cc22c1f6ea
Improved release instructions.
...
The site for reviewing should not be put in the staging area (otherwise
it may be published too early as this area is shared among all commons
components).
The VOTE mail message should contain the commit ID the tag points at,
for better traceability.
2014-12-19 11:23:29 +01:00
e99c2f8940
Don't use branch name in Git buildnumber.
...
As we want to build the artifacts from a clean checkout of a tag, we are
in 'detached head' state, without a branch name.
The build number will retain the most important thing for traceability:
the last commit SHA id.
2014-12-19 10:56:29 +01:00
fc7ee95d9c
Fixed some commands and URLS in the release documentation.
2014-12-19 00:30:20 +01:00
cf4a9d70c9
creating release candidate
2014-12-18 21:29:08 +01:00
1a96973119
Added the command to check Git tag GPG signature
2014-12-18 21:28:04 +01:00
20c577868c
Improved release process documentation with Git.
2014-12-18 21:11:35 +01:00
e27f6e71f2
Update release howto for Git-based release.
2014-12-17 19:03:42 +01:00
3d97f85484
Prepare link gor 3.4 javadoc.
2014-12-17 19:03:14 +01:00
f9437bc642
Candidate release date ...
2014-12-17 19:02:35 +01:00
3b58a282b9
Prevent RAT plugin from complaining about spurious development files.
2014-12-17 19:01:56 +01:00
2ed43de903
Setting up realistic limits for JaCoCo.
...
I think we never met the previous default settings, but did not notice
it in the maven logs before.
2014-12-17 19:01:27 +01:00
88516d95eb
Removed unnecessary attribution, as per discussion on developers list.
2014-12-17 12:01:42 +01:00
44401d183d
Made it clear that NaNs or infinities may yield NaN in moment computations. JIRA: MATH-1146.
2014-12-16 20:45:33 -07:00
26e6114583
Made it clear that NaNs or infinities may yield NaN in moment computations. JIRA: MATH-1146.
2014-12-16 16:49:16 -07:00
d9b951c1fa
Merge branch 'master' of https://git-wip-us.apache.org/repos/asf/commons-math
2014-12-16 09:24:12 -07:00
ae1cbdbd6d
Fixed cut/paste error in NOTICE.
2014-12-16 09:23:48 -07:00
17c903385c
Replace svn-based build number with git-based build number.
2014-12-16 17:03:40 +01:00
f9125d0b87
Updated LICENSE, NOTICE files with netlib attributions for BesselJ. JIRA: MATH-1066.
2014-12-16 08:54:42 -07:00
75872ffe9b
fixed typos in error messages translations
2014-12-16 13:37:58 +01:00
7655cb61e3
Including Ekstazi profile (www.ekstazi.org) to optimize regression tests
2014-12-16 13:37:58 +01:00
8d4fd1dabb
Removed unnecessary call to "super()".
2014-12-16 13:25:10 +01:00
540aa2e7ed
Fixed message format formatting.
2014-12-15 16:48:59 -07:00
b723f82bd5
Added Bessel functions of the first kind, based on NetLib implementation.
...
JIRA: MATH-1066
Based on patch provided by Brian Wignall
2014-12-15 15:11:32 -07:00
376958763c
Merge branch 'master' of https://git-wip-us.apache.org/repos/asf/commons-math
2014-12-15 14:58:44 -07:00
f80f577748
Added Bessel functions of the first kind, based on NetLib implementation.
...
JIRA: MATH-1066
Based on patch provided by Brian Wignall
2014-12-15 13:48:07 -07:00
809f0f89cb
[MATH-1154] Add clarification to ctor javadoc of all distributions to pass a null random generator in case no sampling is needed.
2014-12-15 21:21:09 +01:00
59fe593a42
Javadoc
2014-12-15 17:06:20 +01:00
92a027bc73
Merge branch 'master' of https://git-wip-us.apache.org/repos/asf/commons-math
2014-12-15 16:57:10 +01:00
ed6109d0a2
Javadoc
2014-12-15 16:56:42 +01:00
d5f586bd4e
[MATH-1142] Improve performance of kalman gain calculation of KalmanFilter. Thanks to Arne Schwarz.
2014-12-15 16:50:50 +01:00
2fb2221d48
Fixed integer overflow in KolmogorovSmirnovTest causing 2-sample test
...
to use exact method when the product of the sample sizes exceeds
Integer.MAX_VALUE, resulting in effectively hung execution.
JIRA: MATH-1181
Reported by Gilad
2014-12-15 06:44:53 -07:00
7fb571b16c
Javadoc
2014-12-15 13:48:48 +01:00
de73ad3115
MATH-1180: Sequence of integers
...
Added to "MathArrays": method to create an array containing a sequence of integers.
2014-12-15 13:40:04 +01:00
dabc5ab661
Ignore backup files.
2014-12-14 20:22:36 +01:00
491786ce41
MATH-1172: Simple curve fitter
...
Provides boiler-plate code so that users can readily fit any parametric function.
2014-12-14 18:48:01 +01:00
753f278d10
MATH-1173: Tricubic interpolation
...
New class "TricubicInterpolator" to replace "TricubicSplineInterpolator".
2014-12-14 18:25:08 +01:00
09129d5367
prepared release notes for version 3.4
2014-12-13 22:04:33 +01:00
9f42f18f15
updated download page for the next release (3.4)
2014-12-13 21:41:51 +01:00
4ed24c87c5
fixed missing closing element
2014-12-13 21:40:00 +01:00
5d6e445484
MATH-1178 (typo in userguide)
...
Fixed incorrect example in "stat" section.
[Reported by Dmitriy.]
2014-12-09 00:13:28 +01:00
c8f122c304
Merge branch 'master' of https://git-wip-us.apache.org/repos/asf/commons-math
...
Conflicts:
src/changes/changes.xml
2014-12-03 12:30:26 +01:00
fd7a4362ca
MATH-1173: Tricubic interpolation
...
New class "TricubicInterpolatingFunction" to replace "TricubicSplineInterpolatingFunction".
2014-12-03 12:25:09 +01:00
10b1c517cd
Filter out spurious vertices in polygons boundaries.
2014-12-03 12:20:18 +01:00
17f52a2e55
MATH-1175
...
"LaplaceDistribution": Fixed special case of inverse cumulative distribution.
2014-12-03 12:06:03 +01:00
b4fb13bca9
Connect closest node instead of first node below threshold in polygons.
2014-12-03 10:25:44 +01:00
428a48533c
Fixed comment, the test for MATH-1174 does not fail anymore.
2014-12-02 22:35:27 +01:00
3f13fbeff3
Typo.
2014-12-02 22:29:45 +01:00
d21d3a94ab
Deprecated AVLTree and OrderedTuple.
...
These two classes were dirty hacks used only for the extraction of
boundary from 2D Euclidean BSP trees. They were not really efficient and
produces some problems as topology was not preserved.
A completely new implementation of boundary extraction has been set up
as of 3.4, that preserves more topological information and therefore
should be more robust. This new implementation does not use these
classes anymore.
As AVLTree is not really a math concept, this class is not considered to
belong to [math] scope. OrderedTuple is more a grey zone, but it is
really something inefficient and no need for it is foreseen.
2014-12-02 22:18:57 +01:00
e6aae3a8bf
Fixed a problem with too thin polygons considered to have infinite size.
2014-12-02 22:09:06 +01:00
893ef53c6c
Removed development traces.
2014-12-02 22:06:33 +01:00
Luc Maisonobe
Phil Steitz
Milos Gligoric
Gilles
Thomas Neidhart