baohan
/
scikit-learn
mirror of https://github.com/scikit-learn/scikit-learn.git
0
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
16,706 Commits 31 Branches 142 Tags 892 MiB
Commit Graph

2224 Commits

Author SHA1 Message Date
Arnaud Joly 689058a23e Merge pull request #3412 from brentp/logistic_l1_l2_ex 
[MRG] use more interesting range for C in logistic l1 l2 example.
 2014-07-17 16:54:43 +02:00
Andreas Mueller 5b247f90e4 move around examples for better structure. 2014-07-17 09:35:30 +02:00
Brent Pedersen c8705478e7 use more interesting range for C in logistic l1 l2 example. 
Before this change, output is (http://scikit-learn.org/dev/auto_examples/linear_model/plot_logistic_l1_l2_sparsity.html#example-linear-model-plot-logistic-l1-l2-sparsity-py):
C=10
Sparsity with L1 penalty: 6.25%
score with L1 penalty: 0.9104
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.9093
C=100
Sparsity with L1 penalty: 6.25%
score with L1 penalty: 0.9098
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.9098
C=1000
Sparsity with L1 penalty: 4.69%
score with L1 penalty: 0.9098
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.9098

With this change, output is:
C=100.00
Sparsity with L1 penalty: 6.25%
score with L1 penalty: 0.9110
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.9098
C=1.00
Sparsity with L1 penalty: 9.38%
score with L1 penalty: 0.9104
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.9093
C=0.01
Sparsity with L1 penalty: 85.94%
score with L1 penalty: 0.8625
Sparsity with L2 penalty: 4.69%
score with L2 penalty: 0.8915
 2014-07-16 15:55:48 -06:00
Andreas Mueller d146d23d93 Catch ConvergenceWarning in RandomizedL1 2014-07-16 15:55:09 +02:00
Olivier Grisel 5cab6c9e5c DOC add missing plt.figure() for multiplot support in ROC example 2014-06-25 14:57:11 +02:00
Olivier Grisel dae905c52f DOC remove spurious plt.clf() from ROC example 2014-06-25 14:21:57 +02:00
Olivier Grisel e331935d2d FIX Python 3 support for examples/applications/plot_prediction_latency.py 2014-06-24 16:16:54 +02:00
Olivier Grisel 45a90568bc FIX Python 3 support for examples/applications/plot_out_of_core_classification.py 2014-06-24 14:00:35 +02:00
Olivier Grisel f892622967 FIX Python 3 support for datasets.species_distributions 2014-06-24 10:39:41 +02:00
Olivier Grisel 30347e9a80 FIX python 3 compat for plot_bias_variance.py 2014-06-23 11:04:39 +02:00
Félix-Antoine Fortin b16ffbb04d DOC: fix 2 covariance examples rst math markup. 
Some formulas in plot_mahalanobis_distances and
plot_robust_vs_empirical_covariance were not correctly rendered by sphinx
as some characters were not properly escaped. This is fixed by setting the
docstrings as raw string using r"""

There was also missing braces for the number of features and samples subscript.
 2014-06-06 11:44:36 +02:00
Joel Nothman fc0150d08d DOC/FIX Address @arjoly's comments 2014-06-05 11:20:36 +02:00
Joel Nothman 18c399e20a MAINT deprecate sequences of sequences support 
* add warnings
 * provide alternative binarizer, sklearn.preprocessing.MultiLabelBinarizer
 * fix documentation
 2014-06-05 11:20:35 +02:00
Alexandre Gramfort 0a4ba728ff Merge pull request #2822 from AlexanderFabisch/tsne 
[MRG+1] t-SNE
 2014-06-04 13:53:29 +02:00
bwignall 9bee5b4d2f CLN: Capitalize "Dirichlet" and "Mexican" in example docstrings 2014-06-01 17:43:21 -04:00
bwignall 34fc4d4214 CLN: Capitalize "Gaussian" in example docstrings 2014-06-01 16:53:23 -04:00
bwignall c76a7a195c CLN: Fix typo 2014-05-30 20:08:08 -04:00
Alexander Fabisch f59e08f21f Correct examples 2014-05-28 20:04:46 +02:00
Alexander Fabisch 08c17d24c5 Use PCA initialization in examples 2014-05-28 19:56:10 +02:00
Alexander Fabisch 91e95df436 Integrate PCA initialization 
* update docstring
* add tests
 2014-05-28 19:56:09 +02:00
Olivier Grisel 58bcbaf0bd Add optional PCA init to t-SNE 2014-05-28 19:56:09 +02:00
Alexander Fabisch 25a8c5f05a Adress Gael's comments 
* rename _binary_search to _utils
* indent list elements
* use default parameters in examples
 2014-05-28 19:56:09 +02:00
Alexander Fabisch 8d0fadabbc Remove generalization 2014-05-28 19:56:09 +02:00
Alexander Fabisch 1b668d471c Explain example 2014-05-28 19:56:09 +02:00
Alexander Fabisch 43b41f1ff7 Modify learning schedule 2014-05-28 19:56:09 +02:00
Alexander Fabisch 5de4c7ddc6 Remove example with digits dataset 2014-05-28 19:56:09 +02:00
Alexander Fabisch 8c969cb011 Add t-SNE to other examples 2014-05-28 19:56:09 +02:00
Alexander Fabisch 53e015c961 Simplify optimization 
* remove L-BFGS, gradient descent finds better visualizations (although the cost is often higher)
* learning_rate is a parameter
* learning schedule:
    * initialize embedding with standard deviation 1
    * 100 episodes with early exaggeration
    * 20 iterations with momentum 0.5
    * then switch to momentum 0.8
    * default learning rate is 100, reasonable range is 100-500
 2014-05-28 19:54:15 +02:00
Alexander Fabisch 4f2154a3e7 Compare t-SNE with other manifold learners 2014-05-28 19:54:15 +02:00
Alexander Fabisch 7a1f84eb69 Generalize with average of nearest neighbors 2014-05-28 19:51:08 +02:00
Alexander Fabisch da43e4c230 Add t-SNE 2014-05-28 19:51:08 +02:00
ugurthemaster b929656592 Update plot_adaboost_regression.py 
Library imports have been moved to top of the file.
 2014-05-26 09:48:50 +03:00
Olivier Grisel 40597f1df5 Merge pull request #3169 from mjbommar/issue-3167-eradicate-todense 
PR re: issue 3167 to eradicate .todense()
 2014-05-24 21:02:47 +02:00
Olivier Grisel 1a05e6d3f7 FIX: euclidean divide in plot_image_denoising.py for Py3 support 2014-05-23 13:33:13 +02:00
mjbommar d8099d6d27 Replacing the non-test .todense() methods with .toarray() 2014-05-20 21:16:34 -04:00
Gilles Louppe 8523766e32 Merge pull request #3156 from IvicaJovic/fixes 
DOC Fix wrong confidence interval
 2014-05-18 17:18:05 +02:00
Gael Varoquaux 974fb95915 Remove HMM examples 
Fixes #2646
 2014-05-18 15:54:29 +02:00
IvicaJovic 8dc8673525 fix wrong confidence interval 2014-05-17 22:28:30 +02:00
Rajat Khanduja 7263d58dcf Modified example 'plot_stock_market.py' to use matplotlib.pyplot 
Modified despite #3153 because
- Error is not related to pylab or matplotlib.pyplot
- Removing the symbol 'DELL' from the example solves the problem
 2014-05-16 09:43:50 +05:30
Rajat Khanduja cca6fc7e8b Some more examples updated to use matplotlib.pyplot 2014-05-16 09:42:02 +05:30
Rajat Khanduja 7d10d98492 Fixed some PEP8 violations even present in master branch, in examples/ 
This fixes some of the violations that were not caused by the change to pyplot
from pylab.
 2014-05-15 09:35:24 +05:30
Rajat Khanduja 5e20b8c664 Fixed pep8 violations. Some 'line too long' errors still remain. 2014-05-15 09:23:16 +05:30
Rajat Khanduja 2244098532 Updated examples to use pyplot for plotting instead of pylab 
Some examples not updated because scripts throw error (probably not
because of the change)
 2014-05-15 02:18:07 +05:30
t-aft 9d18b3bfc0 DOC grid_search_digits.py does no do nested CV 
Fixes #3136.
 2014-05-10 17:16:48 +02:00
Arnaud Joly 3a4edd0f5c Merge pull request #2936 from ugurthemaster/master 
Update plot_svm_regression.py
 2014-04-25 21:45:06 +02:00
Olivier Grisel 9792624483 COSMIT: use plural, there are 2 learning curves 2014-04-19 12:38:16 -04:00
Olivier Grisel 56a251c21f Merge pull request #3083 from sdenton4/enh_learning_curves 
[MRG+2] Improved learning curves
 2014-04-17 22:06:58 -04:00
sdenton4 a63da37826 Fixed style errors detected by pep8. 2014-04-17 16:24:23 -04:00
sdenton4 e4f9b5b4b0 Improved the learning curve example. 
Scaled y-axis to make it easier to compare the two methods.
Wrapped duplicate code into a utility function that users can copy, paste
and adapt in their own project.
Added more iterations to cross-validate the GaussianNB model to get a smoother
curve.
 2014-04-17 15:59:15 -04:00
Michael Becker 00bb7a42f9 TruncatedSVD: Calculate explained variance. 
Fixes #3047
 2014-04-17 15:10:34 -04:00