Maxima -- GPL CAS based on DOE-MACSYMA - Code Log


Commit Date  
[5aa3ab] by Jaime Villate Jaime Villate

Changes the default value of gnuplot_view_args in Windows to "-persist ~s",
to reflect the change in v. 4.6.3 of gnuplot

Documents system variables gnuplot_commad, gnuplot_view_args and
gnuplot_file_args.

Changes the default value of plot_option gnuplot_default_term_command, in Windows, to "set term pop".

2013-07-30 14:22:21 Tree
[736075] by Volker van Nek Volker van Nek

update files in contrib gf

2013-07-30 14:18:05 Tree
[2037eb] by Volker van Nek Volker van Nek

updating tests and doc for numth

2013-07-30 14:15:44 Tree
[953269] by Volker van Nek Volker van Nek

revision of GF, introducing EF, extension fields

2013-07-30 14:12:26 Tree
[fe7996] by Volker van Nek Volker van Nek

updating share testsuite

2013-07-30 14:07:08 Tree
[cdc8be] by Volker van Nek Volker van Nek

some work on the stringproc manual

2013-07-30 14:03:51 Tree
[8db6aa] by Volker van Nek Volker van Nek

small changes in stringproc test files

2013-07-30 13:46:23 Tree
[af831b] by Volker van Nek Volker van Nek

two small fixes in printf

2013-07-30 13:43:40 Tree
[522084] by Volker van Nek Volker van Nek

new function readchar in stringproc

2013-07-30 13:41:52 Tree
[7ce44b] by Volker van Nek Volker van Nek

work on the German manual

2013-07-30 13:37:37 Tree
[0f5708] by Leo Butler Leo Butler

Fix issue #2607: lbfgs documentation is incomplete. Add documentation on how to pass functions as arguments to lbfgs, update the examples.

2013-07-27 17:01:48 Tree
[0c3c1e] by Robert Dodier Robert Dodier

Merge branch 'master' of ssh://git.code.sf.net/p/maxima/code

2013-07-22 17:05:01 Tree
[5ead80] by Robert Dodier Robert Dodier

Ensure that list of conversions is constructed before trying to
use it. Resolves SF bug # 2612: "known_units() broken"

2013-07-22 16:55:28 Tree
[886955] by Raymond Toy Raymond Toy

Fix issue #2608

Signal a maxima error about incorrect number of arguments instead of
lisp error. This gives a nicer error message and doesn't dump you to
the Lisp debugger if maxima is started with -g.

2013-07-10 02:57:56 Tree
[3f41fe] by Robert Dodier Robert Dodier

Update fft.dem. Resolves SF bug # 2604: "fft.dem is out of date"
Thanks to Leo Butler for the patch.

2013-07-03 05:59:04 Tree
[66cb77] by Robert Dodier Robert Dodier

Fix up signs of sine and cosine terms in Fourier transform.
Resolves SF bug # 2603: "fft documentation". Thanks to Leo Butler for the patch.

2013-07-03 05:38:26 Tree
[ec587f] by Rupert Swarbrick Rupert Swarbrick

Document and simplify the logic for risplit-expt-sqrt-pow

This took me AGES to understand and I rewrote it wrongly several ways,
so I thought I should write up some hints for the next poor soul that
has to read it...

2013-07-02 23:08:15 Tree
[82bdd9] by Rupert Swarbrick Rupert Swarbrick

Disembowel risplit-expt

We chop the contents of the massive COND form into functions in their
own right. There shouldn't be any change what gets computed and how,
but there should be a great improvement in readability!

2013-07-02 23:07:59 Tree
[ab9d9f] by Rupert Swarbrick Rupert Swarbrick

Call SININT in HYPERTRIGINT1 instead of calling INTEGRATOR directly

This ensures that the special variables that INTEGRATOR
expects ($OPSUBST, VARLIST, GENVAR and STACK) all get bound
correctly. Fixes bug 2591 and doesn't slow down the test suite. The
only changed result in the testsuite is a minor change to the format
of an expression in rtest_integrate_special.

2013-07-02 14:24:56 Tree
[22d9db] by Rupert Swarbrick Rupert Swarbrick

Get rid of the *trig special variable

There's no real reason not to pass this around as an argument, and
doing so makes debugging stuff in trgred much easier, since you don't
have to remember to wrap everything in a (let ((*trig t)) ...) form.

2013-07-02 13:34:00 Tree
[cd26b0] by Rupert Swarbrick Rupert Swarbrick

Fix bug in trigreduce with complicated expressions

This fixes bug 2594. In particular, if you define

testit(a, n) := block([initial: product(cos(k*x), k, 1, n)],
[subst(a, x, initial),
subst(a, x, trigreduce(initial))]
)$

then testit(0,k) gave the right answer for k=1,...,7, but gave
something silly for k=8.

It turns out that there was a logic error in sp1tplus. That function
calls sp1sintcos repeatedly and gets back either a term of the form
sin(foo) or cos(foo) or it gets a sum of such terms. In the latter
case, if it started with a coefficient on the front, it might end up
multiplying the coefficient by the sum. Unfortunately, this meant that
on the next time round, the term wasn't in the right form and
sp1sintcos started giving silly answers. The new code is much more
paranoid and explicitly throws an error if it gets a result in a form
that it doesn't recognise.

2013-07-02 11:30:48 Tree
[ab681c] by Rupert Swarbrick Rupert Swarbrick

Rewrite some of sp1times using dotimes/push/mapc

No change to functionality (provided the implementation was bright
enough to spot that we were throwing away the result of a mapcar), but
slightly easier to read.

2013-07-01 23:24:28 Tree
[867b25] by Rupert Swarbrick Rupert Swarbrick

Switch to using lenient-extended-realp to test $defint limits

This was triggered by Bug 2600, where the old code wasn't happy with
an upper limit of "p(x)", since imagpart(p(x)) simplified to
itself (which is not equal to zero). Hopefully this change strikes the
right balance with

integrate (f(x), x, 0, %i);

or

integrate (f(x), x, 0, [1,2]);

giving an error. But

integrate (sin(x), x, 0, g(y)); => 1 - cos(g(y))

2013-07-01 21:19:37 Tree
[90a78f] by Rupert Swarbrick Rupert Swarbrick

Hush an unused argument warning.

2013-07-01 21:15:28 Tree
[24090c] by Rupert Swarbrick Rupert Swarbrick

Make some-bfloatp and some-floatp deal correctly with rat expressions.

Fixes bug 2602.

2013-07-01 20:57:13 Tree
Older >