- GiNaC-list - ginac.de

symtab problem with multithread
by Kraus Philipp 21 Aug '10

21 Aug '10

Hi, I try to use a GiNaC::symtab in a thread object. I copy a const reference to the threaded method and get problems on reading the symbols. If I take a symbol from the table with GiNaC::ex x = m_symbols["mysymbol"], I would run on > 1 thread into errors. Valgrind shows this information: Possible data race during write of size 4 at 0x1ea5c04 by thread #1 ==38631== at 0x8137: std::_Rb_tree<std::string, std::pair<std::string const, GiNaC::ex>, std::_Select1st<std::pair<std::string const, GiNaC::ex> >, std::less<std::string>, std::allocator<std::pair<std::string const, GiNaC::ex> > >::_M_erase(std::_Rb_tree_node<std::pair<std::string const, GiNaC::ex> >*) (in main) ==38631== by 0x92BF: boost::_bi::storage1<boost::_bi::value<worker<double, 2ul> > >::~storage1() (in main) ==38631== by 0xE0B0: boost::thread* boost::thread_group::create_thread<boost::_bi::bind_t<void, boost::_mfi::mf0<void, worker<double, 2ul> >, boost::_bi::list1<boost::_bi::value<worker<double, 2ul> > > > >(boost::_bi::bind_t<void, boost::_mfi::mf0<void, worker<double, 2ul> >, boost::_bi::list1<boost::_bi::value<mworker<double, 2ul> > > >) (in main) I have copied the table with std::copy( p_syms.begin(), p_syms.end(), std::inserter(m_symbols, m_symbols.begin())); p_syms is the original table and m_symbols should be the copy, but it doesn't work. In my thread object I can use a deep copy of the symbol table, because every thread is independed of the other threads. I'm not sure that the problem is a GiNaC problem, but you can give some tips perhaps. Thanks Phil

2 1

ginsh: valgrind reports 26 bytes lost
by Joerg Arndt 17 Aug '10

17 Aug '10

While we are at it: % valgrind --leak-check=full ginsh ==5463== Memcheck, a memory error detector ==5463== Copyright (C) 2002-2009, and GNU GPL'd, by Julian Seward et al. ==5463== Using Valgrind-3.5.0 and LibVEX; rerun with -h for copyright info ==5463== Command: ginsh ==5463== ginsh - GiNaC Interactive Shell (ginac V1.5.8) __, _______ Copyright (C) 1999-2010 Johannes Gutenberg University Mainz, (__) * | Germany. This is free software with ABSOLUTELY NO WARRANTY. ._) i N a C | You are welcome to redistribute it under certain conditions. <-------------' For details type `warranty;'. Type ?? for a list of help topics. > [pressing ctrl-d to exit] ==5463== ==5463== HEAP SUMMARY: ==5463== in use at exit: 743,496 bytes in 486 blocks ==5463== total heap usage: 3,081 allocs, 2,595 frees, 980,701 bytes allocated ==5463== ==5463== 26 bytes in 1 blocks are definitely lost in loss record 15 of 55 ==5463== at 0x4C26C3A: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) ==5463== by 0x59FEEA2: cln::xmalloc(unsigned long) (cl_malloc.cc:28) ==5463== by 0x5A077DD: cln::cl_make_heap_string(char const*) (cl_st_make1.cc:20) ==5463== by 0x5A05866: cln::cl_prin_globals_init_helper::cl_prin_globals_init_helper() (string.h:83) ==5463== by 0x5B27C05: global constructors keyed to cl_SV_number_aprint.cc (io.h:78) ==5463== by 0x5B27C65: ??? (in /usr/local/lib/libcln.so.6.0.1) ==5463== by 0x59FAF8A: ??? (in /usr/local/lib/libcln.so.6.0.1) ==5463== ==5463== LEAK SUMMARY: ==5463== definitely lost: 26 bytes in 1 blocks ==5463== indirectly lost: 0 bytes in 0 blocks ==5463== possibly lost: 0 bytes in 0 blocks ==5463== still reachable: 743,470 bytes in 485 blocks ==5463== suppressed: 0 bytes in 0 blocks ==5463== Reachable blocks (those to which a pointer was found) are not shown. ==5463== To see them, rerun with: --leak-check=full --show-reachable=yes ==5463== ==5463== For counts of detected and suppressed errors, rerun with: -v ==5463== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 2 from 2)

1 0

remove entry from symtab
by Kraus Philipp 16 Aug '10

16 Aug '10

Hi, I'm using GiNaC 1.5.8 and have declared a GiNaC::symtab with symbols. How can I remove a symbole (by string name or GiNaC::symbol expression) from the table? I don't find anything about the symtab in the documentation Thanks Phil

2 2

GiNaC kudos
by Stephen Montgomery-Smith 10 Aug '10

10 Aug '10

Hi Guys, I just wrote some programs that use GiNaC to multiply polynomials whose degrees can get into the 1000's, and whose coefficients are as large as 10^10000. GiNaC performed extraordinarily well, and even went a little faster than Mathematica! I am very impressed! (If you are interested, I am trying to analyze results of Luria-Delbruck experiments using Bayesian statistics.) Stephen

2 3

arithmetic error on concat two expressions
by Kraus Philipp 07 Aug '10

07 Aug '10

Hi, I'm new at GiNaC and I use the lib for create a gradient descent algorithm (symbolic calculating). I have a formula "a*x^3+b*y^4" and I subsitute the formula in the error function "0.5*(t - <formula>)^2". After the substitution GiNaC returns the expression "(0.5)*(a*x^3+b*y^4-t)^2". I use this code: GiNaC::ex m_expression; GiNaC::symtab m_exprtable; try { m_expression = l_parser( "a*x^3+b*y^4" ); m_exprtable = l_parser.get_syms(); } catch (...) {} GiNaC::ex l_full; GiNaC::symtab l_table(m_exprtable); l_table["formula"] = m_expression; GiNaC::parser l_parser(l_table); try { l_full = l_parser( "0.5*(t - formula)^2" ); // (*) } catch (...) {} std::cout << l_full << std::endl; I using the version 1.5.8 (on 1.5.7 I had got a compiler error). Is this the correct way for concating two symbolic expression? Do I have forgot anything? Thanks Phil

4 7

is ginac thread safe for this case?
by Cristobal Navarro 26 Jul '10

26 Jul '10

will ginac be thread safe for this case? 1) threads access diferent expressions which are independant 2) each thread will increase the expression its working in.. for example thread1...expression1 += x+y thread2.... expression2 += 2*x.... and so on. 2) no function evaluation is done never. 3) however, all expressions share the same global symbols ( eg, expressions made of global symbols x and y ). 4) threads access the same common matrix of expressions, but two threads will never access the same cell. i'am testing at the moment, and is working but in the long run i am not sure if problems will arise. any information, welcome. thanks in advance. Cristobal <http://www.youtube.com/neoideo>

5 22

Store matrix, then read it?
by Cristobal Navarro 07 Jul '10

07 Jul '10

Hello, i am experiencing a little trouble when reading my stored matrix. when i write and read the same matrix i stored, it works. but if i only run the program to read the matrix file, i get this error. terminate called after throwing an instance of 'std::runtime_error' what(): unknown matrix dimensions in archive This is a one-file example source code i made for you guys so you can test, its based on the same problem i have on my bigger code. try running it as it is, then comment the "write" method and run it again so it only reads the matrices, youll get the error eventually. but you will never get it if you write and read, its weird. how can this be solved? //CODE #include <fstream> #include <ginac/ginac.h> using namespace std; using namespace GiNaC; symbol q("q"),v("v"); void read( string filename, GiNaC::matrix* qvCol, GiNaC::matrix* qvMat ); void write( string filename, GiNaC::matrix qvCol, GiNaC::matrix qvMat ); int main(int argc, char** argv){ GiNaC::matrix qvMat(2,2), qvCol(2,1), rm1, rm2; //mat1 qvCol(0,0) = q*v; qvCol(1,0) = 2*q*v; //mat2 qvMat(0,0) = q*v; qvMat(0,1) = 2*q*v; qvMat(1,0) = 3*q*v; qvMat(1,1) = 4*q*v; cout << endl << "m1: " << endl << qvCol << endl; cout << "m2: " << endl << qvMat << endl << endl; write("example.gin", qvCol, qvMat); read("example.gin", &rm1, &rm2); cout << endl << "rm1: " << endl << rm1 << endl; cout << "rm2: " << endl << rm2 << endl << endl; return 0; } void read( string filename, GiNaC::matrix* qvCol, GiNaC::matrix* qvMat ){ cout << "Main::Read File......."; fstream fr(filename.c_str(),ios::in|ios::binary); if( !fr.is_open() ){ printf("bad file\n"); return; } GiNaC::archive_node canCol, canMat; GiNaC::lst syms(q, v); fr >> canCol; fr >> canMat; qvCol->read_archive(canCol, syms); qvMat->read_archive(canMat, syms); fr.close(); cout << "OK: " << filename << endl; } void write( string filename, GiNaC::matrix qvCol, GiNaC::matrix qvMat ){ cout << "Main::Write File......"; fflush(stdout); //abrir archivo para escritura fstream fw(filename.c_str(),ios::out|ios::binary); GiNaC::archive_node anCol, anMat; qvCol.archive( anCol ); qvMat.archive( anMat ); fw << anCol; fw << anMat; fw.close(); cout << "OK: " << filename << endl; } //END

2 2

GiNaC release 1.5.8
by Jens Vollinga 06 Jul '10

06 Jul '10

Hi everybody, GiNaC 1.5.8 is out and available. The changes are: * Changed default behaviour of conjugate() for user-defined functions to avoid wrong evaluations. A user-defined conjugate_func must now be explicitly registered. * Fixed: Parser now handles abbreviations as advertized in the manual. * Fixed two bugs in the GCD code (one infinite loop, one miscalculation). * Fixed dangerous iterator use. * Fixed a memory leak in excompiler. * Fixed several syntax issues wrt to gcc 4.5 and wrt to compile option -std=cxx0x. * Fixed a wrong URL and a texinfo problem in the manual. * Added degree_vector() utility function. As always, this release can be downloaded from http://www.ginac.de/Download.html Best wishes, Jens

1 0

parser and subs problem
by Felipe Bordeu 29 Jun '10

29 Jun '10

Hello, I'm trying to use the ginac parser, but doesn't work as expected. Page 81 of GiNaC 1.5.7 (an open framework ...) tutorial (29 march 2010) " It’s also possible to map input (sub)strings to arbitrary expressions: GiNaC::symbol x, y; GiNaC::symtab table; table["x"] = x+log(y)+1; GiNaC::parser reader(table); GiNaC::ex e = reader("5*x^3 - x^2"); " but I get the error terminate called after throwing an instance of 'std::invalid_argument' what(): find_or_insert_symbol: name "x" does not correspond to a symbol Abandon function find_or_insert_symbol in file ginac/parser/parse_context.cpp:49 can handle only symbols not expressions. So I tried to use only symbols and then do some subs to get the correct expression without success. GiNaC::symbol x, y; GiNaC::ex expr1 = x*y; GiNaC::ex mat1 = GiNaC::symbolic_matrix(2, 2, "K"); GiNaC::ex mat2 = GiNaC::symbolic_matrix(2, 2, "T"); GiNaC::ex expr2 = expr1.subs(GiNaC::lst(x==mat1,y==mat2)); cout << expr2 << endl; cout << expr2.evalm() << endl; and I get the error : [[K00,K01],[K10,K11]]*[[T00,T01],[T10,T11]] terminate called after throwing an instance of 'std::runtime_error' what(): matrix::mul_scalar(): non-commutative scalar Abandon do I have to use indexed objects to get the correct answer??? Thanks -- Felipe

2 1

1by1 matrix to scalar
by Felipe Bordeu 23 Jun '10

23 Jun '10

Hello, I have a little problem, I'm dealing functions made of matrix products. My question is, how can "force" ginac to convert automatically all the 1by1 matrices to scalars?. or do I have to do it manually?? Here is a simple example. void main2(){ ex m1=symbolic_matrix(1, 3, "x"); ex m2=symbolic_matrix(3, 1, "y"); cout << m1 << std::endl; cout << m2 << std::endl; ex m3 = (m1*m2).evalm(); cout << m3 << std::endl; cout << ex_to<matrix>(m3)(0,0)<< std::endl;; ex m4 = m3-ex_to<matrix>(m3)(0,0); cout << m4 << std::endl; }; the output : [[x0,x1,x2]] [[y0],[y1],[y2]] [[y1*x1+y2*x2+y0*x0]] y1*x1+y2*x2+y0*x0 -y1*x1+[[y1*x1+y2*x2+y0*x0]]-y2*x2-y0*x0 the last line "should" be zero ?-- thanks Felipe

4 5