FFTW 3.3.4
* New functions fftw_alignment_of (to check whether two arrays are
equally aligned for the purposes of applying a plan) and fftw_sprint_plan
(to output a description of plan to a string).
* Bugfix in fftw-wisdom-to-conf; thanks to Florian Oppermann for the
bug report.
* Fixed manual to work with texinfo-5.
* Increased timing interval on x86_64 to reduce timing errors.
* Default to Win32 threads, not pthreads, if both are present.
* Various build-script fixes.
CHANGES IN R 3.0.3:
NEW FEATURES:
* On Windows there is support for making .texi manuals using
texinfo 5.0 or later: the setting is in file
src/gnuwin32/MkRules.dist.
A packaging of the Perl script and modules for texinfo 5.2 has
been made available at <URL:
http://www.stats.ox.ac.uk/pub/Rtools/>.
* write.table() now handles matrices of 2^31 or more elements, for
those with large amounts of patience and disc space.
* There is a new function, La_version(), to report the version of
LAPACK in use.
* The HTML version of 'An Introduction to R' now has links to PNG
versions of the figures.
* There is some support to produce manuals in ebook formats. (See
doc/manual/Makefile. Suggested by Mauro Cavalcanti.)
* On a Unix-alike Sys.timezone() returns NA if the environment
variable TZ is unset, to distinguish it from an empty string
which on some OSes means the UTC time zone.
* The backtick may now be escaped in strings, to allow names
containing them to be constructed, e.g. `\``.
* read.table(), readLines() and scan() now warn when an embedded
nul is found in the input.
* KalmanForecast(fast = FALSE) is now the default, and the help
contains an example of how fast = TRUE can be used in this
version. (The usage will change in 3.1.0.)
* strptime() now checks the locale only when locale-specific
formats are used and caches the locale in use: this can halve the
time taken on OSes with slow system functions (e.g. OS X).
* strptime() and the format() methods for classes "POSIXct",
"POSIXlt" and "Date" recognize strings with marked encodings:
this allows, for example, UTF-8 French month names to be read on
(French) Windows.
* iconv(to = "utf8") is now accepted on all platforms (some
implementations did already, but GNU libiconv did not: however
converted strings were not marked as being in UTF-8). The
official name, "UTF-8" is still preferred.
* available.packages() is better protected against corrupt metadata
files.
* Finalizers are marked to be run at garbage collection, but run
only at a somewhat safer later time (when interrupts are
checked). This circumvents some problems with finalizers running
arbitrary code during garbage collection (the known instances
being running options() and (C-level) path.expand()
re-entrantly).
Changelog:
Maxima 5.32 change log
Changes not backward compatible:
* share function graph2d: removed in favor of plot2d
New items in core:
* function limit: limits for fresnel_c and fresnel_s
* function plot2d: adaptive plotting for parametric plots
* functions round, truncate: distribute over lists and other aggregate types
* function zn_characteristic_factors: Shanks characteristic factors
New items in share:
* new package smath: glue code for SMath Studio (notebook interface)
* new package logic: some functions for Boolean expressions
* package lapack: new function zheev (eigenvalues of Hermitian matrix)
Changes in core:
* function labels: make 'labels' an ordinary function (not argument-quoting)
* plotting: more general color specification
* plotting: rework plot options machinery
* plotting: introduce some new options
Other changes:
* Build and install mgnuplot.exe under windows
Bug fixes:
2668 Bigfloat Gamma Inaccurate for Small Inputs
2666 plotdf ignores first of [x, ...] or [y, ...] if there are no other options
2662 Maxima manual adjust
2660 lapack puts binary files in binary/binary-$foo/share/lapack/binary-$foo/lapack
2659 bug_report() directions for creating ticket need updating
2649 Correction for $matrix evaluation
2639 limit asks about internal variable
2634 zgeev does not operate on real matrices
2633 ev(integrate,numer) gives strange result
2630 inverse_jacobi_cn(-2.0, 3.0) generates an error
2629 elliptic_kc(3.0) not accurate
2628 fourier elimination on equality
2624 Erroneous limit result
2615 Numeric evaluation of inverse Jacobi elliptic functions is wrong for some inputs
2282 Problem in LISP with times function in sym
unnumbered bugs:
mailing list 2013-09-26: transcendental functions and compilation
mailing list 2007-04-02: fixes a bug in plot2d-discrete
GCL bug #40646 fixed (problem actually in Maxima)
workaround GCL bug #40154 (some float literals parsed incorrectly)
ensure that 'sign' does not make any changes to assume database.
fix Postscript plotting on windows
fix limit(-(3*n^2 + 1)*(-1)^n/sqrt(n^5 + 8*n^3 + 8),n,inf);
fix bug in ldisplay and ldisp: need to display %t label
fix: inv_mod(0,1) should return false
fix cyclic-p, gf-eval, gf-irr-p
fix gf-cminus-b, gf-prim-p
fix bug in computing (.75b0+%i*1.5b0)^(%i/4)
fix "describe" returning wrong entry with texinfo version 5
plots using the dumb gnuplot terminal were not being displayed in the terminal when the gnuplot_pipes is used
Maxima 5.31 change log
Changes in core:
* Galois fields: revision, introduction of extension fields
* inverse_jacobi_sn: improve accuracy
* Bessel functions: integral of bessel_i, bessel_j of arbitrary order
* matrix functions: revise ModeMatrix, mat_function, dispJordan, diag, JF
* signum: do not declare signum real- or integer-valued
* carg: declare carg real-valued
* parser: add L, l, W, and w to list of exponent marker characters
Changes in share:
* package stringproc: new functions readchar, md5sum, base64
Other changes:
* documentation: minor changes to make Texinfo 5 happy
* build system: changes to make Maxima build correctly on Windows
Bug fixes:
* 2622 base64 type warnings
* 2619 Function inverse_erf - error in numerical evaluation
* 2612 known_units() broken
* 2608 quad_qags(x,x) generates a lisp error
* 2607 lbfgs documentation is incomplete
* 2604 fft.dem is out of date
* 2603 fft documentation
* 2602 lu_factor regression
* 2599 cl-info error handling
* 2597 logcontract and subscripted log
* 2594 Incorrect integration of product of cosines
* 2591 risch gives Lisp error for sinh expression
* 2583 sign error for integrate(x^(8*%i-1),x);
* 2570 Make acos(cos(x)) simplify to x when on correct interval
* 2569 translate rat(1,x) and rat([1]) incorrect
* 2536 Lisp error in limit: \
* 2501 %pi/8 is definitely not an integer
* 2475 limit(x^(mu/x^g),x,inf) --> error
unnumbered bugs:
* mailing list 2013-05-22: "array-lambda error with 5.30.0"
* mailing list 2013-05-07: "defrule regression compare to version 5.29.1"
* fix rk() when the independent variable's initial value is nonzero
* fix bug in asksign
* fix wrong noun form returned for jacobi_nc(elliptic_kc(m)/n,m)
Maxima 5.30 change log
Changes not backward compatible:
* function gf_set: renamed to gf_set_data
Other changes:
* function bfzeta: extend to complex plane
* function gf_set: return a structure
* function gf_exp: allow negative exponents
* functions plot2d, plot3d: accept subscripted variables as independent variables
* package dynamics: revise function rk
* package ezunits: new functions expand_dimensional, solve_dimensional, other changes
* package linearalgebra(??): revise mat_function
* ??: revise functions jordan, JF, diag
Bug fixes:
??: limit((log(y+h)-log(y))/h,h,0,plus)
2561: limit(log(x^2),x,-20) gives 2*log(-20)
2559:
2557: abs_integrate leaks assumptions into enclosing context
2547:
2542:
2541: (FIXED??)
2540:
2531: Integration with inf
2529: limit(x^n,x,minf) gives minf for n positive
2528: imagpart etc. wrong when var declared real and complex
2527: exponent too big in limit
2517:
2509: fresnel_s incorrect for small values
2496:
2458: to_poly_solve gives a wrong solution for cos(x)=sin(3x)
2449: cabs(expr) fails when expr is 0
2442: arrayinfo gives a Lisp error
2441:
2412: Problems with integral (x/(exp(x)+1),x,0,inf)
2362:
2350:
2279:
2219: tex2ooo literals format
472: ctrl-D causes Maxima debugger to loop endlessly (clisp only)
unnumbered bugs:
mailing list circa 2013-03-30: load_pathname unset with maxima -b
mailing list 2013-02-21: gamma_incomplete_regularized(a,0) returning the wrong noun form
mailing list 2012-12-23: "let(simp) behaves different when loaded from batch file"
mailing list 2012-12-14: "Maxima won,t pull out a constant from a linear function"
bern() with zerobern:false
clean up facts introduced by sign functions
Upstream changes:
27 September 2013: statmod 1.4.18
- Update reference for permp().
- bug fix to elda() so that it returns NA for the tests instead of
giving an error when the Fisher information for the slope isNA.
- Exact roles of authors specified in DESCRIPTION file.
- All usage lines in help files wrapped at 90 characters to ensure
that code is not truncated in pdf manual.
Upstream changes:
(But no changes for 0.4.7)
Version 0.4-6 2013/10/30
Added function dasytriangular for the density of asymmetric triangular distribution as described in Mardia (1972)
In this version two functions are available for evaluating the median (medianCircular which allows for the dispersion para
meter and median.circular with only the median reported), however medianCircular is deprecated and it will be removed in t
he next releases.
Added mean.circular.c to calculate the circular mean in C language.
New algorithm for circular median. The implementation is in C language.
Added function for Hodges Lehmann estimate of the median for circular and non circular data.
Several bugs fixed.
Upstream changes:
Changes in Version 1.7-11
o Extended the license from "GPL-2" to "GPL-2 | GPL-3".
o Avoid duplications in Suggests/Imports/Depends in the package's
DESCRIPTION. Use only :: instead of ::: to access certain functions
from other namespaces.
o Added as.list() methods for yearmon/yearqtr.
o Added workaround in Ops.zoo if first argument is not a zoo series
(prompted by Josh Ulrich).
o The transform() method for zoo series now does what method for
data.frame does (instead of just calling it) in order to get
non-standard evaluation.
* New, no 2to3, Python 2 and Python 3 are supported by a common code base.
* New, gufuncs for linear algebra, enabling operations on stacked arrays.
* New, inplace fancy indexing for ufuncs with the ``.at`` method.
* New, ``partition`` function, partial sorting via selection for fast median.
* New, ``nanmean``, ``nanvar``, and ``nanstd`` functions skipping NaNs.
* New, ``full`` and ``full_like`` functions to create value initialized arrays.
* New, ``PyUFunc_RegisterLoopForDescr``, better ufunc support for user dtypes.
* Numerous performance improvements in many areas.
* Support for Python versions 2.4 and 2.5 has been dropped.
* Support for SCons has been removed.
* added Symmetric/Hermitian LDLT factorization routines with rook pivoting algorithm
* 2-by-1 CSD to be used for tall and skinny matrix with orthonormal columns (in LAPCK 3.4.0, we already integrated CSD of a full square orthogonal matrix)
* New stopping criteria for balancing.
Perl extension for Munkres' solution to classical Assignment problem
for square and rectangular matrices. This module extends the solution
of Assignment problem for square matrices to rectangular matrices by
padding zeros. Thus a rectangular matrix is converted to square matrix
by padding necessary zeros.
* The API routine glp_read_mps was changed to remove free rows.
* A bug was fixed in the API routine glp_read_lp.
* The zlib compression library used by some GLPK routines and
included in the package was downgraded from 1.2.7 to 1.2.5 (as
in GLPK 4.50) because of addressability bugs on some 64-bit
platforms.
* A bug was fixed in a routine that reads gzipped files.
* Two API routines glp_get_it_cnt and glp_set_it_cnt were added.
* All obsolete GLPK API routines (prefixed with lpx) were removed
from the package.
* A set of routines that simulate the old GLPK API (as defined
in 4.48) were added; see examples/oldapi/api/lpx.c.
* A namespace bug was fixed in the SQL table drive module.
* The NEWS files have been re-organized.
This file contains news for R >= 3.0.0: news for the 0.x.y, 1.x.y
and 2.x.y releases is in files NEWS.0, NEWS.1 and NEWS.2. The
latter files are now installed when R is installed. An HTML
version of news from 2.10.0 to 2.15.3 is available as
doc/html/NEWS.2.html.
* sum() for integer arguments now uses an integer accumulator of at
least 64 bits and so will be more accurate in the very rare case
that a cumulative sum exceeds 2^53 (necessarily summing more than
4 million elements).
* The example() and tools::Rd2ex() functions now have parameters to
allow them to ignore \dontrun markup in examples. (Suggested by
Peter Solymos.)
* str(x) is considerably faster for very large lists, or factors
with 100,000 levels, the latter as in PR#15337.
* col2rgb() now converts factors to character strings not integer
codes (suggested by Bryan Hanson).
* tail(warnings()) now works, via the new `[` method.
* There is now support for the LaTeX style file zi4.sty which has
in some distributions replaced inconsolata.sty.
* unlist(x) now typically returns all non-list xs unchanged, not
just the "vector" ones. Consequently, format(lst) now also works
when the list lst has non-vector elements.
* The tools::getVignetteInfo() function has been added to give
information about installed vignettes.
* New assertCondition(), etc. utilities in tools, useful for
testing.
* Profiling now records non-inlined calls from byte-compiled code
to BUILTIN functions.
* Various functions in stats and elsewhere that use non-standard
evaluation are now more careful to follow the namespace scoping
rules. E.g. stats::lm() can now find stats::model.frame() even
if stats is not on the search path or if some package defines a
function of that name.
* If an invalid/corrupt .Random.seed object is encountered in the
workspace it is ignored with a warning rather than giving an
error. (This allows R itself to rely on a working RNG, e.g. to
choose a random port.)
* seq() and seq.int() give more explicit error messages if called
with invalid (e.g. NaN) inputs.
* When parse() finds a syntax error, it now makes partial parse
information available up to the location of the error. (Request
of Reijo Sund.)
* Methods invoked by NextMethod() had a different dynamic parent to
the generic. This was causing trouble where S3 methods invoked
via lazy evaluation could lose track of their generic.
(PR#15267)
* Code for the negative binomial distribution now treats the case
size == 0 as a one-point distribution at zero.
* abbreviate() handles without warning non-ASCII input strings
which require no abbreviation.
* read.dcf() no longer has a limit of 8191 bytes per line. (Wish of
PR#15250.)
* formatC(x) no longer copies the class of x to the result, to
avoid misuse creating invalid objects as in PR#15303. A warning
is given if a class is discarded.
* Dataset npk has been copied from MASS to allow more tests to be
run without recommended packages being installed.
* The initialization of the regression coefficients for
non-degenerate differenced models in arima() has been changed and
in some examples avoids a local maximum. (PR#15396)
* termplot() now has an argument transform.x to control the display
of individual terms in the plot. (PR#15329)
* format() now supports digits = 0, to display nsmall decimal
places.
* There is a new read-only par() parameter called "page", which
returns a logical value indicating whether the next plot.new()
call will start a new page.
* Processing Sweave and Rd documents to PDF now renders backticks
* utils::modifyList() gets a new argument keep.null allowing NULL
components in the replacement to be retained, instead of causing
corresponding components to be deleted.
* tools::pkgVignettes() gains argument check; if set to TRUE, it
will warn when it appears a vignette requests a non-existent
vignette engine.
--------------------------------------------------------------------------
NEWS for version 1.1 (released on October 28, 2013)
--------------------------------------------------------------------------
New and Changed Features
========================
o Added a new operator on polyhedra: the positive time elapse.
o In the Java language interface:
- The constraint/generator/... system classes now extend the ArrayList
generic container (rather than Vector);
- Variable objects are now built from a long (rather than int) value,
thereby matching the type used elsewhere for space dimensions;
- added new static method to Variable class
void setStringifier(Variable_Stringifier)
where Variable_Stringifier is an interface allowing for
customization of the output routine for variable's names
(see example in interfaces/Java/tests/Variable_Output_test1.java);
- added value NOT_EQUAL to enumeration Relation_Symbol.
Bugfixes
========
o Portability improved.
o Fixed a precision regression in Polyhedron method
void drop_some_non_integer_points(const Variables_Set&,
Complexity_Class);
o In the Java language interface, fixed a C++/Java conversion error
whereby the construction of a valid Variable object in JNI code
was leading to an exception being thrown. The bug has only been
observed on 32-bit builds.
o In the Java interface, fixed declaration of methods
void drop_some_non_integer_points(...);
so as to accept a Complexity_Class enum value.
o Fixed an issue in method MIP_Problem::OK() whereby the method
was trying to enforce a non-invariant condition.
--------------------------------------------------------------------------
NEWS for version 1.0 (released on June 28, 2012)
--------------------------------------------------------------------------
New and Changed Features
========================
o Significant improvements have been obtained in both time and space
resource usage by the definition of data structures and algorithms
for the case of "sparse rows", i.e., sequences of coefficients
where most of the values are zero.
o The library fully supports two different representations for rows:
the "dense" representation is an array-like representation tailored
to sequences having most of their coefficients different from zero;
the "sparse" representation saves memory space (as well as CPU
cycles) when most of the coefficients in the sequence are zero.
o A generic interface allows for a seamless interaction between the
dense and the sparse row representation. Most library entities
(linear expressions, constraints, generators, congruences, and
their systems) can be built using either representation, specified
as a constructor's argument.
o As a by-product of this sparse/dense refactoring work, efficiency
improvements have been obtained even for those computations that
are still based on the dense row representation.
o Reasonable default values for the row representation are provided
for each library entity, automatically leading to significant
memory space savings even in old client/library code, e.g., when
dealing with constraint systems describing weakly relational
abstractions such as boxes and octagonal shapes.
o If desired, these default values can be customized to user's needs
by changing just a few lines of library code. For instance, the
constraint systems stored inside C_Polyhedron and NNC_Polyhedron
objects can be made to use the sparse representation by just
changing the following line in Polyhedron.defs.hh:
static const Representation default_con_sys_repr = DENSE;
to become
static const Representation default_con_sys_repr = SPARSE;
Bugfixes
========
o Fixed a bug affecting methods
bool BD_Shape<T>::contains(const BD_Shape& y) const;
bool Octagonal_Shape<T>::contains(const Octagonal_Shape& y) const;
whereby the wrong result was obtained when *this is an empty
weakly-relational shape and y is not empty.
o Fixed a bug affecting the PIP solver whereby a wrong result could have
been obtained if the input constraint system contained multiple linear
equality constraints.
--------------------------------------------------------------------------
NEWS for version 0.12.1 (released on April 16, 2012)
--------------------------------------------------------------------------
New and Changed Features
========================
o In the C, Java, OCaml and Prolog interfaces, modified the signature
of the function/method/predicate for setting the deterministic timeout
threshold. The new interfaces take two input values, named `unscaled'
and `scale', that are used to compute the threshold value as
`unscaled * 2^scale'.
o Added new Box<ITV> methods
bool has_upper_bound(Variable var,
Coefficient& n, Coefficient& d, bool& closed) const;
bool has_lower_bound(Variable var,
Coefficient& n, Coefficient& d, bool& closed) const;
to query a non-empty box for the existence and value of its upper/lower
bound on variable `var'. The methods have been also added to all the
available language interfaces.
o Two BibTeX databases of papers related to the Parma Polyhedra Library
have been added to the distribution (in the `doc' directory).
Bugfixes
========
o Restored the support for deterministic timeouts in the PIP solver
(it was removed by accident in PPL 0.12).
o Minor documentation fixes.
o Portability improved.
--------------------------------------------------------------------------
NEWS for version 0.12 (released on February 27, 2012)
--------------------------------------------------------------------------
New and Changed Features
========================
o New configure options `--with-gmp=DIR', `--with-gmp-include=DIR' and
`--with-gmp-lib=DIR' supersede the (now removed) option
`--with-gmp-prefix'. (The old option never really worked; hopefully
this is the last change in this area.)
o New configuration option `--disable-documentation'. When specified
no new documentation is built: only the documentation already present
in the source tree is installed upon `make install'.
o The resolution process for PIP_Problem now better exploits the
integrality of parameters to simplify the newly generated tautological
constraints, the splitting constraints of decision nodes, and the
expressions defining artificial parameters.
o The implementations of the MIP and PIP solvers are based on a new
data structure leading to significant space and time savings when
the tableau matrix is sparse; the benchmarks of the ppl_lpsol demo
show an improvement on the average case, that grows when the toughest
tests in the benchmark suite are considered.
o When the `--check' option is used, the input data for demo ppl_lpsol
is perturbed the same way as GLPK does, thereby allowing for a
meaningful comparison of the results obtained.
o The input routine for PPL numeric datatypes has been extended to
accept the ISO9899 (C99) hexadecimal floating constant syntax.
o The Parma Watchdog Library has been merged into the
Parma Polyhedra Library.
Bugfixes
========
o Corrected a precision bug in methods
Box<ITV>::upper_bound_assign(const Box&)
Box<ITV>::upper_bound_assign_if_exact(const Box&)
whereby, provided any argument is an empty box and under other rather
specific conditions, the computed result was correct but unnecessarily
imprecise.
o Corrected a bug in method
Grid::relation_with(const Constraint&) const
whereby, under specific conditions, the method was creating invalid
Grid_Generator objects and providing an incorrect result.
### 0.9.7 / 04.02.2014
* Avoid exception when reading text objects
* Add test for drawings with text (currenty broken)
* Restore xlsopcodes script which had been mangled in previous commits
* Remove ruby 1.9 from roadmap, it's already working fine
* Fix excel file format documentation which had been mangled in previous commits
either because they themselves are not ready or because a
dependency isn't. This is annotated by
PYTHON_VERSIONS_INCOMPATIBLE= 33 # not yet ported as of x.y.z
or
PYTHON_VERSIONS_INCOMPATIBLE= 33 # py-foo, py-bar
respectively, please use the same style for other packages,
and check during updates.
Use versioned_dependencies.mk where applicable.
Use REPLACE_PYTHON instead of handcoded alternatives, where applicable.
Reorder Makefile sections into standard order, where applicable.
Remove PYTHON_VERSIONS_INCLUDE_3X lines since that will be default
with the next commit.
Whitespace cleanups and other nits corrected, where necessary.
