1 This file records noteworthy changes.
4 * Improved the C-source output: complex numbers are printed correctly (using
5 the STL complex<> template or cln::complex()), rational numbers use cl_RA()
6 in the CLN output, and small integers are printed in a more compact format
7 (e.g. "2.0" instead of "2.0000000e+00").
9 1.0.13 (27 January 2003)
10 * Contracting epsilon tensors with Euclidean indices now works.
11 * Improved dummy index symmetrization in sums.
12 * Added dirac_gammaL/R(), which can be used instead of dirac_gamma6/7()
13 but are single objects, to allow for a more compact notation of Dirac
15 * Powers with negative numeric exponents are printed as fractions in the
17 * Added symbolic_matrix() for the convenient creation of matrices filled
19 * Added collect_common_factors() which collects common factors from the
21 * simplify_indexed() converts "gamma~mu*p.mu" to "p\".
23 1.0.12 (30 October 2002)
24 * Fixed a bug in power::expand() that could produce invalid expressions.
25 * The input parser no longer ignores extra data following accepted input.
26 * Improved the CLN C-source output (integers are printed as integers, and
27 floating point numbers include the precision).
28 * Fixed a problem in the LaTeX-output of negative fractions.
29 * Added print_latex() and print_csrc() to ginsh.
30 * The sprem() function is now public.
32 1.0.11 (18 September 2002)
33 * Fixed a possible memory corruption in contractions of indexed objects with
34 delta or metric tensors.
35 * Computing the derivative of a power series object with respect to a symbol
36 that is not the expansion variable now works correctly.
37 * Several bugfixes in code generation.
40 * Powers of indexed objects are now parenthesized correctly in LaTeX output.
41 * Input parser handles indices (they have to be specified in the same list
43 * Added some limited support for subspaces in the idx and tensor classes.
44 * Fixed a bug in canonicalize() (antisymmetric canonicalization of an already
45 sorted list containing two or more equal objects failed to return 0).
48 * simplify_indexed() now raises/lowers dummy indices to canonicalize the index
49 variance. This allows some simplifications that weren't possible before,
50 like eps~a.b~c~d*X.a*X~b -> 0 and X.a~a-X~a.a -> 0.
51 * Implemented relational::subs().
52 * Fixed bug in simplify_ncmul() for clifford objects.
55 * Improvements in memory usage of the expand() methods.
58 * Fixed LaTeX output of indexed and matrix objects.
59 * Fixed matrix::pow(n) for n==0 and added helper functions to create unit
60 matrices "ex unit_matrix(unsigned, unsigned)".
63 * "(x+1).subs(x==x-1)" now returns the correct result "x" instead of "x-1".
65 1.0.5 (27 January 2002)
66 * (l)degree(s), coeff(s, n) and collect(s) were extended to accept expressions
67 of any class (except add/mul/ncmul/numeric) for "s". They should even work
68 if "s" is a "power" object, as long as the exponent is non-integer, but with
69 some limitations. For example, you can "collect(a*2^x+b*2^x, 2^x)" to get
70 "(a+b)*2^x", but "degree(2^(3*x), 2^x)" yields 0 instead of 3).
71 * Fixed a small output bug.
73 1.0.4 (24 January 2002)
74 * Speedup in expand().
75 * Faster Bernoulli numbers (Markus Nullmeier).
76 * Some minor bugfixes and documentation updates.
78 1.0.3 (21 December 2001)
79 * Fixed a bug where quo() would call vector::reserve() with a negative
81 * Fix several bugs in code generation.
83 1.0.2 (19 December 2001)
84 * Input parser recognizes "sqrt()", which is also used in the output.
85 * divide(a,b,q) only modifies q if the division succeeds; also, divide(a,b,a)
87 * Fixed small bug in dummy index renaming which could cause it to not
88 recognize renamable indices in some cases.
89 * power::degree() and power::ldegree() throw an exception when encountering
90 a non-integer exponent.
91 * Add output-support for Python bindings.
93 1.0.1 (22 November 2001)
94 * Function sqrfree() handles a few more cases now.
95 * Class relational has real canonical ordering now.
96 * Handle obscene libreadline version numbers when building ginsh.
98 1.0.0 (6 November 2001)
99 * Some internal reorganization resulting in a general speed-up.
100 * The last 3 evaluated expressions in ginsh are now referred to with the
101 tokens '%', '%%' and '%%%'. The old '"', '""' and '"""' remain for
102 compatibility but may be removed in a future version of GiNaC.
104 0.9.4 (20 September 2001)
105 * Functions have better support for external scripting languages.
106 * Interface cleanups and bugfixes.
107 * Fix silly bug in evalf() that prevented things like 2^Pi being computed.
109 0.9.3 (16 August 2001)
110 * series expansion now much more consistent for small order expansion.
111 * lsolve() accepts algorithmic hint as parameter.
114 * Epsilon tensor is more functional.
115 * simplify_indexed() is better at detecting expressions that vanish for
117 * Several little bugfixes and consistency enhancements.
120 * Ctors of class numeric are not explicit any more. All built-in callers for
121 pseudofunctions are now templated and default to ex arguments which relaxes
122 the need for explicit ctors.
123 * New functions/methods:
125 - remove_first(), remove_last(), sort() and unique() for lists
126 - symmetrize_cyclic()
128 * Instead of just totally symmetric or antisymmetric, complex symmetries
129 can now be defined for indexed objects. Symmetries are described by a
130 tree of "symmetry" objects that is constructed with the sy_none(),
131 sy_symm(), sy_anti() and sy_cycl() functions. The symmetry of a function
132 with respect to its arguments can also be defined (this is currently
133 only used for the Beta function).
134 * Generalized map() to take a function object instead of a function pointer.
135 This allows passing an arbitrary number of additional state to the
136 function being called.
137 * color_trace(), dirac_trace(), diff(), expand(), evalf() and normal() work
138 better with container classes, e.g. using color_trace() on a relation will
139 take the trace on both sides, using diff() on a matrix differentiates every
141 * diff() works properly with non-commutative products and indexed objects.
142 * New option flag "expand_function_args" for expand().
143 * Supplement some (now deprecated) macros by inlined template functions:
144 - is_of_type(foo, type) -> is_a<type>(foo)
145 - is_ex_of_type(foo, type) -> is_a<type>(foo)
146 - is_exaclty_of_type(foo, type) -> is_exaclty_a<type>(foo)
147 - is_ex_exaclty_of_type(foo, type) -> is_exaclty_a<type>(foo)
148 - ex_to_foobar(baz) -> ex_to<foobar>(baz)
149 * rem(c, p[x], x) (c: numeric, p[x]: polynomial) erroneously returned p[x]
151 * Small bugfixes in pattern matching.
152 * Updated libtool to version 1.4.
155 * In the output and in ginsh, lists are now delimited by { } braces, and
156 matrices are delimited by single [ ] brackets.
157 * simplify_indexed() renames dummy indices so, e.g., "a.i*a.i+a.j*a.j" gets
158 simplified to "2*a.i*a.i" (or "2*a.j*a.j").
159 * New functions/methods:
160 - canonicalize_clifford() (helpful when comparing expressions containing
162 - symmetrize() and antisymmetrize()
163 - numer_denom() (return numerator and denominator in one call)
164 - map() (apply function to subexpressions)
165 - evalm() (evaluate sums, products and integer powers of matrices)
166 * Added a new function match() for performing pattern matching. subs() and
167 has() also accept patterns as arguments. A pattern can be any expression,
168 optionally containing wildcard objects. These are constructed with the
169 call "wild(<unsigned>)" and are denoted as "$0", "$1" etc. in the output
171 * Positive integer powers of non-commutative expressions (except matrices)
172 are automatically expanded.
173 * Removed cint subdirectory, ginaccint is a separate package now due to
174 packaging considerations.
175 * Several little bugfixes.
178 * color and clifford classes are functional and documented.
179 * New "spinidx" class for dotted/undotted indices.
180 * Predefined spinor metric tensor (created by spinor_metric()).
181 * Symbols can have a LaTeX name, e.g. symbol s("s", "\\sigma");
182 * LaTeX output of indexed objects is much nicer.
183 * Fixed some build problems (with recent libreadline).
184 * Semantics of arithmetic operators now follows the C++ rules more strictly.
186 0.8.2 (24 April 2001)
187 * degree(), ldegree(), coeff(), lcoeff(), tcoeff() and collect() work with
188 non-symbols as the second argument in ginsh.
189 * the argument to collect() can be a list of objects in which case the
190 result is either a recursively collected polynomial, or a polynomial in
191 a distributed form with terms like coeff*x1^e1*...*xn^en, as specified by
192 the second argument to collect().
193 * Several bugfixes (including a nasty memory leak in .normal()).
194 * class matrix: solve() doesn't call algorithms redundantly any more and
195 inverse() falls back to solve() which works in more general cases.
197 0.8.1 (16 April 2001)
198 * degree(), ldegree(), coeff(), lcoeff(), tcoeff() and collect() can now
199 be used with constants, functions and indexed expressions as well, so you
200 can use it to collect by powers of Pi or sin(x), or to find the coefficient
203 - it only works with symbols, constants, functions and indexed expressions,
204 trying to find the coefficient of, e.g., "x^2" or "x+y" won't work;
205 - it does not know about dummy index summations; the coefficient of
206 gamma~0 in p.mu*gamma~mu should be p.0 but is returned as 0;
207 - using coeff(), tcoeff(), lcoeff() or collect() on elements of
208 noncommutative products might return wrong or surprising results.
209 * subs() no longer only substitutes symbols and indices but performs a more
210 general "syntactic substitution", i.e. it substitutes whole objects in sub-
211 expressions. You can subs((a+b)^2,a+b==3) and get 9, but subs(a+b+c,a+b==3)
214 - substituting numerics (subs(expr, 2==4)) will not replace then in all
215 occurences; in general, you shouldn't substitute numerics, though.
216 * Added preliminary (re)implementations of color and clifford classes.
217 * simplify_indexed(): contraction of symmetric and antisymmetric tensors
219 * Replaced the various print*() member functions by a single print() that
220 takes a print_context object that determines the output formatting. This
221 should make it easier to add more output types, such as LaTeX output,
222 which is based on work by Stefan Weinzierl.
223 * Added functions to retrieve the properties stored in archive objects
224 outside of unarchive() (for printing or debugging purposes).
225 * Some bugfixes (indexed objects, archive writing).
226 * .collect() on non-polynomials is now algebraically correct.
228 0.8.0 (24 March 2001)
229 * Complete revamp of indexed objects. Instead of multiple classes for
230 indexed things and their indices there is now only one "indexed" class
231 and two types of indices: "idx" for simple indices and "varidx" for
232 indices with variance. There are predefined delta, epsilon and metric
233 tensors, and a function simplify_indexed() that performs canonicalization
234 and dummy index summations. Matrix objects can be indexed for doing simple
236 * Added an option "expand_indexed" to expand() to perform expansion of
237 indexed objects like (a+b).i -> a.i + b.i
238 * Renamed get_indices() to get_free_indices(), which no longer returns
239 dummy indices and checks the consistency of indices in sums.
240 * sqrfree() factorization fixed and improved syntactically.
241 * subs() works on matrices.
242 * Matrices can be constructed from flat list of elements; diagonal matrices
243 can be constructed from list of diagonal elements with diag_matrix().
244 * Fixed memory leak in expand().
245 * Operator% for objects of class ncmul has gone. Use operator* now for that
246 case too, which is much more natural.
248 0.7.3 (28 February 2001)
249 * Several bugfixes and minor performance tunings.
250 * Added a section to the tutorial about adding new algebraic classes to GiNaC.
251 * Closed many in-source documentation gaps.
253 0.7.2 (17 February 2001)
254 * Several bugfixes in power series expansion, one of them critical.
256 0.7.1 (7 February 2001)
257 * Fix problems with Cint that were caused by CLN's overloaded operator new.
258 * Fix compilation errors with GCC3.
259 * normal() handles large sums of fractions better and normalizes the exponent
260 of power expressions.
261 * expand() always expands the exponent and transforms x^(a+b) -> x^a*x^b.
262 * Some bugfixes of series expansion around branch cuts of special functions.
264 0.7.0 (15 December 2000)
265 * Requires CLN 1.1 now. Class numeric doesn't use an indirect pointer to the
266 actual representation any more. This is a speedup.
267 * mul::expand() was reengineered to not allocate excess temporary memory.
268 * Non-integer powers of a symbol are treated as constants by (l)degree() and
269 coeff(). Using these functions on an expression containing such powers used
270 to fail with an internal error message. The side-effect is that collect()
271 can be used on expressions which are not polynomials.
272 * Added a man page for the ginac-config script.
273 * Ctor of numeric from char* honors Digits.
275 0.6.4 (10 August 2000)
276 * Complete revamp of methods in class matrix. Some redundant (and poor)
277 implementations of elimination schemes were thrown out. The code is now
278 highly orthogonal, more flexible and much more efficient.
279 * Some long standing and quite nasty bugs were discovered and fixed in the
280 following functions: add::normal(), heur_gcd(), sr_gcd() and Order_eval().
283 * Derivatives are now assembled in a slightly different manner (i.e. they
284 might 'look' different on first sight). Under certain circumstances this
285 can result in a dramatic speedup because it gives hashing a better chance,
286 especially when computing higher derivatives.
287 * Some series expansions of built-in functions have been reengineered.
288 * The algorithm for computing determinants can be chosen by the user. See
289 ginac/flags.h and ginac/matrix.h.
290 * The Dilogarithm (Li2) now has floating point evaluation, derivative and a
291 proper series expansion.
292 * Namespace 'std' cleanly disentangled, as demanded by ISO/EIC 14882-1998(E).
293 * Some minor bugfixes, one major lsolve()-bugfix and documentation updates.
296 * ginaccint.bin is now launched by a binary program instead of by a scripts.
297 This allows us to write #!-scripts. A small test suite for GiNaC-cint was
299 * Several minor bugfixes.
302 * Cleanup in the interface to Cint. The required version is now Cint 5.14.38.
303 * Several bugfixes in target install.
306 * IMPORTANT: Several interface changes make programs written with GiNaC
307 much clearer but break compatibility with older versions:
308 - f(x).series(x,p[,o]) -> f(x).series(x==p,o)
309 - series(f(x),x,p[,o]) -> series(f(x),x==p,o)
310 - gamma() -> tgamma() (The true Gamma function, there is now also
311 log(tgamma()), called lgamma(), in accord with ISO/IEC 9899:1999.)
312 - EulerGamma -> Euler
313 * #include'ing ginac.h defines the preprocessor symbols GINACLIB_MAJOR_VERSION,
314 GINACLIB_MINOR_VERSION, and GINACLIB_MICRO_VERSION with the respective GiNaC
315 library version numbers.
316 * Expressions can be constructed from strings like this:
317 ex e("2*x+y", lst(x, y));
318 * ex::to_rational() provides a way to extend the domain of functions like
319 gcd() and divide() that only work on polynomials or rational functions (the
320 good old ex::subs() method reverses this process)
321 * Calling diff() on a function that has no derivative defined returns the
322 inert derivative function "Derivative".
323 * Several new timings in the check target. Some of them may be rather rude
324 at your machine, feel free to interrupt them.
326 0.5.4 (15 March 2000)
327 * Some algorithms in class matrix (notably determinant) were replaced by
328 less brain-dead ones and should now have much better performance.
329 * Checks were completely reorganized and split up into three parts:
330 a) exams (small regression tests with predefined input)
331 b) checks (lenghty coherence checks with random input)
332 c) timings (for coherence and crude benchmarking)
333 * Behaviour of .evalf() was changed: it doesn't .evalf() any exponents.
334 * Expanded expressions now remember they are expanded to prevent
335 superfluous expansions.
336 * Small bugfixes and improvements in the series expansion.
338 0.5.3 (23 February 2000)
339 * A more flexible scheme for registering functions was implemented,
340 allowing for remembering, too.
343 0.5.2 (16 February 2000)
344 * Mainly fixes a bug in the packaging of release 0.5.1.
346 0.5.1 (14 February 2000)
347 * Fixes a small number of bugs.
349 0.5.0 (7 February 2000)
350 * Expressions can be written ("archived") to files and read therefrom.
351 * Addition of GiNaC-cint, which lets you write complete programs in
352 an interactive shell-like manner in your favoured programming
355 0.4.1 (13 December 1999)
356 * Series Expansion of Gamma function and some other trigonometric
357 functions at their poles works now.
358 * Many more evaluations of special functions at points where
360 * info_flags::rational doesn't return true for complex extensions
361 any more---use info_flags::crational for the old behaviour.
362 info_flags::integer and -::cinteger work similarly, the same
363 holds for types like info_flags::rational_polynomial.
365 0.4.0 (26 November 1999)
366 * First public release.