Skip to content
This repository
Browse code

Changelog for 0.7

  • Loading branch information...
commit a084f38160618cfb8009e6b25eeb8d566e66dbcc 1 parent 4b00058
Fabian Pedregosa authored February 28, 2011
3  .mailmap
@@ -13,4 +13,5 @@ Peter Prettenhofer <peter.prettenhofer@gmail.com> pprett <peter.prettenhofer@gma
13 13
 Anne-Laure Fouque <afouque@is208050.(none)> Anne-Laure FOUQUE <af216607@is206635.intra.cea.fr>
14 14
 Vincent Dubourg <vincent.dubourg@gmail.com> dubourg <vincent.dubourg@gmail.com>
15 15
 Vincent Dubourg <vincent.dubourg@gmail.com> dubourg <dubourg@PTlami14.(none)>
16  
-Christian Osendorfer <osendorf@gmail.com> osdf <osendorf@gmail.com> 
  16
+Christian Osendorfer <osendorf@gmail.com> osdf <osendorf@gmail.com> 
  17
+James Bergstra <james.bergstra@gmail.com> james.bergstra <james.bergstra@gmail.com>
2  doc/modules/classes.rst
Source Rendered
@@ -46,6 +46,7 @@ Generalized Linear Models
46 46
 
47 47
    linear_model.LinearRegression
48 48
    linear_model.Ridge
  49
+   linear_model.RidgeCV
49 50
    linear_model.Lasso
50 51
    linear_model.LassoCV
51 52
    linear_model.ElasticNet
@@ -114,6 +115,7 @@ Nearest Neighbors
114 115
    :toctree: generated/
115 116
    :template: function.rst
116 117
 
  118
+   neighbors.kneighbors_graph
117 119
    ball_tree.knn_brute
118 120
 
119 121
 
100  doc/whats_new.rst
Source Rendered
... ...
@@ -1,6 +1,94 @@
1 1
 
2 2
 .. currentmodule:: scikits.learn
3 3
 
  4
+
  5
+.. _changes_0_7:
  6
+
  7
+0.7
  8
+===
  9
+
  10
+scikits.learn 0.7 was released in March 2011, roughly three months
  11
+after the 0.6 release. This release is marked by the speed
  12
+improvements in existing algorithms, like k-Nearest Neighbors and
  13
+K-Means algorithm and by the inclusion of an efficient algorithm for
  14
+computing the Ridge Generalized Cross Validation solution. Unlike the
  15
+preceding release, no new modules where added to this release.
  16
+
  17
+Changelog
  18
+---------
  19
+
  20
+  - Performance improvements for Gaussian Mixture Model sampling [Jan
  21
+    Schlüter].
  22
+
  23
+  - Implementation of efficient leave-one-out cross-validated Ridge in
  24
+    :class:`linear_model.RidgeCV` [`Mathieu Blondel`_ ]
  25
+
  26
+  - Better handling of collinearity and early stopping in
  27
+    :func:`linear_model.lars_path` [`Alexandre Gramfort`_ and `Fabian
  28
+    Pedregosa`_ ].
  29
+
  30
+  - Fixes for liblinear ordering of labels and sign of coefficients
  31
+    [Dan Yamins, Paolo Losi, `Mathieu Blondel`_ and `Fabian Pedregosa`_ ].
  32
+
  33
+  - Performance improvements for Nearest Neighbors algorithm in
  34
+    high-dimensional spaces [`Fabian Pedregosa`_ ].
  35
+
  36
+  - Performance improvements for :class:`cluster.KMeans` [`Gael
  37
+    Varoquaux`_ and `James Bergstra`_ ].
  38
+
  39
+  - Sanity checks for SVM-based classes [`Mathieu Blondel`_ ].
  40
+
  41
+  - Refactoring of :class:`neighbors.NeighborsClassifier` and
  42
+    :func:`neighbors.kneighbors_graph`: added different algorithms for
  43
+    the k-Nearest Neighbor Search and implemented a more stable
  44
+    algorithm for finding barycenter weigths [`Fabian Pedregosa`_ ].
  45
+
  46
+  - Documentation improvements: Added
  47
+    :class:`pca.RandomizedPCA`. :class:`linear_model.LogisticRegression`. Also
  48
+    added references of matrices used for clustering and other fixes
  49
+    [`Gael Varoquaux`_, `Fabian Pedregosa`_, `Mathieu Blondel`_,
  50
+    `Olivier Grisel`_, Virgile Fritsch , Emmanuelle Gouillart ]
  51
+
  52
+  - Binded decision_function in classes that make use of liblinear_,
  53
+    dense and sparse variants, like :class:`svm.LinearSVC` or
  54
+    :class:`linear_model.LogisticRegression` [`Fabian Pedregosa`_ ].
  55
+
  56
+  - Performance and API improvements to
  57
+    :func:`metrics.euclidian_distances` and to
  58
+    :class:`pca.RandomizedPCA` [`James Bergstra`_ ] 
  59
+
  60
+  - Fix compilation issues under NetBSD [Kamel Ibn Hassen Derouiche]
  61
+
  62
+  - Allow input sequences of different lengths in :class:`hmm.GaussianHMM`
  63
+    [`Ron Weiss`_ ]
  64
+
  65
+  - Fix bug in affinity propagation caused by incorrect indexing [mxf]
  66
+
  67
+
  68
+People
  69
+------
  70
+
  71
+People that made this release possible preceeded by number of commits:
  72
+
  73
+    - 85  `Fabian Pedregosa`_
  74
+    - 67  `Mathieu Blondel`_
  75
+    - 20  `Alexandre Gramfort`_
  76
+    - 19  `James Bergstra`_
  77
+    - 14  Dan Yamins
  78
+    - 13  `Olivier Grisel`_
  79
+    - 12  `Gael Varoquaux`_
  80
+    - 4  Edouard Duchesnay
  81
+    - 4  `Ron Weiss`_
  82
+    - 2  Satrajit Ghosh
  83
+    - 2  `Vincent Dubourg`_
  84
+    - 1  Emmanuelle Gouillart
  85
+    - 1  Kamel Ibn Hassen Derouiche
  86
+    - 1  Paolo Losi
  87
+    - 1  VirgileFritsch
  88
+    - 1  `Yaroslav Halchenko`_
  89
+    - 1  mxf
  90
+
  91
+
4 92
 .. _changes_0_6:
5 93
 
6 94
 0.6
@@ -299,3 +387,15 @@ of commits):
299 387
     *  2  Vincent Michel
300 388
     *  1  Chris Filo Gorgolewski
301 389
 
  390
+
  391
+.. _Alexandre Gramfort: http://www-sop.inria.fr/members/Alexandre.Gramfort/index.fr.html
  392
+
  393
+.. _Fabian Pedregosa: http://fseoane.net/blog/
  394
+
  395
+.. _Mathieu Blondel: http://www.mblondel.org/journal/
  396
+
  397
+.. _James Bergstra: http://www-etud.iro.umontreal.ca/~bergstrj/
  398
+
  399
+.. _liblinear: http://www.csie.ntu.edu.tw/~cjlin/liblinear/
  400
+
  401
+.. _Yaroslav Halchenko: http://www.onerussian.com/

0 notes on commit a084f38

Fabian Pedregosa

wow you are pretty.

Olivier Grisel

That's nothing, Alexandre is a youtube star: http://www.youtube.com/watch?v=CbGKRRVtGjI almost as popular as the Skateboarding Dog .

Fabian Pedregosa

amazing ...

Alexandre Gramfort

183 views... Impressive ! :)

Mathieu Blondel

"Alexandre Gramfort, Prix thèse Fondation EADS 2010" Retweeted by ogrisel and 100+ others.

Yaroslav Halchenko

I demand a version of the video in English ! ;)

Please sign in to comment.
Something went wrong with that request. Please try again.