]>
www.ginac.de Git - ginac.git/log
Christian Bauer [Thu, 24 Oct 2002 20:37:52 +0000 (20:37 +0000)]
under certain conditions, power::expand_add_2() could produce non-canonical
numeric expairs (see exam_paranoia15())
Christian Bauer [Thu, 24 Oct 2002 20:37:19 +0000 (20:37 +0000)]
added a check for the power::expand_add_2() bug
Christian Bauer [Thu, 24 Oct 2002 20:37:03 +0000 (20:37 +0000)]
documentation update
Christian Bauer [Thu, 24 Oct 2002 18:14:19 +0000 (18:14 +0000)]
input parser no longer ignores extra stuff at the end of accepted input
Christian Bauer [Thu, 17 Oct 2002 18:34:38 +0000 (18:34 +0000)]
- added print_latex() and print_csrc()
- updated the man page
Christian Bauer [Wed, 16 Oct 2002 19:06:28 +0000 (19:06 +0000)]
reverted last change: function_index_* ist not necessarily in namespace GiNaC
(only for predefined GiNaC functions)
Richard Kreckel [Tue, 1 Oct 2002 20:10:00 +0000 (20:10 +0000)]
* irem(), iquo(): throw an exception, when second argument vanishes.
Christian Bauer [Sun, 29 Sep 2002 19:39:00 +0000 (19:39 +0000)]
added missing namespace qualification in is_ex_the_function() macro
Richard Kreckel [Mon, 23 Sep 2002 19:45:56 +0000 (19:45 +0000)]
* numeric::print(): increase precision for C source output.
Richard Kreckel [Wed, 18 Sep 2002 20:01:50 +0000 (20:01 +0000)]
* Finalize version 1.0.11.
Christian Bauer [Wed, 18 Sep 2002 18:30:09 +0000 (18:30 +0000)]
documentation update
Richard Kreckel [Wed, 18 Sep 2002 17:10:52 +0000 (17:10 +0000)]
* Fix citation.
Richard Kreckel [Wed, 18 Sep 2002 17:06:37 +0000 (17:06 +0000)]
* Fix comment.
Richard Kreckel [Wed, 18 Sep 2002 17:05:59 +0000 (17:05 +0000)]
* Fix CLN output of a-2. (Chris Dams)
Christian Bauer [Sun, 15 Sep 2002 21:07:44 +0000 (21:07 +0000)]
re-enabled the assignment operator of class symbol
Christian Bauer [Thu, 5 Sep 2002 20:16:59 +0000 (20:16 +0000)]
fixed possible memory corruption in index contraction
Richard Kreckel [Mon, 2 Sep 2002 20:42:56 +0000 (20:42 +0000)]
* Fix broken C output. (Chris Dams)
Richard Kreckel [Tue, 27 Aug 2002 17:18:07 +0000 (17:18 +0000)]
* Fix CLN output of -2*a-2*b. (Chris Dams)
Christian Bauer [Wed, 7 Aug 2002 21:47:18 +0000 (21:47 +0000)]
implemented the case in pseries::derivative() where s!=var
Christian Bauer [Mon, 5 Aug 2002 18:27:18 +0000 (18:27 +0000)]
documentation update
Christian Bauer [Sat, 3 Aug 2002 19:20:22 +0000 (19:20 +0000)]
updated the autogen.sh script
Richard Kreckel [Wed, 24 Jul 2002 22:15:50 +0000 (22:15 +0000)]
* AUTHORS: Add Chris Dams as contributor of patches.
* NEWS, configure.ac, debian/changelog: Finalize for version 1.0.10.
Christian Bauer [Wed, 24 Jul 2002 20:18:33 +0000 (20:18 +0000)]
added a check for that canonicalize() bug
Christian Bauer [Wed, 24 Jul 2002 19:59:27 +0000 (19:59 +0000)]
documentation update
Christian Bauer [Mon, 22 Jul 2002 20:32:49 +0000 (20:32 +0000)]
canonicalize() failed to return zero for an antisymmetric combination of two
or more equal elements if the elements were already in their canonical order
Christian Bauer [Thu, 18 Jul 2002 21:57:29 +0000 (21:57 +0000)]
physical momenta are 4-dimensional (just to prove that we can do this now :-)
Christian Bauer [Thu, 18 Jul 2002 21:56:10 +0000 (21:56 +0000)]
metric tensors now silently replace their indices' dimensions with their
minimum instead of complaining about mismatching dimensions
Christian Bauer [Thu, 18 Jul 2002 19:24:33 +0000 (19:24 +0000)]
- The dimension of indices is now treated as a kind of "effective" dimension
with respect to contractions; this allows for a limited support of subspaces,
e.g. delta.i[2].j[2] * a.i[3] -> a.j[2] by simplify_indexed(), and
delta.i[2].i[3] -> 2. Likewise, eta~mu[3]~nu[3] * eta.mu[4].nu[4] -> 3 etc.
(The numbers in brackets are the index dimensions which are not visible
in the output.)
Symbolic ("N", "D", ...) dimensions are always assumed to be higher than
numeric ones.
- eps0123 is gone, dirac_trace() generates ordinary epsilon tensors with
4-dimensional indices.
Christian Bauer [Mon, 15 Jul 2002 14:23:56 +0000 (14:23 +0000)]
- added idx::replace_dim() and idx::minimal_dim()
- indices no longer need to be of the same dimension to be recognized as
dummy indices, but they must be comparable (to determine their minimum)
(Nein, Markus, das ist noch nicht das mit den vierdimensionalen Tensoren)
Christian Bauer [Fri, 12 Jul 2002 18:08:49 +0000 (18:08 +0000)]
the input parser now handles (simple) indexed objects; the indices to be used
have to be specified in the same list as the symbols
Christian Bauer [Fri, 12 Jul 2002 18:07:52 +0000 (18:07 +0000)]
documentation update
Christian Bauer [Mon, 8 Jul 2002 21:07:19 +0000 (21:07 +0000)]
moved some common code to the 'tensor' class
Christian Bauer [Wed, 3 Jul 2002 18:19:57 +0000 (18:19 +0000)]
powers of indexed objects are now parenthesized correctly in LaTeX output
Christian Bauer [Tue, 25 Jun 2002 21:36:23 +0000 (21:36 +0000)]
clifford and color objects are not contracted if their representation labels
are different
Richard Kreckel [Tue, 11 Jun 2002 21:47:19 +0000 (21:47 +0000)]
* Insert date and bump patchlevel.
Christian Bauer [Tue, 11 Jun 2002 19:35:18 +0000 (19:35 +0000)]
documentation update
Christian Bauer [Mon, 10 Jun 2002 21:30:50 +0000 (21:30 +0000)]
fixed a crash when constructing expressions like (dirac_ONE()+dirac_gamma5())*dirac_gamma5()*dirac_gamma5()
Christian Bauer [Mon, 27 May 2002 20:47:10 +0000 (20:47 +0000)]
implemented relational::subs()
Christian Bauer [Mon, 27 May 2002 20:47:00 +0000 (20:47 +0000)]
documentation update
Christian Bauer [Sat, 25 May 2002 18:39:43 +0000 (18:39 +0000)]
added an additional dummy index symmetrization run over sums in
simplify_indexed(), to find possible cancellations
Christian Bauer [Fri, 24 May 2002 19:55:35 +0000 (19:55 +0000)]
fixed a bug in the raising/lowering of dummy indices and extended it to work
on single indexed objects also
Christian Bauer [Thu, 23 May 2002 04:15:15 +0000 (04:15 +0000)]
simplify_indexed() raises/lowers dummy indices to canonicalize their variance
as far as possible
Christian Bauer [Mon, 13 May 2002 20:21:57 +0000 (20:21 +0000)]
mentioned the imaginary unit 'I'
Christian Bauer [Fri, 10 May 2002 19:02:41 +0000 (19:02 +0000)]
reactivated an epsilon-tensor simplification hack because of a user request
Christian Bauer [Thu, 25 Apr 2002 19:13:58 +0000 (19:13 +0000)]
it is now enforced that indices of a metric tensor are of the same dimension
Richard Kreckel [Sun, 31 Mar 2002 21:56:25 +0000 (21:56 +0000)]
* Bump timestamps and version numbers.
Richard Kreckel [Thu, 28 Mar 2002 15:44:28 +0000 (15:44 +0000)]
* mul::expand(): Fix bug in aggressive version of expansion of two sums
and enable it.
Richard Kreckel [Thu, 28 Mar 2002 15:38:29 +0000 (15:38 +0000)]
* Clean up the output of all timings to stderr.
Richard Kreckel [Thu, 28 Mar 2002 01:14:02 +0000 (01:14 +0000)]
* mul::expand() (mul.cpp): When multiplying two sums, be careful not to
allocate too much.
Richard Kreckel [Thu, 28 Mar 2002 00:27:14 +0000 (00:27 +0000)]
* expairseq::combine_same_terms_sorted_seq() (expairseq.cpp): make this
ugh, ohhh, ahh, more... legible. ;-)
Richard Kreckel [Thu, 28 Mar 2002 00:16:26 +0000 (00:16 +0000)]
* Bump up version, keep INTERFACE_AGE at zero.
(Yeah sure, it's early. But this way we don't forget it.)
Richard Kreckel [Thu, 28 Mar 2002 00:13:25 +0000 (00:13 +0000)]
* (time_fateman_expand.cpp): new file.
* (Makefile.am): compile it.
* (times.h, times.cpp): declare and call it.
Richard Kreckel [Thu, 28 Mar 2002 00:02:51 +0000 (00:02 +0000)]
* expairseq::expairseq(const epvector&, const ex& oc) (expairseq.cpp):
check assertion: GINAC_ASSERT(is_a<numeric>(oc)).
* expairseq::expairseq(epvector*, const ex& oc) (expairseq.cpp): Likewise.
* expair_rest_is_less (expair.h): New functor...
* expairseq::canonicalize() (expairseq.cpp): ...used here.
Richard Kreckel [Wed, 27 Mar 2002 23:41:42 +0000 (23:41 +0000)]
* power::expand_add(): allocate the precise amount needed for representing
the result. And some minor readability cleanups...
Richard Kreckel [Wed, 27 Mar 2002 23:17:25 +0000 (23:17 +0000)]
* Don't push void to static Bernoulli vector, it breaks with GCC-3.1.
Christian Bauer [Wed, 27 Mar 2002 21:24:32 +0000 (21:24 +0000)]
inlined ex::swap() and provided versions of iter_swap() for ex and expair
objects to make the STL sorting algorithms a little more efficient
Richard Kreckel [Sun, 24 Mar 2002 19:11:12 +0000 (19:11 +0000)]
* Add namespace declarators to is_ex_the_function().
Richard Kreckel [Mon, 18 Mar 2002 22:02:34 +0000 (22:02 +0000)]
* Update dates, touch, and all this...
Richard Kreckel [Fri, 8 Mar 2002 20:35:49 +0000 (20:35 +0000)]
* Bump version (early, but I don't want to forget to set INTERFACE_AGE
to zero.)
Richard Kreckel [Fri, 8 Mar 2002 20:32:59 +0000 (20:32 +0000)]
* matrix:;pow(): Fix exponent==0 case.
* matrix::print(): Implement print_latex.
* Add unit_matrix(unsigend) and unit_matrix(unsigned,unsigned).
Christian Bauer [Wed, 6 Mar 2002 18:01:58 +0000 (18:01 +0000)]
updated the description of degree()/ldegree()
Christian Bauer [Tue, 5 Mar 2002 20:32:16 +0000 (20:32 +0000)]
fixed LaTeX output of indexed objects
Christian Bauer [Tue, 5 Mar 2002 20:32:02 +0000 (20:32 +0000)]
documentation update
Richard Kreckel [Mon, 4 Mar 2002 22:09:29 +0000 (22:09 +0000)]
* Insert date.
Richard Kreckel [Mon, 4 Mar 2002 17:13:29 +0000 (17:13 +0000)]
Prepare for release 1.0.6:
* ginac.m4: quote AM_PATH_GINAC (recommended by Roberto Bagnara).
* configure.ac, debian/changelog: bump up version.
* doc/tutorial/ginac.tex: touch?
Christian Bauer [Thu, 28 Feb 2002 17:53:47 +0000 (17:53 +0000)]
"(x+1).subs(x==x-1)" now returns the correct result "x" instead of "x-1"
Christian Bauer [Thu, 28 Feb 2002 17:53:40 +0000 (17:53 +0000)]
fixed typos in examples
Christian Bauer [Thu, 28 Feb 2002 17:53:32 +0000 (17:53 +0000)]
added some checks for subs() behavior
Christian Bauer [Thu, 28 Feb 2002 17:53:20 +0000 (17:53 +0000)]
documentation update
Christian Bauer [Tue, 29 Jan 2002 22:46:20 +0000 (22:46 +0000)]
added check for consistency of numeric input/output routines
Richard Kreckel [Sun, 27 Jan 2002 20:59:12 +0000 (20:59 +0000)]
* Bump up versions to 1.0.5, adjust dates.
Richard Kreckel [Sat, 26 Jan 2002 16:29:52 +0000 (16:29 +0000)]
* Add Roberto Bagnara.
Richard Kreckel [Sat, 26 Jan 2002 16:29:42 +0000 (16:29 +0000)]
* numeric::print(): Treat imaginary numbers correctly.
Christian Bauer [Fri, 25 Jan 2002 18:33:01 +0000 (18:33 +0000)]
- (l)degree(s), coeff(s, n) and collect(s) were extended to accept expressions
of any class (except add/mul/ncmul/numeric) for "s". They should even work
if "s" is a "power" object, as long as the exponent is non-integer, but with
some limitations. For example, you can "collect(a*2^x+b*2^x, 2^x)" to get
"(a+b)*2^x", but "degree(2^(3*x), 2^x)" yields 0 instead of 3).
Richard Kreckel [Fri, 25 Jan 2002 01:28:01 +0000 (01:28 +0000)]
* Finilize version 1.0.4 (version numbers, copyrights and such rubbish).
Richard Kreckel [Fri, 25 Jan 2002 00:03:44 +0000 (00:03 +0000)]
* mul::expand(): Considerable speedup through caching in nested loop.
Richard Kreckel [Tue, 22 Jan 2002 17:01:10 +0000 (17:01 +0000)]
* bernoulli(): Add Markus Nullmeier's new implementation.
* numeric::numeric({int,unsigned}): Really construct CLN immediates.
Richard Kreckel [Sat, 19 Jan 2002 19:12:47 +0000 (19:12 +0000)]
* Enable EXTRACT_ALL for LaTeX.
Richard Kreckel [Sat, 19 Jan 2002 19:12:24 +0000 (19:12 +0000)]
* Docu: elaborate definition of sqrfree() (by Roberto Bagnara)
Richard Kreckel [Fri, 18 Jan 2002 20:06:01 +0000 (20:06 +0000)]
* Updated.
Richard Kreckel [Fri, 18 Jan 2002 19:59:19 +0000 (19:59 +0000)]
* Updated.
Richard Kreckel [Fri, 18 Jan 2002 19:59:09 +0000 (19:59 +0000)]
* Faster Bernoulli numbers (Markus Nullmeier).
Richard Kreckel [Tue, 15 Jan 2002 14:46:17 +0000 (14:46 +0000)]
* GINAC_CHECK_ERRORS: Don't delete cache_file /dev/null (by Roberto Bagnara).
Richard Kreckel [Fri, 11 Jan 2002 18:48:39 +0000 (18:48 +0000)]
* Document the rounding behavior as dicussed on ginac-list.
Richard Kreckel [Fri, 21 Dec 2001 21:06:52 +0000 (21:06 +0000)]
* Adjust for 1.0.3.
Richard Kreckel [Fri, 21 Dec 2001 20:32:23 +0000 (20:32 +0000)]
* Fix thinkos in print_csrc context.
(Cebix: read the implementations of ex::compare() and basic::compare()!!!)
Christian Bauer [Thu, 20 Dec 2001 14:24:45 +0000 (14:24 +0000)]
fixed a bug where quo() would call vector::reserve() with a negative argument
Christian Bauer [Thu, 20 Dec 2001 14:24:34 +0000 (14:24 +0000)]
added check for quo() bug
Christian Bauer [Thu, 20 Dec 2001 14:24:27 +0000 (14:24 +0000)]
documentation update
Richard Kreckel [Wed, 19 Dec 2001 22:06:45 +0000 (22:06 +0000)]
* Packaging update.
Richard Kreckel [Wed, 19 Dec 2001 22:06:24 +0000 (22:06 +0000)]
* Remove check for empty seq in Python output.
Richard Kreckel [Wed, 19 Dec 2001 19:17:26 +0000 (19:17 +0000)]
* Old and future changes: relativity at work...
Richard Kreckel [Wed, 19 Dec 2001 19:16:48 +0000 (19:16 +0000)]
* Huh, huh...
Richard Kreckel [Wed, 19 Dec 2001 19:16:16 +0000 (19:16 +0000)]
* Added output-support for Python bindings and LaTeX printing for
indexed objects (Pearu Peterson).
* relational.*: Removed superfluous is_equal_same_type()...
* inifcns_trans.cpp: ...added extern declaration as workaround for
compiler-bug triggered by changes to relational.*,
* input_parser.yy: added missing namespace specification.
Richard Kreckel [Tue, 18 Dec 2001 19:03:48 +0000 (19:03 +0000)]
* Added some long-pending docu.
Christian Bauer [Tue, 18 Dec 2001 18:48:06 +0000 (18:48 +0000)]
- input parser recognizes "sqrt()", which is also used in the output
- divide(a,b,q) only modifies q if the division succeeds; also, divide(a,b,a)
works now
- fixed small bug in dummy index renaming which could cause it to not
recognize renamable indices in some cases
- power::degree() and power::ldegree() throw an exception when encountering
a non-integer exponent
Richard Kreckel [Thu, 22 Nov 2001 19:33:58 +0000 (19:33 +0000)]
* Bump version number.
Richard Kreckel [Thu, 22 Nov 2001 19:32:56 +0000 (19:32 +0000)]
* Add Pearu's patch canonicalizing class relational, augmented by missing
logic in compare_same_type().
Richard Kreckel [Thu, 22 Nov 2001 19:31:34 +0000 (19:31 +0000)]
* Minor cleanups to avoid copying of symbols and stuff like that.
Christian Bauer [Tue, 20 Nov 2001 20:42:05 +0000 (20:42 +0000)]
readline versions like "4.735beta5-pl2" should now be correctly recognized