Site Xml Log


Commit Date  
[898374] (version-5_12_99rc1) by robert_dodier robert_dodier

Commit minor changes from trunk to release branch.

2007-08-04 19:44:13 Tree
[cdaad5] (release-5_13_0-base) by lbutler lbutler

This commit was manufactured by cvs2svn to create branch 'RELEASE-
5_13_0-BRANCH'.

2007-08-04 16:30:59 Tree
[25a9c3] by robert_dodier robert_dodier

Put share_testsuite.mac and contrib/rtest_lsquares.mac on list of share files.

2007-08-04 16:30:58 Tree
[b7e75b] by riotorto riotorto

Some small changes to avoid errors in texi2pdf

2007-08-03 22:28:40 Tree
[ac526b] by robert_dodier robert_dodier

Repair rtest8 #69 (put named variable in expression instead of %%).

2007-08-03 15:54:25 Tree
[3504f0] by robert_dodier robert_dodier

Increase tolerance for rtest8 #69 (find_root example) so it succeeds.

2007-08-03 05:39:48 Tree
[b6996c] by robert_dodier robert_dodier

(1) New global variable float_approx_equal_tolerance, which is
the tolerance applied by the function float_approx_equal
(for comparison of floating point values in the testsuite).
(2) Rename dfloat_approx_equal to float_approx_equal
(other float functions are all float_foo; Maxima uses only
double precision so no need to try to distinguish double float).

2007-08-03 05:38:49 Tree
[590244] by robert_dodier robert_dodier

Make simplode behave according to documentation
(separator string is now "" instead of "&").

2007-08-03 05:34:44 Tree
[9d2cf1] by robert_dodier robert_dodier

Adjust an expected result to accomodate current testsuite code
(expected results are not evaluated now).

2007-08-03 05:33:39 Tree
[b3df73] by robert_dodier robert_dodier

Assign greater tolerance to float_approx_equal_tolerance (new global) for all tests.

2007-08-03 05:32:23 Tree
[e6a765] by robert_dodier robert_dodier

Update test cases and expected results to be appropriate
for current versions of defstruct code and testsuite code.

2007-08-03 05:31:13 Tree
[65aee9] by robert_dodier robert_dodier

(1) Assign a greater tolerance for floating point numbers.
(2) Change all Lisp strings in expected results to Maxima strings.

2007-08-03 05:29:37 Tree
[9b1e7f] by robert_dodier robert_dodier

Change sconcat to concat throughout, so that results
contain Maxima strings instead of Lisp strings.
That makes it easier to test that expected results are produced.

2007-08-03 05:27:51 Tree
[c2bdc5] by robert_dodier robert_dodier

Load package before running test.

2007-08-03 05:25:21 Tree
[adf952] by robert_dodier robert_dodier

New script to set up testsuite_files so that run_testsuite runs
tests in share packages. Initial revision runs all rtest*.mac scripts in share
except for all the ones in share/contrib/diffequations/tests/.

2007-08-03 05:19:26 Tree
[5635b7] by andrejv andrejv

Fixed a typo

2007-07-31 14:18:24 Tree
[e55eac] by andrejv andrejv

Updating the graphs package.

2007-07-31 11:38:18 Tree
[985768] by hgeyer hgeyer

Added the vector3d package of Walter Eastes (eastes@infinet.com) as
share/contrib/vector3d. This closes patch tracker item #974304.

This commit is based on a newer version (than in the patch tracker)
with clarified copyright status. I made some minor changes to
vector3d.usg to make the tone more appropriate with regard to the
current status of the other vector packages in maxima.

2007-07-30 13:03:48 Tree
[f78eeb] by robert_dodier robert_dodier

Change convention for residuals to lhs minus rhs (was rhs minus lhs).
New convention is consistent with definition of error as y = foo(x) + error
which is typical in statistics texts. Code, tests, and documentation all changed.

2007-07-30 04:31:53 Tree
[2bd1f4] by robert_dodier robert_dodier

In $VANDERMONDE_MATRIX, throw in an "ignore" declaration to quiet
a warning message generated by SBCL for rtest2.
(This file is autoloaded to get the Hessian function.)

2007-07-30 03:02:37 Tree
[829740] by robert_dodier robert_dodier

Replace previous lsquares (least squares) package.
Features of new version:

(1) Formulate numerical problem as a minimization (solved via lbfgs)
instead of a system of equations (solved via mnewton).
The minimization formulation is more likely to converge
(since there is always a notion of the correct direction, namely downhill)
(2) Separate functions for exact vs numerical estimates
(3) Separate functions to construct the mean square error expression,
residuals, and residual mean square error.
Residual MSE function supersedes the DETCOEF global variable
(4) Test script with many examples. Includes examples from previous
implementation and others
(5) Revised documentation

Files:

* share/contrib/lsquares.mac: replaced implementation
* share/contrib/rtest_lsquares.mac: test script
* doc/info/lsquares.texi: revised documentation

Implementation note: Values obtained for numerical estimates vary
from one Lisp implementation to another. That's disconcerting but
I don't know how to make them all agree. The test script fudges
this point by testing some examples with a greater tolerance.

2007-07-30 02:57:52 Tree
[e3cc50] by dgildea dgildea

Fixes for bugs 1741705 and 1748168.

src/defint.lisp
intsc1 no longer requires limits of integration to be multiples of %pi.
uses pretty-good-floor-or-ceiling to find nearest period.
substitutes lower limit into integrand expression only as a last resort, as
substitution tends to cause intsc to miss discontinuities in antiderivative.

tests/rtestint.mac
new test cases from bug reports.

2007-07-29 16:17:37 Tree
[1afb6c] by willisbl willisbl

o fixes for subset and partition_set (unreported bugs)
o fix typo in source code comment

2007-07-29 12:35:34 Tree
[bc0bef] by andrejv andrejv

Added simpsum:true in two places to adapt to the new sum behaviour.

2007-07-27 16:54:01 Tree
[a97907] by riotorto riotorto

Style changes and Spanish translation

2007-07-26 14:21:19 Tree
Older >