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
|
pyem (0.5.6) Thu, 16 Nov 2006 21:02:02 +0900
* correct examples
* correct exceptions msg strings in gauss_mix, which
were buggy
* add examples from website to the package, so that above errors
do not appear again
pyem (0.5.6) Thu, 16 Nov 2006 14:18:19 +0900
* bump to 0.5.6
* Add __str__ and __repr__ for GM and GMM classes
* Add regularization method (but not used yet).
* Change 'f<8' to N.float64 for ctype enabled densities
* Move 'Magic numbers' into a separated python file, misc.py
pyem (0.5.5) Tue, 24 Oct 2006 18:30:54 +0900
* Fix a bug inmultiple_gaussian_den which prevents
full covariance mode to work in GMM.
* Add consistency tests in tests/test_gmm_em.py to avoid the
above bug from being detected from now on.
pyem (0.5.5) Mon, 23 Oct 2006 18:48:15 +0900
* Bump to 0.5.5
* Added bic method to GMM.
* A few fixes for docstrings
* Added bibliography to the doc
pyem (0.5.4) Fri, 20 Oct 2006 12:52:01 +0900
* Bump to 0.5.4
* Online EM is basically implemented, with tests. The API
should be fixed to be choherent (lacks the Trainer Class, too).
The class is not imported by default, still (available as _OnGMM)
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.3) Thu, 12 Oct 2006 21:08:21 +0900
* Change the layout and setup.py for inclusion to scipy.
* Initial script for online em.
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.3) Tue, 03 Oct 2006 18:28:13 +0900
* Update tests so that they work within the numpy test framework
* Update version to 0.5.3
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.2) Tue, 29 Aug 2006 14:53:49 +0900
* Add a class method to GM to create a model directly from
w, mean and var values (uses decorator: python 2.4)
* Add a plot1d method to GM to plot a GM in one dimension (where
the confidence ell kind of plot does not make sense). This draws
each Gaussian pdf, fill the area on the confidence interval
(matplotlib is really cool)
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.2) Mon, 28 Aug 2006 13:20:13 +0900
* Add a plot method to Gm class, so that it is easier
to plot a GM model interactively (needs matplotlib)
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.2) Thu, 24 Aug 2006 19:42:28 +0900
* put version to 0.5.2
* Correct a bug with init method in GMM (change class method
to object method).
* modify the setup for a more flexible system
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5.1) Thu, 17 Aug 2006 11:54:41 +0900
* put version to 0.5.1
* Update to last numpy (change axis args between 1.0b1 and 1.0b2,
change type args of ones and zeros)
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.5) Fri, 04 Aug 2006 23:10:37 +0900
* put version to 0.5.0
* implement confidence interval using chi2
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.4) Fri, 04 Aug 2006 19:37:47 +0900
* put version to 0.4.2
* adapt to new version of numpy (1.0b2SVN)
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.4) Fri, 14 Jul 2006 17:49:57 +0900
* put version to 0.4.1
* change some import due to recent changes in
numpy
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.4) Fri, 14 Jul 2006 16:24:05 +0900
* put version to 0.4
* Heavy refactoring of EM and GMM into classes (see below)
* add a module gauss_mix, which implements Gaussian Mixtures.
* GMM is now a class, which is initialized using a Gaussian Mixture.
a GMM can be trained.
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
pyem (0.3) Thu, 13 Jul 2006 19:48:54 +0900
* put version to 0.3
* Refactoring kmean code into new module.
* Refactoring tests code into test module.
* Replace matrixmultiply and outerproduct calls by dot to use fast BLAS if
available. Not everything is done yet
-- David Cournapeau <david@ar.media.kyoto-u.ac.jp>
|
