1 This file records noteworthy changes.
4 * Added a structure<T> template class for the easy creation of user-defined
6 * Improved the speed of subs() and normal() by the use of maps instead of
8 * Many internal reorganizations and optimizations.
9 * Caveats for class implementors:
10 - basic::copy() and basic::destroy() are gone; classes derived from
11 basic can use the defaults for the assignment operator and copy
13 - basic::subs() and basic::normal() take an 'exmap &' instead of two
15 - basic::subs() now descends into subexpressions (if accessible via
16 nops()/op()/let_op()). If you have a custom implementation of subs()
17 that calls basic::subs() after substituting subexpressions, this needs
18 to be changed to a call to subs_one_level().
21 * Removed deprecated macros is_ex_a, is_ex_exactly_a and friends for good.
22 * The scalar_products mechanism allows the specification of an index dimension.
23 * Removed dirac_gamma6/7().
24 * Added ex::to_polynomial().
25 * subs() accepts an optional "options" argument. The option
26 subs_option::subs_algebraic enables "smart" substitutions in products and
28 * Added stream manipulators "dflt", "latex", "python", "python_repr", "tree",
29 "csrc", "csrc_float", "csrc_double", "csrc_cl_N", "index_dimensions" and
30 "no_index_dimensions" to control the output format. Calling basic::print()
31 directly is now deprecated.
32 * Made the hashing more simple and efficient.
33 * Caveats for class implementors:
34 - basic::subs(): third argument changed from "bool" to "unsigned"
35 - unarchiving constructor and basic::unarchive(): "const" removed from
37 - basic::let_op() should only be implemented if write access to
38 subexpressions is desired
39 - simplify_ncmul() renamed to eval_ncmul()
40 - simplified_ncmul() renamed to hold_ncmul()
41 - nonsimplified_ncmul() renamed to reeval_ncmul()
44 * Improved the C-source output: complex numbers are printed correctly (using
45 the STL complex<> template or cln::complex()), rational numbers use cl_RA()
46 in the CLN output, and small integers are printed in a more compact format
47 (e.g. "2.0" instead of "2.0000000e+00").
48 * function_options::set_return_type() and function_options::do_not_evalf_params()
51 1.0.13 (27 January 2003)
52 * Contracting epsilon tensors with Euclidean indices now works.
53 * Improved dummy index symmetrization in sums.
54 * Added dirac_gammaL/R(), which can be used instead of dirac_gamma6/7()
55 but are single objects, to allow for a more compact notation of Dirac
57 * Powers with negative numeric exponents are printed as fractions in the
59 * Added symbolic_matrix() for the convenient creation of matrices filled
61 * Added collect_common_factors() which collects common factors from the
63 * simplify_indexed() converts "gamma~mu*p.mu" to "p\".
65 1.0.12 (30 October 2002)
66 * Fixed a bug in power::expand() that could produce invalid expressions.
67 * The input parser no longer ignores extra data following accepted input.
68 * Improved the CLN C-source output (integers are printed as integers, and
69 floating point numbers include the precision).
70 * Fixed a problem in the LaTeX-output of negative fractions.
71 * Added print_latex() and print_csrc() to ginsh.
72 * The sprem() function is now public.
74 1.0.11 (18 September 2002)
75 * Fixed a possible memory corruption in contractions of indexed objects with
76 delta or metric tensors.
77 * Computing the derivative of a power series object with respect to a symbol
78 that is not the expansion variable now works correctly.
79 * Several bugfixes in code generation.
82 * Powers of indexed objects are now parenthesized correctly in LaTeX output.
83 * Input parser handles indices (they have to be specified in the same list
85 * Added some limited support for subspaces in the idx and tensor classes.
86 * Fixed a bug in canonicalize() (antisymmetric canonicalization of an already
87 sorted list containing two or more equal objects failed to return 0).
90 * simplify_indexed() now raises/lowers dummy indices to canonicalize the index
91 variance. This allows some simplifications that weren't possible before,
92 like eps~a.b~c~d*X.a*X~b -> 0 and X.a~a-X~a.a -> 0.
93 * Implemented relational::subs().
94 * Fixed bug in simplify_ncmul() for clifford objects.
97 * Improvements in memory usage of the expand() methods.
100 * Fixed LaTeX output of indexed and matrix objects.
101 * Fixed matrix::pow(n) for n==0 and added helper functions to create unit
102 matrices "ex unit_matrix(unsigned, unsigned)".
105 * "(x+1).subs(x==x-1)" now returns the correct result "x" instead of "x-1".
107 1.0.5 (27 January 2002)
108 * (l)degree(s), coeff(s, n) and collect(s) were extended to accept expressions
109 of any class (except add/mul/ncmul/numeric) for "s". They should even work
110 if "s" is a "power" object, as long as the exponent is non-integer, but with
111 some limitations. For example, you can "collect(a*2^x+b*2^x, 2^x)" to get
112 "(a+b)*2^x", but "degree(2^(3*x), 2^x)" yields 0 instead of 3).
113 * Fixed a small output bug.
115 1.0.4 (24 January 2002)
116 * Speedup in expand().
117 * Faster Bernoulli numbers (Markus Nullmeier).
118 * Some minor bugfixes and documentation updates.
120 1.0.3 (21 December 2001)
121 * Fixed a bug where quo() would call vector::reserve() with a negative
123 * Fix several bugs in code generation.
125 1.0.2 (19 December 2001)
126 * Input parser recognizes "sqrt()", which is also used in the output.
127 * divide(a,b,q) only modifies q if the division succeeds; also, divide(a,b,a)
129 * Fixed small bug in dummy index renaming which could cause it to not
130 recognize renamable indices in some cases.
131 * power::degree() and power::ldegree() throw an exception when encountering
132 a non-integer exponent.
133 * Add output-support for Python bindings.
135 1.0.1 (22 November 2001)
136 * Function sqrfree() handles a few more cases now.
137 * Class relational has real canonical ordering now.
138 * Handle obscene libreadline version numbers when building ginsh.
140 1.0.0 (6 November 2001)
141 * Some internal reorganization resulting in a general speed-up.
142 * The last 3 evaluated expressions in ginsh are now referred to with the
143 tokens '%', '%%' and '%%%'. The old '"', '""' and '"""' remain for
144 compatibility but may be removed in a future version of GiNaC.
146 0.9.4 (20 September 2001)
147 * Functions have better support for external scripting languages.
148 * Interface cleanups and bugfixes.
149 * Fix silly bug in evalf() that prevented things like 2^Pi being computed.
151 0.9.3 (16 August 2001)
152 * series expansion now much more consistent for small order expansion.
153 * lsolve() accepts algorithmic hint as parameter.
156 * Epsilon tensor is more functional.
157 * simplify_indexed() is better at detecting expressions that vanish for
159 * Several little bugfixes and consistency enhancements.
162 * Ctors of class numeric are not explicit any more. All built-in callers for
163 pseudofunctions are now templated and default to ex arguments which relaxes
164 the need for explicit ctors.
165 * New functions/methods:
167 - remove_first(), remove_last(), sort() and unique() for lists
168 - symmetrize_cyclic()
170 * Instead of just totally symmetric or antisymmetric, complex symmetries
171 can now be defined for indexed objects. Symmetries are described by a
172 tree of "symmetry" objects that is constructed with the sy_none(),
173 sy_symm(), sy_anti() and sy_cycl() functions. The symmetry of a function
174 with respect to its arguments can also be defined (this is currently
175 only used for the Beta function).
176 * Generalized map() to take a function object instead of a function pointer.
177 This allows passing an arbitrary number of additional state to the
178 function being called.
179 * color_trace(), dirac_trace(), diff(), expand(), evalf() and normal() work
180 better with container classes, e.g. using color_trace() on a relation will
181 take the trace on both sides, using diff() on a matrix differentiates every
183 * diff() works properly with non-commutative products and indexed objects.
184 * New option flag "expand_function_args" for expand().
185 * Supplement some (now deprecated) macros by inlined template functions:
186 - is_of_type(foo, type) -> is_a<type>(foo)
187 - is_ex_of_type(foo, type) -> is_a<type>(foo)
188 - is_exaclty_of_type(foo, type) -> is_exaclty_a<type>(foo)
189 - is_ex_exaclty_of_type(foo, type) -> is_exaclty_a<type>(foo)
190 - ex_to_foobar(baz) -> ex_to<foobar>(baz)
191 * rem(c, p[x], x) (c: numeric, p[x]: polynomial) erroneously returned p[x]
193 * Small bugfixes in pattern matching.
194 * Updated libtool to version 1.4.
197 * In the output and in ginsh, lists are now delimited by { } braces, and
198 matrices are delimited by single [ ] brackets.
199 * simplify_indexed() renames dummy indices so, e.g., "a.i*a.i+a.j*a.j" gets
200 simplified to "2*a.i*a.i" (or "2*a.j*a.j").
201 * New functions/methods:
202 - canonicalize_clifford() (helpful when comparing expressions containing
204 - symmetrize() and antisymmetrize()
205 - numer_denom() (return numerator and denominator in one call)
206 - map() (apply function to subexpressions)
207 - evalm() (evaluate sums, products and integer powers of matrices)
208 * Added a new function match() for performing pattern matching. subs() and
209 has() also accept patterns as arguments. A pattern can be any expression,
210 optionally containing wildcard objects. These are constructed with the
211 call "wild(<unsigned>)" and are denoted as "$0", "$1" etc. in the output
213 * Positive integer powers of non-commutative expressions (except matrices)
214 are automatically expanded.
215 * Removed cint subdirectory, ginaccint is a separate package now due to
216 packaging considerations.
217 * Several little bugfixes.
220 * color and clifford classes are functional and documented.
221 * New "spinidx" class for dotted/undotted indices.
222 * Predefined spinor metric tensor (created by spinor_metric()).
223 * Symbols can have a LaTeX name, e.g. symbol s("s", "\\sigma");
224 * LaTeX output of indexed objects is much nicer.
225 * Fixed some build problems (with recent libreadline).
226 * Semantics of arithmetic operators now follows the C++ rules more strictly.
228 0.8.2 (24 April 2001)
229 * degree(), ldegree(), coeff(), lcoeff(), tcoeff() and collect() work with
230 non-symbols as the second argument in ginsh.
231 * the argument to collect() can be a list of objects in which case the
232 result is either a recursively collected polynomial, or a polynomial in
233 a distributed form with terms like coeff*x1^e1*...*xn^en, as specified by
234 the second argument to collect().
235 * Several bugfixes (including a nasty memory leak in .normal()).
236 * class matrix: solve() doesn't call algorithms redundantly any more and
237 inverse() falls back to solve() which works in more general cases.
239 0.8.1 (16 April 2001)
240 * degree(), ldegree(), coeff(), lcoeff(), tcoeff() and collect() can now
241 be used with constants, functions and indexed expressions as well, so you
242 can use it to collect by powers of Pi or sin(x), or to find the coefficient
245 - it only works with symbols, constants, functions and indexed expressions,
246 trying to find the coefficient of, e.g., "x^2" or "x+y" won't work;
247 - it does not know about dummy index summations; the coefficient of
248 gamma~0 in p.mu*gamma~mu should be p.0 but is returned as 0;
249 - using coeff(), tcoeff(), lcoeff() or collect() on elements of
250 noncommutative products might return wrong or surprising results.
251 * subs() no longer only substitutes symbols and indices but performs a more
252 general "syntactic substitution", i.e. it substitutes whole objects in sub-
253 expressions. You can subs((a+b)^2,a+b==3) and get 9, but subs(a+b+c,a+b==3)
256 - substituting numerics (subs(expr, 2==4)) will not replace then in all
257 occurences; in general, you shouldn't substitute numerics, though.
258 * Added preliminary (re)implementations of color and clifford classes.
259 * simplify_indexed(): contraction of symmetric and antisymmetric tensors
261 * Replaced the various print*() member functions by a single print() that
262 takes a print_context object that determines the output formatting. This
263 should make it easier to add more output types, such as LaTeX output,
264 which is based on work by Stefan Weinzierl.
265 * Added functions to retrieve the properties stored in archive objects
266 outside of unarchive() (for printing or debugging purposes).
267 * Some bugfixes (indexed objects, archive writing).
268 * .collect() on non-polynomials is now algebraically correct.
270 0.8.0 (24 March 2001)
271 * Complete revamp of indexed objects. Instead of multiple classes for
272 indexed things and their indices there is now only one "indexed" class
273 and two types of indices: "idx" for simple indices and "varidx" for
274 indices with variance. There are predefined delta, epsilon and metric
275 tensors, and a function simplify_indexed() that performs canonicalization
276 and dummy index summations. Matrix objects can be indexed for doing simple
278 * Added an option "expand_indexed" to expand() to perform expansion of
279 indexed objects like (a+b).i -> a.i + b.i
280 * Renamed get_indices() to get_free_indices(), which no longer returns
281 dummy indices and checks the consistency of indices in sums.
282 * sqrfree() factorization fixed and improved syntactically.
283 * subs() works on matrices.
284 * Matrices can be constructed from flat list of elements; diagonal matrices
285 can be constructed from list of diagonal elements with diag_matrix().
286 * Fixed memory leak in expand().
287 * Operator% for objects of class ncmul has gone. Use operator* now for that
288 case too, which is much more natural.
290 0.7.3 (28 February 2001)
291 * Several bugfixes and minor performance tunings.
292 * Added a section to the tutorial about adding new algebraic classes to GiNaC.
293 * Closed many in-source documentation gaps.
295 0.7.2 (17 February 2001)
296 * Several bugfixes in power series expansion, one of them critical.
298 0.7.1 (7 February 2001)
299 * Fix problems with Cint that were caused by CLN's overloaded operator new.
300 * Fix compilation errors with GCC3.
301 * normal() handles large sums of fractions better and normalizes the exponent
302 of power expressions.
303 * expand() always expands the exponent and transforms x^(a+b) -> x^a*x^b.
304 * Some bugfixes of series expansion around branch cuts of special functions.
306 0.7.0 (15 December 2000)
307 * Requires CLN 1.1 now. Class numeric doesn't use an indirect pointer to the
308 actual representation any more. This is a speedup.
309 * mul::expand() was reengineered to not allocate excess temporary memory.
310 * Non-integer powers of a symbol are treated as constants by (l)degree() and
311 coeff(). Using these functions on an expression containing such powers used
312 to fail with an internal error message. The side-effect is that collect()
313 can be used on expressions which are not polynomials.
314 * Added a man page for the ginac-config script.
315 * Ctor of numeric from char* honors Digits.
317 0.6.4 (10 August 2000)
318 * Complete revamp of methods in class matrix. Some redundant (and poor)
319 implementations of elimination schemes were thrown out. The code is now
320 highly orthogonal, more flexible and much more efficient.
321 * Some long standing and quite nasty bugs were discovered and fixed in the
322 following functions: add::normal(), heur_gcd(), sr_gcd() and Order_eval().
325 * Derivatives are now assembled in a slightly different manner (i.e. they
326 might 'look' different on first sight). Under certain circumstances this
327 can result in a dramatic speedup because it gives hashing a better chance,
328 especially when computing higher derivatives.
329 * Some series expansions of built-in functions have been reengineered.
330 * The algorithm for computing determinants can be chosen by the user. See
331 ginac/flags.h and ginac/matrix.h.
332 * The Dilogarithm (Li2) now has floating point evaluation, derivative and a
333 proper series expansion.
334 * Namespace 'std' cleanly disentangled, as demanded by ISO/EIC 14882-1998(E).
335 * Some minor bugfixes, one major lsolve()-bugfix and documentation updates.
338 * ginaccint.bin is now launched by a binary program instead of by a scripts.
339 This allows us to write #!-scripts. A small test suite for GiNaC-cint was
341 * Several minor bugfixes.
344 * Cleanup in the interface to Cint. The required version is now Cint 5.14.38.
345 * Several bugfixes in target install.
348 * IMPORTANT: Several interface changes make programs written with GiNaC
349 much clearer but break compatibility with older versions:
350 - f(x).series(x,p[,o]) -> f(x).series(x==p,o)
351 - series(f(x),x,p[,o]) -> series(f(x),x==p,o)
352 - gamma() -> tgamma() (The true Gamma function, there is now also
353 log(tgamma()), called lgamma(), in accord with ISO/IEC 9899:1999.)
354 - EulerGamma -> Euler
355 * #include'ing ginac.h defines the preprocessor symbols GINACLIB_MAJOR_VERSION,
356 GINACLIB_MINOR_VERSION, and GINACLIB_MICRO_VERSION with the respective GiNaC
357 library version numbers.
358 * Expressions can be constructed from strings like this:
359 ex e("2*x+y", lst(x, y));
360 * ex::to_rational() provides a way to extend the domain of functions like
361 gcd() and divide() that only work on polynomials or rational functions (the
362 good old ex::subs() method reverses this process)
363 * Calling diff() on a function that has no derivative defined returns the
364 inert derivative function "Derivative".
365 * Several new timings in the check target. Some of them may be rather rude
366 at your machine, feel free to interrupt them.
368 0.5.4 (15 March 2000)
369 * Some algorithms in class matrix (notably determinant) were replaced by
370 less brain-dead ones and should now have much better performance.
371 * Checks were completely reorganized and split up into three parts:
372 a) exams (small regression tests with predefined input)
373 b) checks (lenghty coherence checks with random input)
374 c) timings (for coherence and crude benchmarking)
375 * Behaviour of .evalf() was changed: it doesn't .evalf() any exponents.
376 * Expanded expressions now remember they are expanded to prevent
377 superfluous expansions.
378 * Small bugfixes and improvements in the series expansion.
380 0.5.3 (23 February 2000)
381 * A more flexible scheme for registering functions was implemented,
382 allowing for remembering, too.
385 0.5.2 (16 February 2000)
386 * Mainly fixes a bug in the packaging of release 0.5.1.
388 0.5.1 (14 February 2000)
389 * Fixes a small number of bugs.
391 0.5.0 (7 February 2000)
392 * Expressions can be written ("archived") to files and read therefrom.
393 * Addition of GiNaC-cint, which lets you write complete programs in
394 an interactive shell-like manner in your favoured programming
397 0.4.1 (13 December 1999)
398 * Series Expansion of Gamma function and some other trigonometric
399 functions at their poles works now.
400 * Many more evaluations of special functions at points where
402 * info_flags::rational doesn't return true for complex extensions
403 any more---use info_flags::crational for the old behaviour.
404 info_flags::integer and -::cinteger work similarly, the same
405 holds for types like info_flags::rational_polynomial.
407 0.4.0 (26 November 1999)
408 * First public release.