // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Open , open_close , Empty , &p::close_drawer >, // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Empty , open_close , Open , &p::open_drawer >, a_row < Empty , cd_detected , Stopped , &p::store_cd_info >, // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Playing , stop , Stopped , &p::stop_playback >, a_row < Playing , pause , Paused , &p::pause_playback >, a_row < Playing , open_close , Open , &p::stop_and_open >, // +---------+-------------+---------+---------------------+--------------- -------+ a_row < Paused , end_pause , Playing , &p::resume_playback >, a_row < Paused , stop , Stopped , &p::stop_playback >, a_row < Paused , open_close , Open , &p::stop_and_open > // +---------+-------------+---------+---------------------+--------------- -------+ > {}; // Replaces the default no-transition response. template <class FSM,class Event> void no_transition(Event const& e, FSM&,int state) { std::cout << "no transition from state " << state << " on event " << typeid(e).name() << std::endl; } }; // Pick a back-end typedef msm::back::state_machine<player_> player; // // Testing utilities. // static char const* const state_names[] = { "Stopped", "Open", "Empty", "Playing", "Paused" }; void pstate(player const& p) { std::cout << " -> " << state_names[p.current_state()[0]] << std::endl; } void test() { player p; // needed to start the highest-level SM. This will call on_entry and mark the start of the SM p.start(); // tests some flags std::cout << "CDLoaded active:" << std::boolalpha << p.is_flag_active<CDLoaded>() << std::endl; //=> false (no CD yet) // go to Open, call on_exit on Empty, then action, then on_entry on Open p.process_event(open_close()); pstate(p); p.process_event(open_close()); pstate(p); p.process_event(cd_detected("louie, louie")); p.process_event(play()); // at this point, Play is active std::cout << "PlayingPaused active:" << std::boolalpha << p.is_flag_active<PlayingPaused>() << std::endl;//=> true std::cout << "FirstSong active:" << std::boolalpha << p.is_flag_active<FirstSongPlaying>() << std::endl;//=> true // make transition happen inside it. Player has no idea about this event but it's ok. p.process_event(NextSong());pstate(p); //2nd song active p.process_event(NextSong());pstate(p);//3rd song active p.process_event(PreviousSong());pstate(p);//2nd song active std::cout << "FirstSong active:" << std::boolalpha << p.is_flag_active<FirstSongPlaying>() << std::endl;//=> false std::cout << "PlayingPaused active:" << std::boolalpha << p.is_flag_active<PlayingPaused>() << std::endl;//=> true p.process_event(pause()); pstate(p); std::cout << "PlayingPaused active:" << std::boolalpha << p.is_flag_active<PlayingPaused>() << std::endl;//=> true // go back to Playing // as you see, it starts back from the original state p.process_event(end_pause()); pstate(p); p.process_event(pause()); pstate(p); p.process_event(stop()); pstate(p); std::cout << "PlayingPaused active:" << std::boolalpha << p.is_flag_active<PlayingPaused>() << std::endl;//=> false std::cout << "CDLoaded active:" << std::boolalpha << p.is_flag_active<CDLoaded>() << std::endl;//=> true // event leading to the same state p.process_event(stop()); pstate(p); } } int main() { test(); return 0; }

1 0

boost::hash and string conflicts
by Erik Scorelle 07 Apr '11

07 Apr '11

Hi, We have been using boost hash to hash filenames, but have found with some of our user data that certain strings will produce the same hash code ( "0012g6" and "0012fu" for example). Is there a recommended way to predict or resolve these sorts of conflicts? Thanks in advance!

4 5

[bjam] "Access is denied."
by Phil Bouchard 07 Apr '11

07 Apr '11

Greetings, I am trying to run bjam on Cygwin with the gcc toolset but I keep getting the following error messages: $ bjam toolset=gcc Access is denied. Access is denied. Access is denied. bjam works fine with msvc. Any help is appreciated. Thanks, -Phil

2 2

I just spent most of the day learning about range and lambda.
by Jake Brownson 07 Apr '11

07 Apr '11

Great work guys, those libraries are awesome. Boost makes c++ into a pretty nice place to be. I've been using serialization, preprocessor, and shared_ptr for awhile now which are also pretty great too.

4 3

compilation errors of Parallel BGL
by Da Zheng 07 Apr '11

07 Apr '11

Hello, I try to build Parallel BGL in Ubuntu. I use gcc version 4.4.5, boost v1.46.1, openmpi v1.4.1-3ubuntu1. I believe my environment fulfills your requirements. Then I download parallel-bgl-0.7.0.tar.gz, and try to compile it, but I get some compilation errors. Below shows all my steps of compiling the library. cmake works fine. zhengda@zhengda-desktop:~/graph/parallel-bgl-0.7.0$ BOOST_ROOT=/home/zhengda/graph/boost_1_46_1 cmake . -- The C compiler identification is GNU -- The CXX compiler identification is GNU -- Check for working C compiler: /usr/bin/gcc -- Check for working C compiler: /usr/bin/gcc -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Check for working CXX compiler: /usr/bin/c++ -- Check for working CXX compiler: /usr/bin/c++ -- works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Found MPI: /usr/lib/libmpichcxx.so -- Boost version: 1.46.1 -- Found the following Boost libraries: -- mpi -- serialization -- filesystem -- system -- Configuring done -- Generating done -- Build files have been written to: /home/zhengda/graph/parallel-bgl-0.7.0 When I compile the code, it shows the following errors. zhengda@zhengda-desktop:~/graph/parallel-bgl-0.7.0$ make Scanning dependencies of target boost_graph_mpi [ 2%] Building CXX object libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/mpi_process_group.cpp.o In file included from /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/mpi_process_group.hpp:807, from /home/zhengda/graph/parallel-bgl-0.7.0/libs/graph_parallel/src/mpi_process_group.cpp:13: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp: In function ‘int boost::graph::distributed::receive(const boost::graph::distributed::mpi_process_group&, int, int, T&)’: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:722: error: ‘stderr’ was not declared in this scope /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:724: error: there are no arguments to ‘fprintf’ that depend on a template parameter, so a declaration of ‘fprintf’ must be available /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:724: note: (if you use ‘-fpermissive’, G++ will accept your code, but allowing the use of an undeclared name is deprecated) /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp: In function ‘typename boost::enable_if<boost::mpi::is_mpi_datatype<T>, std::pair<int, unsigned int> >::type boost::graph::distributed::receive(const boost::graph::distributed::mpi_process_group&, int, int, T*, size_t)’: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:743: error: ‘stderr’ was not declared in this scope /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:745: error: there are no arguments to ‘fprintf’ that depend on a template parameter, so a declaration of ‘fprintf’ must be available /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp: In function ‘int boost::graph::distributed::receive(const boost::graph::distributed::mpi_process_group&, int, int, T&) [with T = int]’: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:609: instantiated from ‘void boost::graph::distributed::mpi_process_group::global_trigger_launcher<Type, Handler>::receive(const boost::graph::distributed::mpi_process_group&, int, int, boost::graph::parallel::trigger_receive_context, int) const [with Type = int, Handler = void (*)(const boost::graph::distributed::mpi_process_group&, int, int, int, bool)]’ /home/zhengda/graph/parallel-bgl-0.7.0/libs/graph_parallel/src/mpi_process_group.cpp:1110: instantiated from here /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:722: error: ‘fprintf’ was not declared in this scope /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp: In function ‘int boost::graph::distributed::receive(const boost::graph::distributed::mpi_process_group&, int, int, T&) [with T = boost::graph::distributed::mpi_process_group::outgoing_messages]’: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:609: instantiated from ‘void boost::graph::distributed::mpi_process_group::global_trigger_launcher<Type, Handler>::receive(const boost::graph::distributed::mpi_process_group&, int, int, boost::graph::parallel::trigger_receive_context, int) const [with Type = boost::graph::distributed::mpi_process_group::outgoing_messages, Handler = void (*)(const boost::graph::distributed::mpi_process_group&, int, int, boost::graph::distributed::mpi_process_group::outgoing_messages&, bool)]’ /home/zhengda/graph/parallel-bgl-0.7.0/libs/graph_parallel/src/mpi_process_group.cpp:1110: instantiated from here /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/detail/mpi_process_group.tpp:722: error: ‘fprintf’ was not declared in this scope make[2]: *** [libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/mpi_process_group.cpp.o] Error 1 make[1]: *** [libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/all] Error 2 make: *** [all] Error 2 So I modify boost/graph/distributed/detail/mpi_process_group.tpp, and insert "#include <stdio.h>". It seems the error above is solved, but I get a new error. This time I don't know how to fix this error. zhengda@zhengda-desktop:~/graph/parallel-bgl-0.7.0$ make [ 2%] Building CXX object libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/mpi_process_group.cpp.o [ 5%] Building CXX object libs/graph_parallel/src/CMakeFiles/boost_graph_mpi.dir/tag_allocator.cpp.o Linking CXX static library ../../../lib/libboost_graph_mpi.a [ 5%] Built target boost_graph_mpi Scanning dependencies of target adjlist_build_test [ 8%] Building CXX object libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/adjlist_build_test.cpp.o In file included from /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/adjacency_list.hpp:3961, from /home/zhengda/graph/parallel-bgl-0.7.0/libs/graph_parallel/test/adjlist_build_test.cpp:10: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/adjlist/serialization.hpp: In function ‘std::vector<int, std::allocator<int> > boost::detail::parallel::available_process_files(const std::string&)’: /home/zhengda/graph/parallel-bgl-0.7.0/boost/graph/distributed/adjlist/serialization.hpp:97: error: conversion from ‘boost::filesystem3::path’ to non-scalar type ‘std::string’ requested make[2]: *** [libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/adjlist_build_test.cpp.o] Error 1 make[1]: *** [libs/graph_parallel/test/CMakeFiles/adjlist_build_test.dir/all] Error 2 make: *** [all] Error 2 Did I do something wrong? Why can I not compile the library? Thanks, Da

3 2

Re: [Boost-users] (no subject)
by Gabriel Marchand 07 Apr '11

07 Apr '11

Thank you for this solution. It compiles. Nevertheless, I still don't achieve what I want to do and I don't understand a few things (I attached the code): 1/ I run the connected_component() function after defining the graph graph_t G(num_nodes) and I expect that there should be as many components as num_nodes as I didn't add the edges yet. However the return of the function tells me there is 1 component ?? After adding the edges between all vertices, the number of components is 1, as it should be. But when I use the function clear_vertex(), I expect to create a new component as the vertex cleaned contain no more edges. The number of components is still one. What I am doing wrong here? 2/I need to know the size in number of vertices of the components, therefore I need to use something like for (i = 0; i != num_nodes; ++i) cout << "Vertex " << i <<" is in component " << component[i] << endl; cout << endl; but of course with our changings done this can't be used. Do you have an idea how I could extract this information. Gabriel --- En date de : Jeu 7.4.11, Cedric Laczny <cedric.laczny(a)gmx.de> a écrit : De: Cedric Laczny <cedric.laczny(a)gmx.de> Objet: Re: [Boost-users] (no subject) À: boost-users(a)lists.boost.org Date: Jeudi 7 avril 2011, 6h50 Alright, now that was a somehow tricky one... Details s. below. On Wednesday, 6. April 2011 12:55:52 Gabriel Marchand wrote: > Hi, I am new to boost and I face the same problem mentioned here with > "connected_components()" using a mutable graph (hence using a list for the > vertex container). I try the detailed solution of Aaron Windsor in the > previous post but I still get some errors at compilation. Here is my > little code: > > > > #include <boost/graph/adjacency_list.hpp> > #include <boost/graph/connected_components.hpp> > > using namespace std; > using namespace boost; > > Box& b = conf.GetBox(); > > > if (!ai) { > ai = new AtomList(); > *ai = conf.GetAtoms().AtomSelector(selecti, b); > } > > typedef boost::adjacency_list< > listS, > listS, > undirectedS, > property<vertex_index_t, size_t>, > no_property> > > > graph_t; > > graph_t g(ai->size()); > property_map<graph_t, vertex_index_t>::type index = get(vertex_index, > g); graph_traits<graph_t>::vertex_iterator vi, vi_end; > graph_traits<graph_t>::vertices_size_type cnt = 0; > for(tie(vi,vi_end) = vertices(g); vi != vi_end; ++vi) > put(index, *vi, cnt++); > > for (unsigned int i = 0; i < ai->size(); ++i) { > Cartesian& crdi = ai->GetElement(i)->pos; > for (unsigned int j = i + 1; j < ai->size(); ++j) { > Cartesian& crdj = ai->GetElement(j)->pos; > double r = conf.GetBox().CartesianDistance(crdi, crdj); > if (r < rcut) { > add_edge(vertex(i,g), vertex(j,g), g); // Add edge to Map > if distance criteria is satisfied > > } > } > } > > vector<int> component(num_vertices(g)); > int num_clusters = connected_components(g, &component[0], > vertex_index_map(index)); > In the example for connected_components() you used, you probably found the "&component[0]" notation and simply used it. However, this does _not_ work for graphs with listS, only for vecS. So basically what you have to do is to define your own components map and _not_ use that wrapper around a vector. As boost seems to model vertices with listS as void* this can hardly work as it is missing the index that it would normally have when using vecS. Also, for vecS, vertices seem to be modeled as int and therefore can serve as indices for vectors directly. So, to make your code compiles, you should use: typedef graph_traits<graph_t>::vertex_descriptor Vertex; map< Vertex, unsigned int > index_std_map; typedef associative_property_map< map< Vertex, unsigned int > > IndexMap; IndexMap index( index_std_map ); // put() the values by iterating over all vertices instead of getting the vertex_index from the graph via "get(vertex_index, G)". With this, you can define your own components-map: vector<int> v_component(num_vertices(G)); iterator_property_map< vector<int>::iterator, IndexMap > component( v_component.begin(), index ); and call connected_components() via: int num_clusters = connected_components(G, component, vertex_index_map(index)); > > > > The compilation is successfull until I add the last line: > int num_clusters = connected_components(g, &component[0], > vertex_index_map(index)); > > > Hope somebody can give a hint > gabriel Hope that I could help and that this solution will be correct for you. Please, of course, test the whole approach in order to make sure it works as you expect it to. Generally it should, as long as the mapping from vertex to index/offset is correct. If there are further questions about the problem or my solution, feel free to mail again. Best, Cedric _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1

Simple Lambda Problem
by Richard Howard 07 Apr '11

07 Apr '11

I am currently trying to get to grips with using Lambda, but I have hit a problem fairly early on. Can anyone explain what I am doing wrong. bool isCar(std::string const & val) { return val.find("car") != std::string::npos; } int main() { std::list<std::string> l; l.push_back("car"); l.push_back("blue car"); l.push_back("red van"); l.push_back("bike"); l.push_back("blue car"); // ... l.remove_if(&isCar); // this works and is how I have done things in the past but I was hoping to avoid the need for the separate function by using Lambda. // I have tried the following l.remove_if(boost::lambda::var(boost::lambda::bind(&std::string::find, boost::lambda::_1, "car")) != std::string::npos); // I am not sure why this does not work, can anyone point me in the right direction? }

2 1

[gil] Processing huge images
by Franz Alt 07 Apr '11

07 Apr '11

Hi all, I'm very new in Boost.GIL so I hope anyone could help me with my question. Imagine several images all with identical dimensions (e.g. 10000x10000). All images form one huge "virtual" image. Because every image is very big (~ 50M) I couldn't copy them together to one image in memory. I want to perform some algorithms (e.g. x-gradient) over the huge image with GIL in a (hopefully) elegant way. The user should use GIL on the huge image "as normal" as possible. Does anyone know how I could handle this problem? Regards Franz

4 6