March 2012 - Boost-users - lists.boost.org

Troubles with boost's shared_array, among other things
by Ted Byers 24 Mar '12

24 Mar '12

I highlighted in bold the remarks of most interest to boost users. The rest just shows you what I have done, and provides the context in which my problem arises. I commented out the problem code below, in an effort to determine why the core dump happened, only to see another problem, with the execution path not even entering my regressionPCA constructor. My program is short and simple, designed to test my PCA regression class. My basicGenerator, declared in data.generator.h, is already known to work flawlessly. This program is the unti test for my regressionPCA class. #include <iostream> #include <vector> #include "data.generator.h" #include "reg.class.h" int main(int argc, char* argv[]) { std::cout << "Flag A" << std::endl; basicGenerator bg; std::cout << "Sample size: " << bg.get_sampleSize() << std::endl; bg.makeData(); std::cout << "Flag B" << std::endl; std::vector<std::vector<double> > x; std::vector<double> y; bg.getDataForRegression(x,y); unsigned int imax = y.size(); for (unsigned int i = 0 ; i < imax ; i++) { std::cout << i << "\t" << y[i] << "\t" << x[i][0] << "\t" << x[i][1] << std::endl; } std::cout << "Flag C" << std::endl; regressionPCA rpca(x,y); return 0; } Here is the tail end of the output it produces: 89 0.416356 0.565407 -0.114605 90 -0.258883 -0.585597 -0.183372 91 0.833355 0.914344 0.81171 92 -0.0338814 -0.00264442 -0.118973 93 1.13764 0.41599 1.33175 94 -1.86323 -1.90867 -1.35118 95 0.907604 1.14917 0.621669 96 2.1166 1.06194 1.1703 97 0.159543 0.14446 -0.665135 98 -0.508617 -0.370597 -0.703225 99 2.69086 2.75267 1.40633 Flag C Ted@Ted-acer-i7w7 ~/New.System/tests $ As you can see, the last output is from the flag I set immediately before creating my instance of regressionPCA on the stack. Here is the declaration of that class. #ifndef REG_CLASS_H #define REG_CLASS_H #include <iosfwd> #include <vector> //#include <boost/smart_ptr/shared_array.hpp> #include <gsl/gsl_linalg.h> class regressionPCA { private: unsigned int nrows, ncols; // boost::shared_array<double> B, Y, U, V, S; regressionPCA(void) {}; // makes default construction impossible public: regressionPCA(const std::vector<std::vector<double> >&, const std::vector<double>&); void Reset(const std::vector<std::vector<double> >&, const std::vector<double>&); inline void setNrows(unsigned int v) { nrows = v;}; inline void setNcols(unsigned int v) { ncols = v;}; inline unsigned int getNrows(void) const { return nrows; }; inline unsigned int getNcols(void) const { return ncols; }; }; #endif Please note that instead of just dying without producing output, the program died most ungracefully with a core dump when I included boost's shared_array. I have no idea why. Here is the implementation of the regressionPCA class: #include "reg.class.h" #include <iostream> #include <algorithm> #include <gsl/gsl_linalg.h> regressionPCA::regressionPCA(const std::vector<std::vector<double> >&data, const std::vector<double>& Ydata) { std::cout << "Flag ALPHA" << std::endl; Reset(data,Ydata); } void regressionPCA::Reset(const std::vector<std::vector<double> >&data, const std::vector<double>& Ydata) { unsigned int r(0),c(0),n(0); std::cout << "r: " << r << "\tc: " << c << "\tn: " << n << std::endl; c = data.size(); std::cout << "r: " << r << "\tc: " << c << "\tn: " << n << std::endl; r = data[0].size(); std::cout << "r: " << r << "\tc: " << c << "\tn: " << n << std::endl; setNrows(r); setNcols(c); n = r * c; std::cout << "r: " << r << "\tc: " << c << "\tn: " << n << std::endl; double* B = new double[c]; double* Y = new double[r]; double* U = new double[n]; double* V = new double[c*c]; double* S = new double[c]; /* boost::shared_array<double> B(new double[c]), boost::shared_array<double> B(new double[c]), Y(new double[r]), U(new double[n]), V(new double[c*c]), S(new double[c]); B = Btmp; Y = Ytmp; U = Utmp; V = Vtmp; S = Stmp;*/ double *bptr = U.get(); std::cout << "Flag1" << std::endl; std::vector<std::vector<double> >::const_iterator it = data.begin(), end = data.end(); while (it != end) { bptr = std::copy(it->begin(), it->end(),bptr); ++it; } std::cout << "Flag2" << std::endl; bptr = Y.get(); std::copy(Ydata.begin(),Ydata.end(),bptr); gsl_matrix_view Um = gsl_matrix_view_array(U.get(), getNrows(), getNcols()); gsl_matrix_view Ym = gsl_vector_view_array(Y.get(), getNrows()); gsl_matrix_view Bm = gsl_vector_view_array(B.get(), getNcols()); gsl_matrix_view Sm = gsl_vector_view_array(S.get(), getNcols()); gsl_matrix_view Vm = gsl_matrix_view_array(U.get(), getNcols(), getNcols()); std::cout << "Flag3" << std::endl; gsl_linalg_SV_decomp_jacobi(Um,Vm,Sm); std::cout << "Flag4" << std::endl; gsl_linalg_SV_solve(Um,Vm,Sm,Ym,Bm); std::cout << std::endl << std::endl << "Sv = " << gsl_vector_get(Sm,0) << "\t" << gsl_vector_get(Sm,1) << std::endl; std::cout << std::endl << std::endl << "V = " << std::endl; std::cout << "\t" << gsl_matrix_get(Vm,0,0) << "\t" << gsl_matrix_get(Vm,0,1) << std::endl; std::cout << "\t" << gsl_matrix_get(Vm,1,0) << "\t" << gsl_matrix_get(Vm,1,1) << std::endl; std::cout << std::endl << std::endl << "Beta = " << gsl_vector_get(Bm,0) << "\t" << gsl_vector_get(Bm,1) << std::endl; }; Note, I know the execution path does not get into the constructor becuase the line "std::cout << "Flag ALPHA" << std::endl;" does not produce output. So, a couple mysteries. 1) Why does merely including boost's shared_array cause a core dump? 2) when I resort to naked array pointers (yes, I know about the memory leak, but that is the least of my troubles right now, and I would have avoided that had boost's shared array worked for me), why does execution not enter my constructor? Any insights would be appreciated. BTW: If you notice me doing something silly with GSL, I'd appreciate knowing about it, even though my problems do not appear to be related to GSL.

1 0

[BGL] Implementing own graph types: Overloading in namespace boost or ADL?
by Philipp Moeller 23 Mar '12

23 Mar '12

The documentation does not explicitly mention which of both approaches is recommended to implement the free function of the graph concepts for a custom graph. Usually the graph will be in its own namespace. Should the free functions be declared in the boost namespace or in the same namespace as the graph? This also would determine if BGL functions should be called qualified or unqualified outside of namespace boost. -- Philipp Moeller GeometryFactory

2 3

[Asio] Binding tcp socket to a particular network interface
by De Sousa, Dan 23 Mar '12

23 Mar '12

I'm trying to figure out the best cross platform solution to guarantee that a particular boost socket always binds to a particular network interface (eth0, eth4, etc). I know how to resolve the interface to an address and I see that I can specify an tcp endpoint to tcp socket on construction or with bind, but I also need to specify the port. Since I could be connecting to multiple servers in one instance, and have to deal with multiple instances of this library on the same machine, I can't rely a fixed port number. So my question(s): Is there a reliable cross platform way to get an endpoint to an address without a port? Or to do what boost tcp socket does by default, to get an unbound port for an address? Or generally guarantee that the socket is bound to a particular network interface? Short of finding a way to resolve ports, I don't see a way to do this repeatedly. For some context, we are trying to eliminate routing tables to specify the interface to particular servers. Thanks in advance for any suggestions! Dan ________________________________ IMPORTANT: The information contained in this email and/or its attachments is confidential. If you are not the intended recipient, please notify the sender immediately by reply and immediately delete this message and all its attachments. Any review, use, reproduction, disclosure or dissemination of this message or any attachment by an unintended recipient is strictly prohibited. Neither this message nor any attachment is intended as or should be construed as an offer, solicitation or recommendation to buy or sell any security or other financial instrument. Neither the sender, his or her employer nor any of their respective affiliates makes any warranties as to the completeness or accuracy of any of the information contained herein or that this message or any of its attachments is free of viruses.

1 0

[Regex] Is it possible to match any "linebreak"
by Christoph Duelli 23 Mar '12

23 Mar '12

Is there a way to have a regex like [[:digit:]]{3}([^\n]+)\n? have \n match any line breaking character? (The processed files might have Unix or DOS line endings.) I know I could modify the regex [[:digit:]]{3}([^\n\r]+)[\n\r]* but I was hoping there might be an easier solution like a character class [[:linebreak:]], or maybe an escaped character like \R or something like that. (I am using Boost 1.43 in the relevant project.) Thank you and best regards Christoph

2 4

BOOST_PP_ARRAY_PUSH_BACK
by petros 23 Mar '12

23 Mar '12

Hi, the following piece of code: #include <boost/preprocessor/array.hpp> #include <boost/preprocessor/array/push_back.hpp> #include <boost/preprocessor/iteration/local.hpp> #define barfoo (2, (1, 2) ) BOOST_PP_ARRAY_PUSH_BACK( barfoo, 3 ) gived the following compilation error ( MSVC2010, win7 ) : 1>Build started 3/22/2012 4:06:39 PM. 1>ClCompile: 1> opexp.cpp 1>opexp.cpp(6): error C2059: syntax error : 'constant' Any ideas why this happens ? Any help will be greatly appreciated. TIA, Petros

2 4

[Interprocess] How to interrupt interprocess_condition::wait()?
by wright_left 23 Mar '12

23 Mar '12

I had a simple concurrent queue that was working fine shuffling data between threads. Now I want to use it between processes. I switched out the boost::condition_variable for a boost::interprocess::interprocess_condition. The queue funcitons normally, except that I can't interrupt the thread waiting on the queue. Previously I simply called boost::thread::interrupt() and the thread would die (with the interrupt exception I assume). But interprocess_condition::wait() doesn't seem to care about the interrupt. Am I missing something? Do I need to use the interprocess_condition since the queue is in shared memory? Or can I go back to using the boost::condition_variable? -- View this message in context: http://boost.2283326.n4.nabble.com/Interprocess-How-to-interrupt-interproce… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

problems with boost bind and boost thread
by Marco 23 Mar '12

23 Mar '12

Hi, I have problems with the attached code (test.cpp, extracted from the original class only with the relevant code). The member template constructor of CCycleThread takes an argument and passes this to boost bind. The binding is to a another member template method which is the thread function for boost thread: boost::thread(boost::bind(&CCycleThread::execute<T>,this,t) The template parameter type is deduced by the argument type passed by the user. If the type is an ordinary function the code works fine. If the argument itself is another boost bind object the compilation fails! I tested the following within the constructor: auto test = boost::bind(&CCycleThread::execute<T>,this,t) // -> ok test(); // -> compile error The call of the operator() of the bind object is the problem. Interestingly, if one binding is boost::bind and the other one is std::bind the code works. Any ideas what's the problem? How can I fix this? Using std and boost bind is more a hack than a real solution :/

4 7

Clarification on McGregor common subgraph library
by Suhas Chakravarty 22 Mar '12

22 Mar '12

Hi The documentation for the mcgregor common subgraph library, at http://www.boost.org/doc/libs/1_48_0/libs/graph/doc/mcgregor_common_subgrap…, says that " "mcgregor_common_subgraphs_maximum_unique(...); Explores the *entire* search space and invokes the user_callback afterward with each of the largest, unique discovered subgraphs. Returning false from the user_callback will *not* terminate the search because it is invoked after the search has been completed. " Given graph 1 A |__ B_| (Self loop on B) | C / \ D E \ / F | G / \ \ H \ / I And graph 2 A |__ B_| (Self loop on B) | C | D | E / \ \ F \ / G The application of mcgregor_common_subgraphs_maximum_unique gives me the following mapping: G1 <-> E2 H1 <-> F2 I1 <-> G2 But in addition, I was also expecting the mappings A1<-A2, B1<->B2 and C1<-E2, D1<-F2, since these are separate, maximum subgraphs I see these (along with other smaller subgraphs) when I use simply "mcgregor_common_subgraphs" or "mcgregor_common_subgraphs_unique". Why do I not see them with the "maximum" version? Thanks Suhas

2 1

[mpi] all_reduce and MPI_IN_PLACE like feature
by Alain O Miniussi 22 Mar '12

22 Mar '12

Hi, MPI_Allreduce offers the possibility to use the same buffer for the send and recv data: http://www.open-mpi.org/doc/v1.4/man3/MPI_Allreduce.3.php#toc8 It seems that boost's all_reduce does not allow such a feature. First, I would like to know if the previous statement is correct and, if yes, if it was just overlooked or if there was a reason for that? Just in case, I filed a ticket regarding that issue, along with a tentative patch, in https://svn.boost.org/trac/boost/ticket/6704 Regards, Alain

1 0

Convert asio::error_code into boost::system::error_code
by Jens Weller 22 Mar '12

22 Mar '12

Hi, I'm in a usecase, where we use a library, which brings its own asio lib with it: http://www.resiprocate.org/ReTurn_Overview If this is correct, they're using asio 1.2 in its header only version. I've build an API which uses boost::asio to send data over the network. One of our goals is now, to also include stun/turn, so we have to use this library. To enable the errorreporting, I need to convert the asio::error_code coming from reTurn to boost::system::error_code, so that I can call my notify method with it: void notifyError(const boost::system::error_code& ec); any expierence in this? The constructors don't seem to fit, as error_category is in asio a enum, and in boost::system a class f.e. gz Jens Weller -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de

2 2