- GiNaC-list - ginac.de

'full' differentiation?
by Ralf Stephan 02 Aug '04

02 Aug '04

Hello, is there an easy way to get 'full' differentiation, ie, of all variables, instead of only the main variable? Example: Given symbols x,u, diff(x/u, x) would then produce (u-u'x)/u^2. At the moment, I'm abusing psi to get the effect but that is certainly not the best way. sincerely, ralf

2 1

get_free_indices() returns wrong result
by varg＠thsun1.jinr.ru 02 Aug '04

02 Aug '04

Hello! This simple program: #include <iostream> #include <ginac/ginac.h> using namespace std; using namespace GiNaC; int main(int argc, char** argv) { symbol A("A"); symbol i_sym("i"); idx i(i_sym, 3); ex test = indexed(A, i)*indexed(A,i); exprseq free_ind = test.get_free_indices(); cout << "Free indices of expression " << test << ": "; cout << free_ind << endl; return 0; } gives: Free indices of expression (A.i)^2: (.i) but not just an empty sequence, as I expected. This patch seems to fix the inconsistency: diff -u -r GiNaC-orig/ginac/indexed.cpp GiNaC/ginac/indexed.cpp --- GiNaC-orig/ginac/indexed.cpp 2004-06-10 02:19:36.000000000 +0400 +++ GiNaC/ginac/indexed.cpp 2004-07-26 22:31:52.000000000 +0400 @@ -499,10 +499,31 @@ return free_indices; } +// checks if index is symbolic and is_exactly_a<idx> +struct is_a_symb_idx : public std::unary_function<ex, bool> { + bool operator()(const ex & e) + { + return is_exactly_a<idx>(e) && ex_to<idx>(e).is_symbolic(); + } +}; + exvector power::get_free_indices() const { - // Return free indices of basis - return basis.get_free_indices(); + exvector tmp_inds = basis.get_free_indices(); + + if (exponent.info(info_flags::even)) { + // if the exponent is an even number, then any "free" index + // of base is actually a dummy index, if it + // 1) has no variance, + // 2) is not a numeric index + exvector really_free; + remove_copy_if(tmp_inds.begin(), tmp_inds.end(), + back_inserter(really_free), is_a_symb_idx()); + return really_free; + } else { + return tmp_inds; + } + } /** Rename dummy indices in an expression.

2 1

doc patch (overloaded user functions)
by Ralf Stephan 28 Jul '04

28 Jul '04

Hello, I would like to suggest the following patch: Regards, ralf --- ginac.texi~ Fri Jul 23 16:29:57 2004 +++ ginac.texi Wed Jul 28 11:30:44 2004 @@ -6307,6 +6307,14 @@ option which is explained in the next section. +@subsection Overloaded functions + +If you ever intend to construct functions that differ only in the number +of parameters but have the same name then the macros of the previous +section no longer can accomplish the task. In that case, we can only +advise you read the source code and how we did it in the cases of +@code{psi} and @code{zeta}. It can be found in the files @code{inifcns.h}, +(a)code{inifcns_gamma.cpp} and @code{inifcns_nstdsums.cpp}, respectively. @node Printing, Structures, Symbolic functions, Extending GiNaC @c node-name, next, previous, up

1 0

Majordomo is dead, long live Mailman!
by Christian Bauer 18 Jul '04

18 Jul '04

Hello! The GiNaC mailing lists are now running under the Mailman mailing list software. You should have received an Email with your personal subscription password and some instructions on how to change your subscription settings and how to unsubscribe. To fend off spam, only subscribed members can post to the lists now. If you are using multiple Email addresses for posting, you can subscribe all of them, and set all except one to "nomail" on the subscription options page, so you won't get duplicate messages. The list archives have been preserved and can be accessed here: http://thep.physik.uni-mainz.de/pipermail/ginac-list/ http://thep.physik.uni-mainz.de/pipermail/ginac-devel/ Have fun, Christian -- / Physics is an algorithm \/ http://www.uni-mainz.de/~bauec002/

1 0

doc patch
by Ralf Stephan 11 Jun '04

11 Jun '04

Hello, I thought this was missing from the documentation: --- ginac.texi.orig Mon Mar 15 18:47:17 2004 +++ ginac.texi Thu Jun 10 09:41:09 2004 @@ -4905,6 +4905,12 @@ @item @code{binomial(n, m)} @tab binomial coefficients @cindex @code{binomial()} +@item @code{bernoulli(n)} +@tab Bernoulli numbers +@cindex @code{bernoulli()} +@item @code{fibonacci(n)} +@tab Fibonacci numbers +@cindex @code{fibonacci()} @item @code{Order(x)} @tab order term function in truncated power series @cindex @code{Order()}

2 3

g++-3.3 and template specializations
by Kai Ludwig 08 Jun '04

08 Jun '04

Hi, there are some error messages, g++-3.3 gives me concerning missing template<> statements in GiNaC-1.2.1 Kai /usr/local/include/ginac/container.h:665: error: explicit specialization not preceded by `template <>' In file included from /usr/local/include/ginac/normal.h:29, from /usr/local/include/ginac/ginac.h:31, from ../genericRateLaw.hpp:32, from ../ecco.hpp:41, from eccoTest.cpp:1: /usr/local/include/ginac/lst.h:35: error: explicit specialization not preceded by `template <>' /usr/local/include/ginac/lst.h:38: error: explicit specialization not preceded by `template <>' /usr/local/include/ginac/lst.h:41: error: explicit specialization not preceded by `template <>' /usr/local/include/ginac/lst.h:44: error: explicit specialization not preceded by `template <>' In file included from /usr/local/include/ginac/ncmul.h:26, from /usr/local/include/ginac/structure.h:29, from /usr/local/include/ginac/ginac.h:42, from ../genericRateLaw.hpp:32, from ../ecco.hpp:41, from eccoTest.cpp:1: /usr/local/include/ginac/exprseq.h:35: error: explicit specialization not preceded by `template <>' -- http://echempp.sourceforge.net Kai Ludwig Institut für Organische Chemie Auf der Morgenstelle 18 72076 Tübingen Tel.: 07071/29-73049 Mail: kai.ludwig(a)uni-tuebingen.de

1 1

Re: Error while building GiNaC-1.2.1
by Kai Ludwig 08 Jun '04

08 Jun '04

Hi, seems that bison/flex (yacc/lex) have to be installed on the system, before GiNaC can be compiled. Nothing is said about that in the README/INSTALL files. The configure script checks it, but does not fail. Anyway, after providing them, g++-3.3 compiles GiNaC-1.2.1 fine. Cheers, Kai -------- Original Message -------- Subject: Error while building GiNaC-1.2.1 From: "Kai Ludwig" <kai.ludwig(a)uni-tuebingen.de> Date: Tue, June 8, 2004 9:42 am To: <ginac-devel(a)ginac.de> Hi All, First of all: You're doing a great job with GiNaC! I've tried to compile the GiNaC-1.2.1 version with g++-3.3.3-41 under a SuSE9.1 Linux system. Got the following: g++ -DHAVE_CONFIG_H -I. -I. -I.. -g -O2 -MT input_lexer.lo -MD -MP -MF .deps/input_lexer.Tpo -c input_lexer.cc -fPIC -DPIC -o .libs/input_lexer.oIn file included from /home/cbauer/projects/GiNaC-1.2/ginac/input_lexer.ll:35:input_lexer.h:43: error: conflicting types for `char ginac_yytext[]' input_lexer.cc:446: error: previous declaration as `char*ginac_yytext' input_lexer.h:43: error: storage size of `ginac_yytext' isn't known input_lexer.h:43: error: storage size of `ginac_yytext' isn't known make[2]: *** [input_lexer.lo] Error 1 make[2]: Leaving directory `/usr/local/src/ginac/GiNaC-1.2.1/ginac' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/usr/local/src/ginac/GiNaC-1.2.1' make: *** [all] Error 2 Cheers, Kai -- http://echempp.sourceforge.net Kai Ludwig Institut für Organische Chemie Auf der Morgenstelle 18 72076 Tübingen Tel.: 07071/29-73049 Mail: kai.ludwig(a)uni-tuebingen.de -- http://echempp.sourceforge.net Kai Ludwig Institut für Organische Chemie Auf der Morgenstelle 18 72076 Tübingen Tel.: 07071/29-73049 Mail: kai.ludwig(a)uni-tuebingen.de

1 0

[Fwd: error in cln check]
by Kai Ludwig 08 Jun '04

08 Jun '04

Hi All, I posted the message below to Richard last week. I recognized on his homepage that he could be a bit busy. Anyway, maybe it is also relevant for GiNaC users. Therefore, I decided to post it again to the ginac lists. Meanwhile, I've installed cln-1.1.5 from a rpm-package located at "http://mirror.mcs.anl.gov/suse/i386/9.1/suse/i586/" That should work. Kai -------- Original Message -------- Subject: error in cln check From: "Kai Ludwig" <kai.ludwig(a)uni-tuebingen.de> Date: Fri, June 4, 2004 4:17 pm To: <richard.kreckel(a)ginac.de> Hello Richard, I've just compiled the cln-1.1.7 with the g++-3.3.3-41 on a SuSE Linux 9.1 platform and got the following linker error: g++ -g -O2 -I../include -I../include -I. -I../src -I../src/base -I../src/float -I../src/float/lfloat -I../src/integer -I../src/base/random -I../src/base/digitseq -I../src/base/digit -I../src/base -I../src/base -S exam_I_sqrtp.cc -o exam_I_sqrtp.sgcc -g -O2 -I../include -I../include -I. -I../src -I../src/base -I../src/float -I../src/float/lfloat -I../src/integer -I../src/base/random -I../src/base/digitseq -I../src/base/digit -I../src/base -I../src/base -c exam_I_sqrtp.s -o exam_I_sqrtp.o/bin/sh ../libtool --mode=link g++ -g -O2 exam.o exam_I.o exam_RA.o exam_SF.o exam_FF.o exam_DF.o exam_LF.o exam_I_gcd.o exam_I_sqrtp.o .../src/libcln.la -lm -o exammkdir .libs g++ -g -O2 exam.o exam_I.o exam_RA.o exam_SF.o exam_FF.o exam_DF.o exam_LF.o exam_I_gcd.o exam_I_sqrtp.o -o .libs/exam .../src/.libs/libcln.so /usr/lib/libgmp.so -lm -Wl,--rpath -Wl,/usr/local/libexam.o(.text+0x42): In function `_GLOBAL__I__Z15test_elementaryv': .../include/cln/float.h:649: undefined reference to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun'exam_SF.o(.text+0x41): In function `_GLOBAL__I__Z11test_sfloatv': .../include/cln/float.h:649: undefined reference to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun'exam_FF.o(.text+0x41): In function `_GLOBAL__I__Z11test_ffloatv': .../include/cln/float.h:649: undefined reference to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun'exam_FF.o(.text+0x46): In function `_GLOBAL__I__Z11test_ffloatv': .../include/cln/ffloat.h:289: undefined reference to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun'exam_DF.o(.text+0x41): In function `_GLOBAL__I__Z11test_dfloatv': .../include/cln/float.h:649: undefined reference to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun'exam_DF.o(.text+0x46):../include/cln/dfloat.h:290:more undefined references to `_GLOBAL__I_cl_module__cl_ieee__firstglobalfun' follow../src/.libs/libcln.so: undefined reference to `_GLOBAL__I_cl_module__cl_SV_ringelt__firstglobalfun'../src/.libs/libcln.so: undefined reference to `_GLOBAL__I_cl_module__cl_GV_number__firstglobalfun'../src/.libs/libcln.so: undefined reference to `_GLOBAL__I_cl_module__cl_GV_I__firstglobalfun'../src/.libs/libcln.so: undefined reference to `_GLOBAL__I_cl_module__cl_SV_number__firstglobalfun'../src/.libs/libcln.so: undefined reference to `_GLOBAL__I_cl_module__cl_symbol__firstglobalfun' Cheers Kai -- http://echempp.sourceforge.net Kai Ludwig Institut für Organische Chemie Auf der Morgenstelle 18 72076 Tübingen Tel.: 07071/29-73049 Mail: kai.ludwig(a)uni-tuebingen.de

1 0

power rules
by econrad 22 May '04

22 May '04

Here's an annoying issue I have: I'm using GiNaC as a symbolic front end for a dynamical systems tool I'm building that does alot of numerical bifurcation analysis and such (I'm using GiNaC to generate all the symbolic info I need up front, translating this into numerical prodedures, and then using LAPACK/BLAS and some new continuation algorithms on the what I've generated). Models typically have terms like "a*k^n/(p^n+k^n)" in the RHS of the ODEs, where a, k, p are real numbers and n is typically an integer. Calculation of the jacobian then has terms like -a*k^n*n*p^(n-1)/(p^n+k^n)^2. Problem is, GiNaC uses the safe assumption and calculates -a*k^n*n*p^n*(p^n+k^n)^(-2)*p^(-1), which introduces a singularity into the derivatives that isn't really there. I understand the rational behind not automatically making p^n/p^1=p^(n-1) when one doesn't know anything about n, but nevertheless, I know that I'll only be using a real n>1 (or in any case, I'll only use n that makes this a valid reduction). How can I force this power transformation to take place... I'm sure I can do some convoluted matching and replacing to get this to work, but a user should be able to give an option to "eval" or to "normal" to get this operation performed at their own risk, yes? Have I missed something? Any help would be appreciated. Emery Emery Conrad 4089 Derring Hall Virginia Tech, Blacksburg VA 540.231.5508

2 1

GiNaC 1.2.1 now available
by Christian Bauer 23 Apr '04

23 Apr '04

Hello! GiNaC 1.2.1 is now available! The changes are: - Fixed infinite recursion in atan2_evalf() and improved atan2_eval(). - Added automatic evaluations for trigonometric functions with negative arguments (e.g. sin(-2) -> -sin(2)). - Fixed a static initialization order goof-up. - Fixed various bugs in series expansion. If you had problems using GiNaC 1.2 on certain platforms because you were getting segmentation faults all the way, this release should solve the issue. You can download GiNaC 1.2.1 as always from ftp://ftpthep.physik.uni-mainz.de/pub/GiNaC Bye, Jens and Christian -- / Physics is an algorithm \/ http://www.uni-mainz.de/~bauec002/

1 0