baohan
/
scikit-learn
mirror of https://github.com/scikit-learn/scikit-learn.git
0
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
10,497 Commits 31 Branches 142 Tags 892 MiB
Commit Graph

33 Commits

Author SHA1 Message Date
James Bergstra 5d6540d054 ENH: renaming iter_limit -> max_iter 2012-10-16 16:13:00 +02:00
James Bergstra 812376955a ENH: adding iter_limit to libsvm 
This commit adds support for a new libsvm parameter, which permits setting a
hard limit on the while loop in the svm.cpp file, in the Solver::Solve
function. Without this functionality, it has been observed that libsvm hangs
without converging in a reasonable time -- it can take days instead of
seconds, and perhaps it is in fact an infinite loop, I don't know.  This
commit allows the user to set a hard limit on the number of libsvm iterations,
and modifies the libsvm bindings to raise a new exception of SolverTimeout
when this limit is reached.
 2012-10-16 16:13:00 +02:00
Andreas Mueller 254e15fa55 ENH fix error by removing unnecessary test. 2012-09-03 11:02:14 +02:00
Nelle Varoquaux 4631a2192b FIX sparse OneClassSVM was using the wrong parameter 2012-09-02 20:47:46 +02:00
Andreas Mueller ee7d78ce47 Make sparse svm base class ABC 2012-06-26 14:16:04 +02:00
Andreas Mueller 66d00f4748 MISC trying to remove scale_C 2012-05-05 21:57:04 +02:00
Nick Wilson c90a329080 Add verbose parameter to SVMs (fixes #250) 
Add a verbose parameter to SVMs to get verbose output from libsvm
printed to stdout while fitting a model.

Note that this setting takes advantage of a per-process runtime setting
in libsvm that, if enabled, may not work properly in a multithreaded
context.
 2012-03-15 00:16:06 -07:00
Nick Wilson 0321ded656 Stop passing keyword arguments for positional args 
The base SVM class constructors take positional arguments, but some of
their subclasses were passing some of the arguments in as keyword
arguments.
 2012-03-14 23:15:54 -07:00
Andreas Mueller 0e8312312b ENH: moved class_weight parameter in svms from fit to ``__init__``. 2012-02-07 00:24:02 +01:00
Lars Buitinck 72707abf3f ENH merge sparse and dense SVMs, part 2 
* keep sparse.* behavior: will transform dense data into sparse data
  when fitting (we might want to deprecate this along with the classes)
* document sparsity support in classes.py
 2012-01-27 17:41:25 +01:00
Lars Buitinck bc6255e8e7 ENH merge sparse and dense SVMs, part 1 
* enable sparse matrix support in formerly dense versions
* remove DenseBaseLibSVM
* move sparse.libsvm to libsvm_sparse
 2012-01-27 17:41:22 +01:00
Andreas Mueller 5c2a8696e3 COSMIT removed unused imports 2012-01-23 21:13:55 +01:00
Lars Buitinck 263d6c02f7 ENH merge dense/sparse LinearSVC, part 1: no more SparseBaseLibLinear 2012-01-16 17:55:47 +01:00
Lars Buitinck a6e7178d20 COSMIT remove superfluous imports in svm/sparse/base.py 2012-01-08 22:17:21 +01:00
Alexandre Gramfort db74b538e5 Merge remote-tracking branch 'origin/master' into n_samples_scaling 
Conflicts:
	sklearn/svm/sparse/base.py
	sklearn/svm/tests/test_sparse.py
 2011-12-18 12:45:58 +01:00
Alexandre Gramfort 8ec89a9fc2 ENH : C scaling of sparse models 2011-12-18 12:39:26 +01:00
Andreas Mueller 25c93f4af6 COSMIT: pep8 2011-12-15 17:24:53 +01:00
Alexandre Gramfort fdbd584a93 s/C_scale_n_samples/scale_C 2011-11-11 21:36:19 -05:00
Alexandre Gramfort 6def64c3e6 ENH : adding support for scaling by n_samples of the C parameter in libsvm and liblinear models 2011-11-11 16:43:03 -05:00
Fabian Pedregosa 09677e0193 FIX: do not use reshape in libsvm sparse bindings. 
This should fix issues of people using this in parallel.
 2011-11-07 15:05:37 +01:00
Andy 07b4121ea3 BUG: minor mistake in earlier commit. 2011-11-04 14:34:51 +01:00
Andy 8878ebe9da ENH Sparse SVM: removed cache_size parameter from fit method. Is now part of constructur. 2011-11-04 13:34:54 +01:00
Andy d961d79d52 BUG: cache_size also for sparse SVMs 2011-11-03 17:28:43 +01:00
Olivier Grisel 95d497f550 Merge branch 'master' of github.com:scikit-learn/scikit-learn 2011-11-01 12:26:23 +01:00
Olivier Grisel 8c65edf86a ENH: more informative error messages when input has invalid shapes 2011-11-01 12:25:13 +01:00
Lars Buitinck 3bde2bf97b ENH accept matrix input throughout 
Simple solution: replace np.asanyarray with np.asarray throughout.
Also, use utils.as_float_array where appropriate.
 2011-10-21 13:51:12 +02:00
Lars Buitinck 87c8664d1a FIX Issue 379 and use the opportunity to refactor libsvm code 
Sparse SVC's would try to use the dense decision_function, since
SparseBaseLibSVM did not override that method. Solution: factor out
non-common code to DenseBaseLibSVM. (@diogojc's test program still
crashes, but with an appropriate error message.)

Also, leverage superclass __init__ in SparseBaseLibSVM
 2011-10-09 04:28:56 +02:00
Peter Prettenhofer 3b34162524 fix: use None as default value for class_weight and sample_weight for sparse OneClassSVM; ample_weight -> sample_weight 2011-10-02 18:21:46 +02:00
Peter Prettenhofer 178c6e255a check if y.shape[0] == X.shape[0]; this is especially troublesome for svm.sparse because most people are not aware of the sparse matrix - KFold troubles.. 2011-10-02 17:49:53 +02:00
Fabian Pedregosa d89cee1d44 Use 1 / n_features as value for gamma. 
This fixes https://github.com/scikit-learn/scikit-learn/issues/332
 2011-09-13 12:21:08 +02:00
Fabian Pedregosa ddf4b72109 Move project directory from scikits.learn to sklearn 2011-09-02 12:06:57 +02:00
Fabian Pedregosa 68f27e3790 Revert "Move project directory from scikits.learn to sklearn" 
This reverts commit fd0d3b879d.
 2011-09-02 12:03:18 +02:00
Fabian Pedregosa fd0d3b879d Move project directory from scikits.learn to sklearn 2011-09-02 11:38:24 +02:00