1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
|
<?xml version="1.0"?>
<!-- $Author: hoschek3 $, $Revision: 1.1 $, $Date: 2004/02/25 23:25:18 $ -->
<document>
<properties>
<author email="whoschek.AT.lbl.DOT.gov">Wolfgang Hoschek</author>
<title>Changelog</title>
</properties>
<body>
<!-- ##################################################################################### -->
<section name="Version 1.2.0 (Sep 10, 2004)">
<ul>
<li>Removed com.isml.math.* package (If you need Complex numbers, download and use <a href="http://jakarta.apache.org/commons/math/userguide/complex.html"> Jakarta Commons Math</a> instead)</li>
<li>Documentation is now auto generated</li>
<li>Fixed Algebra.normInfinity(DoubleMatrix1D x) (T.J.Hunt@open.ac.uk)</li>
<li>Upgraded package util.concurrent from version 1.3.2 to 1.3.4 (package name edu.oswego... becomes EDU.oswego...)</li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.1.0 (Oct 29, 2003)">
<ul>
<li>This version breaks the <a href="ibm-1.4.1-2xpentium4-2.8GHz.txt"> 1.9 Gflop/s barrier</a>
on JDK ibm-1.4.1, RedHat 9.0, 2x IntelXeon@2.8 GHz.</li>
<li>Redesigned documentation and web site</li>
<li>Moved build system to ANT</li>
<li>Upgraded package util.concurrent from version 1.3.1 to 1.3.2, which is now a separate jar</li>
<li>Upgraded package corejava to corejava-2002-11-16, which is now licensed under LGPL.</li>
<li>Removed packages ViolinStrings, edu.cornell.lassp.houle.RngPack, jal.*</li>
<li>Colt now requires jdk 1.2.x</li>
<li>Fixed covariance bug in cern.jet.stat.Descriptive (scott white "scottw@eml.cc")</li>
<li>Fixed bug in cern.jet.random.Normal.nextDouble in combination with setState() (nick.collier@verizon.net)</li>
<li>Fixed bug in cern.jet.stat.quantile.EquiDepthHistogram (Brett Neumeier random@rnd.cx)</li>
<li>RCDoubleMatrix2D was unnecessarily restricted in shape (double)columns*rows
<= Integer.MAX_VALUE). Is now only limited by nonZeroes <= Integer.MAX_VALUE</li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.0.3 (Nov 22, 2002)">
<ul>
<li>This version breaks the <a href="ibm-1.4.1-2xpentium4-2.2GHz.txt"> 1.3 Gflop barrier</a>
on JDK ibm-1.4.1, RedHat 8.0, 2x Pentium4@2.2 GHz.</li>
<li>Upgraded package util.concurrent from version 1.3.0 to 1.3.1</li>
<li>Fixed bug in cern.jet.random.Uniform.nextInt (jan.cheyns@intec.rug.ac.be)</li>
<li>Fixed small bug in cern.colt.Partitioning and cern.colt.matrix.doublealgo.Partitioning</li>
<li>Fixed Poisson.staticNextInt()</li>
<li>Fixed infinite loop in Logarithmic.nextDouble() for some parameter combinations</li>
<li>Fix for CholeskyDecomposition.solve (MG Ferreira "mgf@webmail.co.za")</li>
<li>cern.jet.stat.quantile.QuantileFinderFactory should ignore N if if known_N
is false.</li>
<li>class PrintFormat is unreliable - do not use it. matrix print formatting
is now reverted back to use corejava.Format instead.</li>
<li> MersenneTwister had a <a target="_top" href="http://www.math.keio.ac.jp/matumoto/emt.html">small
problem</a>, which is now fixed with Matsumotos code from 2002/1/26. </li>
<li>cern.jet.stat.Probability.studentT and cern.jet.random.StudentT were only
defined for values >= 0</li>
<li>cern.colt.bitvector.BitVector.replaceFromToWith(int from, int to, boolean
value) treated leading partial unit incorrectly (Olivier Janssens "oj@adhoc.be")
</li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.0.2 (Nov 08, 2001)">
<ul>
<li>Added GenericSorting.mergeSort for "stable" in-place sorting of
arbitrary shaped data</li>
<li>Upgraded the ViolinStrings package to V1.0. Michael Schmeling reports that
two split methods which were contributed by Eric Jablow have been added and
some minor documentation bugs have been fixed.</li>
<li>Upgraded package util.concurrent from version 1.2.5 to 1.3.0</li>
<li> Added 2d matrix versions of daxpy(), dcopy(), dscal(), dswap() to Blas,
SeqBlas, SmpBlas (Martin.Schimschak@dresdner-bank.com)</li>
<li>MersenneTwister now also allows zero seed (Martin.Schimschak@dresdner-bank.com)</li>
<li>Fix for serious SparseDoubleMatrix2D and Open<type>Map performance degradation</li>
<li>RCDoubleMatrix2D.assign(otherMatrix) got faster</li>
<li>RCDoubleMatrix2D.assign(double) got faster</li>
<li>DenseDoubleMatrix{1,2,3}D.assign(double) got faster</li>
<li>DenseDoubleMatrix2D.zMult got a bit faster for very large problem sizes</li>
<li>Algebra.normInfinity(DoubleMatrix1D x) and Algebra.normInfinity(DoubleMatrix2D
A) were incorrect (Martin.Schimschak@dresdner-bank.com)</li>
<li>SeqBlas.dger() and SeqBlas.dtrmv() were incorrect (Martin.Schimschak@dresdner-bank.com)</li>
<li>ObjectArrayList.toList() threw IndexOutOfBoundsException</li>
<li>RCDoubleMatrix2D was unnecessarily restricted in shape (double)columns*rows
<= Integer.MAX_VALUE). Is now only limited by nonZeroes <= Integer.MAX_VALUE</li>
<li>cern.colt.matrix.linalg.Property.equals(..) gave wrong results when comparing
exotic special cases NaN, inf, -inf.</li>
<li>cern.colt.matrix.doublealgo.Sorting was not NaN aware. In compliance with
the JDK sorts, it now swaps NaNs to the end.</li>
<li>Property.isIdentity bug fix</li>
<li>Polynomial.p1evl javadoc was misleading</li>
<li>Added matrix mergesorts. Matrix quicksorts alone were not enough, because
matrices are frequently sorted successively by multiple columns. In order
to preserve the relative order of equal elements a so-called "stable"
sort was needed. Mergesort is such a thing.
<blockquote>
<p>To enable this, some minor changes were necessary. Please update your
source code as follows</p>
<table width="75%" border="1" cellspacing="0">
<tr valign="top">
<td><b>Old style</b></td>
<td><b>New style</b></td>
</tr>
<tr valign="top">
<td height="59"><tt>cern.colt.matrix.doublealgo.Sorting.quickSort(...)</tt></td>
<td height="59">
<p><tt>cern.colt.matrix.doublealgo.Sorting.quickSort.sort(...)</tt>
or
<tt>cern.colt.matrix.doublealgo.Sorting.mergeSort.sort(...)</tt></p>
</td>
</tr>
</table>
</blockquote>
</li>
<li>matrix.viewSorted(...) methods now use mergesort instead of quicksort.</li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.0.1 (May 15, 2000)">
<ul>
<li>New blocked algorithm improved performance of out-of-cache matrix-matrix
and matrix-vector mult (Blas.dgemm and dgemv).</li>
<li>Added parallel implementations of assign(function) to the BLAS (Basic Linear
Algebra System). See <tt>cern.colt.matrix.linalg</tt>.</li>
<li>Fixed a bad bug in GenericPermuting. This one caused incorrect results to
LU solving.</li>
<li>SmpBlas.dgemm and dgemv threw exceptions when asking for transpositions.</li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.0.0">
<ul>
<li>Powerful, extendible and performant histogram abstractions, improving ways
for the statistical study of experimental data. </li>
<li>Matrix functionality and performance improvements</li>
<li>Sequential and parallel implementations of the BLAS (Basic Linear Algebra
System). See <tt>cern.colt.matrix.linalg</tt>.</li>
<li>Sparse row compressed matrix implementation - <tt>RCDoubleMatrix2D</tt></li>
<li>Major performance improvements for many matrix operations, in particular
optimizations detecting function objects for scaling (+,-,*,/, ...), delegating
to optimized internal implementations</li>
<li>For performance added generalized matrix-matrix and matrix-vector mult:
<tt>C = alpha*A*B + beta*C </tt>and <tt>z = alpha*A*y + beta*z</tt> with or
without transpositions</li>
<li>Added the long awaited histogram packages <tt>hep.aida </tt>, <tt>hep.aida.ref</tt>
and <tt>hep.aida.bin</tt>. The stable bins formerly found in cern.jet.histo
have been moved to hep.aida.bin and cern.jet.histo has been dropped. </li>
<li>Added histogram and OLAP cube operators to <tt>cern.colt.matrix.doublealgo.Statistic</tt>.</li>
<li>Added sampling views to cern.colt.matrix.doublealgo.Statistic </li>
<li>Efficient matrix quicksort using precomputation; added to <tt>cern.colt.matrix.doublealgo.Sorting</tt></li>
<li>Stencil methods for finite difference operations in <tt>cern.colt.matrix.doublealgo.Stencil</tt></li>
<li>DoubleMatrix1D.getNonZeros(IntArrayList indexList, DoubleArrayList valueList,
int maxCardinality) had a bug, hence Algebra.inverse sometimes produced wrong
answers. </li>
<li>QRDecomposition</li>
<li>matrix.viewStrides() sometimes gave wrong shape</li>
<li>matrix.zMult(...) sometimes threw ArrayIndexOutOfBoundsExceptions</li>
<li>DynamicBin1D.max() sometimes gave wrong answer -infinity</li>
<li>Distance matrix in cern.colt.matrix.doublealgo.Statistic now works on user
defined distance functions</li>
<li>Moved cern.colt.matrix.DoubleTransform to cern.colt.matrix.doublealgo.Transform
(sorry) </li>
</ul>
</section>
<!-- ##################################################################################### -->
<section name="Version 1.0 Beta4">
<ul>
<li>Renamed to lower case: EDU.oswego.cs.dl.util.concurrent --> edu.oswego.cs.dl.util.concurrent (resolves collisions of multiple edu packages)</li>
<li>cern.colt.matrix.linalg.Algebra.trace() now returns a double.</li>
<li>Improved build process (makefile)</li>
<li>package cern.colt.matrix - bug fixes and performance improvements</li>
<li>Stronger separation of matrix interfaces and matrix implementations:
<ul>
<li>the implementations of matrix data structures moved from package cern.colt.matrix
to package cern.colt.matrix.imp</li>
<li>matrix factories moved from cern.colt.matrix.algo to cern.colt.matrix</li>
</ul>
</li>
<li>Formatter objects - Matrix formatting much extended and now also possible
using java.lang.Double.toString(double)</li>
<li>Moved cern.jet.math.Unit and cern.jet.math.PhysicalConstants to cern.clhep
</li>
<li>Uniform.nextDouble() now returns a value between min and max, as specified
upon instance construction</li>
<li> package cern.colt.matrix.linalg - Linear Algebra </li>
<li>Additional convenience methods in matrix factories</li>
<li>cern.colt.matrix.DoubleTransform: Additional convenience methods for element-by-element
transformations</li>
<li>Introduced Object matrices</li>
<li>A few additional methods in DynamicBin1D related to sampling, bootstrapping,
correlation</li>
<li>Performance log of compute intensive matrix computations</li>
<li>Now also redistributing <tt>util.concurrent</tt> framework - Standardized,
efficient utility classes commonly encountered in concurrent programming</li>
<li>package cern.clhep - Starting to port CLHEP (C++)</li>
</ul>
</section>
</body>
</document>
|
