- GiNaC-list - ginac.de

Bug Report
by Jon Graves 20 Jul '11

20 Jul '11

OS: Ubuntu 11.04 Compiled with g++: g++ -o test test.cpp -lcln -lginac GiNaC ver: 1.60 Sample version of program attached. Program generates Hermite polynomial to high degree. It then subs into it a series of numbers. The output RANDOMLY generates an incorrect answer. No answer after a substitution should ever be zero, yet zeros inevitably appear. The problem occurs once out of every 3 executions, on average.

2 1

Release 1.6.1
by Jens Vollinga 18 Jul '11

18 Jul '11

Hi everybody, GiNaC 1.6.1 is out and available. The changes are the following: * Fixed several internal assertions for the debug version of GiNaC. * More verbose report of (auto-)configuration problems. * Fixed a compile problem with ginsh using gcc 4.6. As always, this release can be downloaded from http://www.ginac.de/Download.html Regards, Jens

1 0

3 index tensors
by Gregory Stiesberg 09 Jul '11

09 Jul '11

Hello, Given a matrix of expressions A and vectors q,v,w, I'm trying to evaluate a tensor expression of the type diff(A_ij, q_k) * v_k * w_j . With i,j,k of type idx, and A,q of type Matrix, It would be nice to get a 3 index tensor from indexed( diff(A(i,j),q(k)), i,j,k) or by diff(indexed(A,i,j),indexed(q,k)) but these don't work. After looking at the definition of epsilon_tensor the simplest option seems to be to write a class indexed_from_function3 able to return a 3-index tensor from an arbitrary function of three indices: for example ex f(int i, int j, int k){ return( diff(A(i,j),q(k)) ); } ex tensexp0=indexed_from_function3(f, i,j,k) * indexed(v, k) * indexed(w, j) Is there already a way to construct an arbitrary 3-index tensor that I've not found yet? Thanks in advance, Gregg Stiesberg

1 0

Subresultant calculation
by Joachim Redies 07 Jul '11

07 Jul '11

Hello everbody, is it possible to calculate the subresultant of two multivariate polynomials (inherited from ex) with little development effort and knowledge about the mathematical background? I have seen that sr_gcd in normal.cpp uses subresultants. Is it possible to isolate the subresultant calulation? Thanks, Joachim

1 0

Windows build
by Kraus Philipp 02 Jul '11

02 Jul '11

Hello, I would like to use Ginac under Windows (with Visual Studio). Is there a build script like cmake to build the Ginac library? I have build Ginac with cygwin but in my opinion I can't link again the cygwin binaries with VS, because the linker creates different calls, so if I use Ginac with VS I must create the binary with VS Thanks Phil

2 2

issues with ginac-1.6.0
by PG CLARK 25 May '11

25 May '11

this is probably me having installed the new library wrong but when I run my newly compiled program I get this error message ./main: error while loading shared libraries: libginac.so.2: cannot open shared object file: No such file or directory any ideas what might be the problem? I ran the make check after installing and it passed fine? regards peter clark ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

3 3

Re: [GiNaC-list] issues with ginac-1.6.0
by PG CLARK 25 May '11

25 May '11

Please disregard that. It was a bug in ubuntu 11 that just got fixed. That's the good news. Bad news is my program still gets killed for over use of ram and swap space. Are you really sure 1.6.0 has the improvement to simplified index in them? Tell me do we have an algorithm for taking 2 symetry trees acting on the same indexes and merging them to create one with the symetrys of both? If we had that I'm pretty sure I could write a far more efficient simplification algorithm. Quoting Stephen Montgomery-Smith <stephen(a)missouri.edu>: > You probably need to do > > sudo ldconfig > > PG CLARK wrote: >> Sorry thats a school boy error yes I'm running ubuntu 11.4, kernal >> 2.6.38-8-generic, gnome 2.32.1 (yes most of this isn't relevant but >> I'm going for completeness) >> >> 1.9GiB ram, AMD Athlon 64 X2 dual core processor >> >> ginac 1.6.0 was installed with the standard install commands included >> in the install file >> ./configure >> make >> sudo make install >> make check >> >> all tests passed >> >> my terminal after I tried ldconfig looked like this. >> >> peter@peter-ubuntu:~$ man ldconfig >> peter@peter-ubuntu:~$ ldconfig >> /sbin/ldconfig.real: Can't create temporary cache file >> /etc/ld.so.cache~: Permission denied >> peter@peter-ubuntu:~$ cd Documents >> peter@peter-ubuntu:~/Documents$ gcc main.cpp -o main -lcln -lginac >> peter@peter-ubuntu:~/Documents$ ./main >> ./main: error while loading shared libraries: libginac.so.2: cannot >> open shared object file: No such file or directory >> >> For give me but prior to using this system all I did was a bit of >> darwin tinkering and before that MS-dos. I'm not too clear on the >> specifics of how linux handles it's libraries. >> >> Quoting "Richard B. Kreckel"<kreckel(a)ginac.de>: >> >> >>> You didn't tell us about your system or how and where you installed >>> the library. Buy maybe you do have to run ldconfig? >>> >>> -richy. >>> -- >>> Richard B. Kreckel >>> <http://www.ginac.de/~kreckel/> >>> _______________________________________________ >>> GiNaC-list mailing list >>> GiNaC-list(a)ginac.de >>> https://www.cebix.net/mailman/listinfo/ginac-list >>> >>> >> >> >> ---------------------------------------------------------------- >> This message was sent using IMP, the Internet Messaging Program. >> >> >> _______________________________________________ >> GiNaC-list mailing list >> GiNaC-list(a)ginac.de >> https://www.cebix.net/mailman/listinfo/ginac-list >> >> >> >> > > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

1 0

Release 1.6.0
by Jens Vollinga 22 May '11

22 May '11

Hi everybody, GiNaC 1.6.0 is out and available. You can find the changes at the end of this email. Note: This release is not binary compatible to the previous one. Also, we changed the name of the library filename from libginac-1.5 to libginac. The ABI versioning is now done in recommended libtool manner. The dynamic library will be created as libginac.so.2.0.0 this time. We also stopped to create extra branches in the repository for ABI compatible versions. The latest releases will from now on always be in master. Changes: * Fixed compilation issues with clang. * Fixed compilation issues with VC++. * Fixed a segfault in fsolve() and improved its numerical algorithm. * Fixed several memory leaks in power::eval(). * Fixed the substitution algorithm that led to wrong results (exp(log(x))->x). * Fixed a bug in add:eval() (real_part() had returned wrong results). * Fixed infinite loop bug when unarchiving realsymbol and possymbol. * Fixed bugs in the GCD code. * Fixed empty string bug in symbol::get_name(). * Fixed bugs concerning STL vector sizes that can potentially cause segfaults. * Fixed a bug in is_polynomial(). * Removed autogen dependencies. * Improved simplify_indexed() algorithm. * Changed the connection between library version and package version. * power::series() can handle now more expressions with singularities in the exponent. * Added conjugate() methods to functions cosh, sinh, tanh. As always, this release can be downloaded from http://www.ginac.de/Download.html Best wishes, Jens

1 0

is_polynomial gives strange results
by Florent Hivert 20 May '11

20 May '11

Hi there, I'm a regular user of GiNaC through PyNaC and Sage and I'm usually very happy with it. Good work guys ! However during on of my class I was quite surprised by a result from Sage. I've been able to trace back the result to GiNaC so I'd like to report it here. Note that it's the first time I go into the arcane of GiNaC so I may be wrong. Here is a code that reproduce the problem: #include <iostream> #include <ginac/ginac.h> using namespace std; using namespace GiNaC; int main() { symbol x("x"); ex poly; poly = sqrt(x*x+1)*sqrt(x+1); cout << poly << endl; cout << poly.is_polynomial(x) << endl; return 0; } The output is: popcorn-*/Sage-Work/GiNaC $ ./ginac sqrt(1+x)*sqrt(1+x^2) 1 so that sqrt(1+x)*sqrt(1+x^2) is considered as a polynomial in x ! Here are the standard info: I just installed cln-1.3.2.tar.bz2 and ginac-1.5.8.tar.bz2 from the sources. I'm using: popcorn-~/install $ c++ --version c++ (SUSE Linux) 4.5.0 20100604 [gcc-4_5-branch revision 160292] In a popcorn-~/install $ uname -a Linux popcorn.rouba.net 2.6.34.7-0.7-desktop #1 SMP PREEMPT 2010-12-13 11:13:53 +0100 x86_64 x86_64 x86_64 GNU/Linux Cheers, Florent Hivert

3 4

issues with simplify_indexed()
by PG CLARK 17 May '11

17 May '11

I've been trying to get a little program of mine to work and well it just doesn't terminate. Somewhere it gets snagged in an infinite loop in the library. I'm pretty sure this is a library bug and I've alerted development but in the mean while has an one encountered this before? Has any one figured out a work around? I've whittled my program down to the smallest most simple component I could get to non terminate. The code looks like this. #include <iostream> #include <ginac/ginac.h> using namespace std; using namespace GiNaC; int main() { symbol i_sym("i"), j_sym("j"), k_sym("k"), l_sym("l"), m_sym("m"), n_sym("n"); idx i(i_sym, 2), j(j_sym, 2), k(k_sym, 2), l(l_sym, 2), m(m_sym, 2), n(n_sym, 2); symbol alpha_sym("\\alpha"), beta_sym("\\beta"), gamma_sym("\\gamma"); idx Va(alpha_sym, 3), Vb(beta_sym, 3), Vg(gamma_sym, 3); symbol U("U"), v("v"), h("h"); exvector IJKLMNVec; idx IJKLMNIA[] ={i, j, k, l, m, n}; IJKLMNVec.assign (IJKLMNIA,IJKLMNIA+6); ex myEx = (indexed(h, IJKLMNVec)*indexed(U, i, Va) *indexed(U, k, Vb)*indexed(U, m, Vg)*indexed(v, j, Va) *indexed(v, l, Vb)*indexed(v, n, Vg)).simplify_indexed(); cout << myEx << endl; return 0; } This code never makes it to the cout code. it get trapped in a loop before then. Any one seen this before? ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

2 3