Main Changes Scilab 2.7 -> 3.0
==============================
NEW FUNCTIONALITIES
===================
- continue instruction added
- [a(i,j),b(..),..]=foo(..) syntax is now handled
- arpack algorithms for sparse matrix eigenvalues/eigenvectors computation added
- Levenberg-Marquardt algorithm of Minpack used to define function lsqrsolve
- New graphics mode improved
- beta and legendre function added (thanks to B Pincon)
- Scicos has been extended to handle non causal systems description (supported by RNTL)
- Matlab to Scilab translator has been re-written, and extended to current Matlab syntax.
OTHERS IMPROVEMENTS
===================
- Better discontinuities handling in Scicos
- More efficient number parsing (thanks to B Pincon)
- More efficient 3D Zoom (thanks to E Segre)
- Improved scipad, shortcuts, language handling, colorization, bug fixes.. (thanks to E Segre)
- for loop variable is no more removed at the end of the loops
- figure sizes are now taken into account in Postscript export
- Windows GUI improved
- Graphic windows refresh improved under Windows
- Scilab has been adapated to IA64 architecture
- Helvetica (Arial on Windows) fonts added (corresponding font identifiers are
from 6 to 9) (thanks to B Pincon)
- N dimensionnal matrices efficiency improved (thanks to B Pincon)
- logical expressions shortcuts handled in if
- cd, ls, clc, clf, tohome functions added
- functions with no argument can be called without () if called as a command
Changes:
Substantial change/improvement to simplify command.
Added some more code to fix floating point inaccuracies.
Improved solving. y=(x+1/x)^3 solves both ways, now.
Full simplify after every loop in derivative command added.
Move derivative and taylor commands to "diff.c".
Added "-m" option to Mathomatic (memory size multiplier).
It sets the size of equation spaces at run time.
important changes:
- Ports of Numeric functions
A bunch of Numeric functions were ported to numarray in the new
libnumeric module. To get these import from numarray.numeric. Most
notable among these are put, putmask, take, argmin, and argmax. Also
added were sort, argsort, concatenate, repeat and resize. These are
independent ports/implementations in C done for the purpose of best
Numeric compatibility and small array performance. The numarray
versions, which handle additional cases, still exist and are the
default in numarray proper.
- Faster matrix multiply
The setup for numarray's matrix multiply was moved into C-code. This
makes it faster for small matrices.
- bug fixes
For a complete list of changes see:
http://sourceforge.net/project/shownotes.php?release_id=250453
# XXX: following comment still true for buildlink3?
# R does ugly things in the configure to get the default list of arguments
# for linking Fortran, which exposes the buildlink2 tricks so clean up
# afterwards
Changes:
Renamed the Mathomatic executable to "mathomatic".
Add inverse Laplace transform. Usage: laplace inverse x
Fixed bug in solving for zero, which occurred when there were no
variables in the divisor.
Some improvement in solve routine.
This only worked with gnumeric-1.0.x, and depends on gnome1 structures.
gnumeric is at 1.2.x now, uses gnome2 and comes with an own plotting
package.
Should fix PR pkg/26009 by MLH.
This release fixes a bug related to post incrementing, some regression failures related to compiling under cygwin, and a number of gcc compiler warnings.
Changes:
Slight improvement to integrate and laplace commands.
Integrate/laplace x on (a+x)/(b+c) works now.
Added more examples to the documentation (am.htm).
2 to 10 times speedup for simplify command by fixing polynomial factoring.
The primary purpose of this release is to correct problems on big-endian
64 bit platforms. It also introduces some test code to verify that the
functions and constants work properly.
Changes since 11.2:
Doubled the default amount of memory consumption to 11 megabytes.
This allows equations twice as large.
Fixed bug factoring infinity.
Slight improvement to "integrate" command.
Trap window resize signal and set number of screen columns and rows.
Code cleanup.
Fixed factoring of constants in factor command.
Added "laplace" command.
Allow "#" as a comment, as long as it is not immediately followed by
a number.
Enabled shelling out for all versions with the "!" command.
Improvement and speedup for poly_gcd() and poly2_gcd().
Slight improvement in polynomial factoring made by deleting 1 line of code.
Use getopt(3) to parse command line options.
Made "-c" and "-h" options work together. If both are specified,
work in HTML mono mode.
CHANGES IN R VERSION 1.9.1
NEW FEATURES
o as.Date() now has a method for "POSIXlt" objects.
o mean() has a method for "difftime" objects and so summary()
works for such objects.
o legend() has a new argument 'pt.cex'.
o plot.ts() has more arguments, particularly 'yax.flip'.
o heatmap() has a new 'keep.dendro' argument.
o The default barplot method now handles vectors and 1-d arrays
(e.g., obtained by table()) the same, and uses grey instead of
heat color palettes in these cases. (Also fixes PR#6776.)
o nls() now looks for variables and functions in its formula in
the environment of the formula before the search path, in the
same way lm() etc look for variables in their formulae.
DEPRECATED & DEFUNCT
o Support for non-IEEE-754 arithmetic (which has been untested
for some time) will be removed in the next full release.
o Direct use of R INSTALL|REMOVE|BATCH|COMPILE|SHLIB is
deprecated: use R CMD instead.
o The gnome/GNOME graphics device is deprecated and will be
removed in the next full release.
Plus Bug Fixes.
Changes since 11.1c:
Moved notification of polynomial factoring success to debug level 1 or higher.
Removed "NO_COMPLEX_ROOTS" compile-time define option.
I am thinking of making this a run-time option.
Increased security of "SECURE" compile-time define.
Absolutely no file operations are allowed or linked in.
Added check for terminal with isatty().
This allows better piping into Mathomatic.
Split up change-log ("changes.txt" and "changes_old.txt").
Changed "x!" to mean gamma(x+1) for Laplace transforms.
Fixed parsing of things like "2e#", which failed before,
thinking it was scientific notation.
Wrote "primes/twin.c", to find twin primes.
Added "help usage" command.
Allow output of "help" command to be redirected to a file.
Fixed parsing of variable names (like "pig", which didn't work right).
Added "integrate" command.
Fixed "derivative" command.
Code cleanup.
Fixed "taylor" command.
The step from 1.0.12 to 1.2.12 is too large to mention all changes.
pkgsrc issues:
-use gnome2
-add python support (python23 provides a shared lib)
-XXX I've enabled libgda/gnomedb, this seems to be completelely
non-functional to me, but perhaps it is just my setup...
Changes since 11.1:
Added Java code generation (usage: "code java").
Added "list export" option, to display equations in exportable format.
They can be cut-and-pasted to another math program with this option.
Added a blank line after every equation listed with the "list" and
"code" commands. Looks better.
Made "MAX_VAR_LEN" completely flexible (can be set to any value; set to 80).
Improved factoring of polynomials with repeated factors by trying to
differentiate with respect to every variable.
Changes:
- Fixed "makefile" to be completely portable.
No longer requires GNU make.
- Simplified some code.
- Converted token_type.kind to enum type for better type checking.
- Removed all "unsigned" variables and type casts.
- Simplified "simplify" command code. Some speed up.
- Moved "*.in" and "fix*" to directory "tests".
- Implemented long variable names (up to 40 characters).
building even though it uses INSTALL_DATA on them so set SHAREMODE=644
rather than have to patch an ever increasing number of files.
Idea from Julio M. Merino Vidal.
Changes:
- Removed calc() fraction code for accuracy because of fraction slack
that was implemented in version 10.9c.
- Solve increase power function wasn't working with odd number roots.
Fixed. Appears to be another gcc optimizer bug.
- Improved simplify command.
- Tweaked factor_constants().
- Some changes to "makefile" for "readline" support.
"GNU make" or "gmake" required now.
- Fixed solving of "x^(1/99)=x".
- Fixed readline bug.
